Transport Layer: UDP and TCP CS491G: Computer Networking Lab V. Arun Transport Layer 3-1 Slides...
-
Upload
miles-marsh -
Category
Documents
-
view
238 -
download
7
Transcript of Transport Layer: UDP and TCP CS491G: Computer Networking Lab V. Arun Transport Layer 3-1 Slides...
Transport Layer UDP and TCP
CS491G Computer Networking Lab
V Arun
Transport Layer 3-1Slides adapted from Kurose and Ross
Transport Layer 3-2
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-3
Transport services and protocols
provide logical communication between app processes running on different hosts
transport protocols run in end systems send side breaks app
messages into segments passes to network layer
recv side reassembles segments into messages passes to app layer
more than one transport protocol available to apps Internet TCP and UDP
application
transportnetworkdata linkphysical
logical end-end transport
application
transportnetworkdata linkphysical
Transport Layer 3-4
Transport vs network layer network layer
logical communication between hosts
transport layer logical communication between processes relies on and
enhances network layer services
12 kids in Annrsquos house sending letters to 12 kids in Billrsquos house
hosts = houses processes = kids app messages =
letters in envelopes transport protocol =
Ann and Bill who demux to in-house siblings
network-layer protocol = postal service
household analogy
Transport Layer 3-5
Internet transport-layer protocols
reliable in-order delivery (TCP) congestion control flow control connection setup
unreliable unordered delivery UDP no-frills extension of
ldquobest-effortrdquo IP services not
available delay guarantees bandwidth
guarantees
application
transportnetworkdata linkphysical
application
transportnetworkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
logical end-end transport
Transport Layer 3-6
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-7
Multiplexingdemultiplexing
process
socket
use header info to deliverreceived segments to correct socket
demultiplexing at receiverhandle data from multiplesockets add transport header (later used for demultiplexing)
multiplexing at sender
transport
application
physical
link
network
P2P1
transport
application
physical
link
network
P4
transport
application
physical
link
network
P3
Transport Layer 3-8
How demultiplexing works
host receives IP datagrams each datagram has source and
destination IP address each datagram carries one
transport-layer segment each segment has source and
destination port number host uses IP addresses amp port
numbers to direct segment to right socket
source port dest port
32 bits
applicationdata
(payload)
other header fields
TCPUDP segment format
Transport Layer 3-9
Connectionless demultiplexing
recall created socket has host-local port
DatagramSocket mySocket1 = new DatagramSocket(12534)
when host receives UDP segment checks destination IP
and port in segment directs UDP segment
to socket bound to that (IPport)
recall when creating datagram to send into UDP socket must specify
destination IP address destination port IP datagrams with same dest (IP port) but different source IP addresses andor source port numbers will be directed to same socket
Transport Layer 3-10
Connectionless demux example
DatagramSocket serverSocket = new DatagramSocket
(6428)
transport
application
physical
link
network
P3transport
application
physical
link
network
P1
transport
application
physical
link
network
P4
DatagramSocket mySocket1 = new DatagramSocket (5775)
DatagramSocket mySocket2 = new DatagramSocket (9157)
source port 9157dest port 6428
source port 6428dest port 9157
source port dest port
source port dest port
Transport Layer 3-11
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
demux receiver uses all four values to direct segment to right socket
server host has many simultaneous TCP sockets each socket identified
by its own 4-tuple web servers have
different socket each client non-persistent HTTP
will have different socket for each request
Transport Layer 3-12
Connection-oriented demux example
transport
application
physical
link
network
P3transport
app
physical
link
P4
transport
application
physical
link
network
P2
source IPport A9157dest IP port B80
source IPport B80dest IPport A9157
host IP address
A
host IP address
C
network
P6P5P3
source IPport C5775dest IPport B80
source IPport C9157dest IPport B80
three segments all destined to IP address B dest port 80 are demultiplexed to different sockets
server IP
address B
server socket also port 80
Transport Layer 3-13
Connection-oriented demux example
transport
application
physical
link
network
P3transport
app
physical
link
transport
application
physical
link
network
P2
source IPport A9157dest IP port B80
source IPport B80dest IPport A9157
host IP address
A
host IP address
C
server IP
address B
network
P3
source IPport C5775dest IPport B80
source IPport C9157dest IPport B80
P4
server socket also port 80threaded server
Transport Layer 3-14
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-15
UDP User Datagram Protocol [RFC 768] no frills bare bones
transport protocol for ldquobest effortrdquo service UDP segments may be lost delivered out-of-
order connectionless
no sender-receiver handshaking
each UDP segment handled independently
UDP uses streaming
multimedia apps (loss tolerant rate sensitive)
DNS SNMP
reliable transfer over UDP add reliability at
application layer application-specific
error recovery
Transport Layer 3-16
UDP segment header
source port dest port
32 bits
applicationdata
(payload)
UDP segment format
length checksum
length in bytes of UDP segment
including header
no connection establishment (which can add delay)
simple no connection state at sender receiver
small header size no congestion control
UDP can blast away as fast as desired
why is there a UDP
Transport Layer 3-17
UDP checksum
sender treat segment contents
including header fields as sequence of 16-bit integers
checksum addition (onersquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (flipped bits) in segments
Transport Layer 3-18
Internet checksum exampleexample add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Note when adding numbers a carryout from the most significant bit needs to be added to the
result
Q1 Sockets and multiplexing TCP uses more information in packet
headers in order to demultiplex packets compared to UDPA TrueB False
Transport Layer 3-19
Q2 Sockets UDP Suppose we use UDP instead of TCP under
HTTP for designing a web server where all requests and responses fit in a single packet Suppose a 100 clients are simultaneously communicating with this web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-20
Q3 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-21
Q4 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server Do all of the sockets at the server have the same server-side port numberA YesB No
Transport Layer 3-22
Q5 UDP checksums Letrsquos denote a UDP packet as (checksum
data) ignoring other fields for this question Suppose a sender sends (0010 1110) and the receiver receives (00111110) Which of the following is true of the receiverA Thinks the packet is corrupted and
discards the packetB Thinks only the checksum is corrupted
and delivers the correct data to the application
C Can possibly conclude that nothing is wrong with the packet
D A and CTransport Layer 3-23
Transport Layer 3-24
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-25
TCP Overview RFCs 79311221323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) inits sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order
byte steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
Transport Layer 3-26
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement number
receive window
Urg data pointerchecksum
FSRPAUheadlen
notused
options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 3-27
TCP seq numbers ACKssequence numbers
byte stream ldquonumberrdquo of first byte in segmentrsquos data
acknowledgementsseq of next byte expected from other side
cumulative ACKQ how receiver handles out-of-order segmentsA TCP spec doesnrsquot say - up to implementor
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent ACKed
sent not-yet ACKed(ldquoin-flightrdquo)
usablebut not yet sent
not usable
window size N
sender sequence number space
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
outgoing segment from sender
Transport Layer 3-28
TCP seq numbers ACKs
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt oflsquoCrsquo echoesback lsquoCrsquo
simple telnet scenario
Host BHost A
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43 data = lsquoCrsquo
Seq=43 ACK=80
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-2
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-3
Transport services and protocols
provide logical communication between app processes running on different hosts
transport protocols run in end systems send side breaks app
messages into segments passes to network layer
recv side reassembles segments into messages passes to app layer
more than one transport protocol available to apps Internet TCP and UDP
application
transportnetworkdata linkphysical
logical end-end transport
application
transportnetworkdata linkphysical
Transport Layer 3-4
Transport vs network layer network layer
logical communication between hosts
transport layer logical communication between processes relies on and
enhances network layer services
12 kids in Annrsquos house sending letters to 12 kids in Billrsquos house
hosts = houses processes = kids app messages =
letters in envelopes transport protocol =
Ann and Bill who demux to in-house siblings
network-layer protocol = postal service
household analogy
Transport Layer 3-5
Internet transport-layer protocols
reliable in-order delivery (TCP) congestion control flow control connection setup
unreliable unordered delivery UDP no-frills extension of
ldquobest-effortrdquo IP services not
available delay guarantees bandwidth
guarantees
application
transportnetworkdata linkphysical
application
transportnetworkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
logical end-end transport
Transport Layer 3-6
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-7
Multiplexingdemultiplexing
process
socket
use header info to deliverreceived segments to correct socket
demultiplexing at receiverhandle data from multiplesockets add transport header (later used for demultiplexing)
multiplexing at sender
transport
application
physical
link
network
P2P1
transport
application
physical
link
network
P4
transport
application
physical
link
network
P3
Transport Layer 3-8
How demultiplexing works
host receives IP datagrams each datagram has source and
destination IP address each datagram carries one
transport-layer segment each segment has source and
destination port number host uses IP addresses amp port
numbers to direct segment to right socket
source port dest port
32 bits
applicationdata
(payload)
other header fields
TCPUDP segment format
Transport Layer 3-9
Connectionless demultiplexing
recall created socket has host-local port
DatagramSocket mySocket1 = new DatagramSocket(12534)
when host receives UDP segment checks destination IP
and port in segment directs UDP segment
to socket bound to that (IPport)
recall when creating datagram to send into UDP socket must specify
destination IP address destination port IP datagrams with same dest (IP port) but different source IP addresses andor source port numbers will be directed to same socket
Transport Layer 3-10
Connectionless demux example
DatagramSocket serverSocket = new DatagramSocket
(6428)
transport
application
physical
link
network
P3transport
application
physical
link
network
P1
transport
application
physical
link
network
P4
DatagramSocket mySocket1 = new DatagramSocket (5775)
DatagramSocket mySocket2 = new DatagramSocket (9157)
source port 9157dest port 6428
source port 6428dest port 9157
source port dest port
source port dest port
Transport Layer 3-11
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
demux receiver uses all four values to direct segment to right socket
server host has many simultaneous TCP sockets each socket identified
by its own 4-tuple web servers have
different socket each client non-persistent HTTP
will have different socket for each request
Transport Layer 3-12
Connection-oriented demux example
transport
application
physical
link
network
P3transport
app
physical
link
P4
transport
application
physical
link
network
P2
source IPport A9157dest IP port B80
source IPport B80dest IPport A9157
host IP address
A
host IP address
C
network
P6P5P3
source IPport C5775dest IPport B80
source IPport C9157dest IPport B80
three segments all destined to IP address B dest port 80 are demultiplexed to different sockets
server IP
address B
server socket also port 80
Transport Layer 3-13
Connection-oriented demux example
transport
application
physical
link
network
P3transport
app
physical
link
transport
application
physical
link
network
P2
source IPport A9157dest IP port B80
source IPport B80dest IPport A9157
host IP address
A
host IP address
C
server IP
address B
network
P3
source IPport C5775dest IPport B80
source IPport C9157dest IPport B80
P4
server socket also port 80threaded server
Transport Layer 3-14
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-15
UDP User Datagram Protocol [RFC 768] no frills bare bones
transport protocol for ldquobest effortrdquo service UDP segments may be lost delivered out-of-
order connectionless
no sender-receiver handshaking
each UDP segment handled independently
UDP uses streaming
multimedia apps (loss tolerant rate sensitive)
DNS SNMP
reliable transfer over UDP add reliability at
application layer application-specific
error recovery
Transport Layer 3-16
UDP segment header
source port dest port
32 bits
applicationdata
(payload)
UDP segment format
length checksum
length in bytes of UDP segment
including header
no connection establishment (which can add delay)
simple no connection state at sender receiver
small header size no congestion control
UDP can blast away as fast as desired
why is there a UDP
Transport Layer 3-17
UDP checksum
sender treat segment contents
including header fields as sequence of 16-bit integers
checksum addition (onersquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (flipped bits) in segments
Transport Layer 3-18
Internet checksum exampleexample add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Note when adding numbers a carryout from the most significant bit needs to be added to the
result
Q1 Sockets and multiplexing TCP uses more information in packet
headers in order to demultiplex packets compared to UDPA TrueB False
Transport Layer 3-19
Q2 Sockets UDP Suppose we use UDP instead of TCP under
HTTP for designing a web server where all requests and responses fit in a single packet Suppose a 100 clients are simultaneously communicating with this web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-20
Q3 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-21
Q4 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server Do all of the sockets at the server have the same server-side port numberA YesB No
Transport Layer 3-22
Q5 UDP checksums Letrsquos denote a UDP packet as (checksum
data) ignoring other fields for this question Suppose a sender sends (0010 1110) and the receiver receives (00111110) Which of the following is true of the receiverA Thinks the packet is corrupted and
discards the packetB Thinks only the checksum is corrupted
and delivers the correct data to the application
C Can possibly conclude that nothing is wrong with the packet
D A and CTransport Layer 3-23
Transport Layer 3-24
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-25
TCP Overview RFCs 79311221323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) inits sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order
byte steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
Transport Layer 3-26
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement number
receive window
Urg data pointerchecksum
FSRPAUheadlen
notused
options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 3-27
TCP seq numbers ACKssequence numbers
byte stream ldquonumberrdquo of first byte in segmentrsquos data
acknowledgementsseq of next byte expected from other side
cumulative ACKQ how receiver handles out-of-order segmentsA TCP spec doesnrsquot say - up to implementor
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent ACKed
sent not-yet ACKed(ldquoin-flightrdquo)
usablebut not yet sent
not usable
window size N
sender sequence number space
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
outgoing segment from sender
Transport Layer 3-28
TCP seq numbers ACKs
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt oflsquoCrsquo echoesback lsquoCrsquo
simple telnet scenario
Host BHost A
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43 data = lsquoCrsquo
Seq=43 ACK=80
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-3
Transport services and protocols
provide logical communication between app processes running on different hosts
transport protocols run in end systems send side breaks app
messages into segments passes to network layer
recv side reassembles segments into messages passes to app layer
more than one transport protocol available to apps Internet TCP and UDP
application
transportnetworkdata linkphysical
logical end-end transport
application
transportnetworkdata linkphysical
Transport Layer 3-4
Transport vs network layer network layer
logical communication between hosts
transport layer logical communication between processes relies on and
enhances network layer services
12 kids in Annrsquos house sending letters to 12 kids in Billrsquos house
hosts = houses processes = kids app messages =
letters in envelopes transport protocol =
Ann and Bill who demux to in-house siblings
network-layer protocol = postal service
household analogy
Transport Layer 3-5
Internet transport-layer protocols
reliable in-order delivery (TCP) congestion control flow control connection setup
unreliable unordered delivery UDP no-frills extension of
ldquobest-effortrdquo IP services not
available delay guarantees bandwidth
guarantees
application
transportnetworkdata linkphysical
application
transportnetworkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
logical end-end transport
Transport Layer 3-6
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-7
Multiplexingdemultiplexing
process
socket
use header info to deliverreceived segments to correct socket
demultiplexing at receiverhandle data from multiplesockets add transport header (later used for demultiplexing)
multiplexing at sender
transport
application
physical
link
network
P2P1
transport
application
physical
link
network
P4
transport
application
physical
link
network
P3
Transport Layer 3-8
How demultiplexing works
host receives IP datagrams each datagram has source and
destination IP address each datagram carries one
transport-layer segment each segment has source and
destination port number host uses IP addresses amp port
numbers to direct segment to right socket
source port dest port
32 bits
applicationdata
(payload)
other header fields
TCPUDP segment format
Transport Layer 3-9
Connectionless demultiplexing
recall created socket has host-local port
DatagramSocket mySocket1 = new DatagramSocket(12534)
when host receives UDP segment checks destination IP
and port in segment directs UDP segment
to socket bound to that (IPport)
recall when creating datagram to send into UDP socket must specify
destination IP address destination port IP datagrams with same dest (IP port) but different source IP addresses andor source port numbers will be directed to same socket
Transport Layer 3-10
Connectionless demux example
DatagramSocket serverSocket = new DatagramSocket
(6428)
transport
application
physical
link
network
P3transport
application
physical
link
network
P1
transport
application
physical
link
network
P4
DatagramSocket mySocket1 = new DatagramSocket (5775)
DatagramSocket mySocket2 = new DatagramSocket (9157)
source port 9157dest port 6428
source port 6428dest port 9157
source port dest port
source port dest port
Transport Layer 3-11
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
demux receiver uses all four values to direct segment to right socket
server host has many simultaneous TCP sockets each socket identified
by its own 4-tuple web servers have
different socket each client non-persistent HTTP
will have different socket for each request
Transport Layer 3-12
Connection-oriented demux example
transport
application
physical
link
network
P3transport
app
physical
link
P4
transport
application
physical
link
network
P2
source IPport A9157dest IP port B80
source IPport B80dest IPport A9157
host IP address
A
host IP address
C
network
P6P5P3
source IPport C5775dest IPport B80
source IPport C9157dest IPport B80
three segments all destined to IP address B dest port 80 are demultiplexed to different sockets
server IP
address B
server socket also port 80
Transport Layer 3-13
Connection-oriented demux example
transport
application
physical
link
network
P3transport
app
physical
link
transport
application
physical
link
network
P2
source IPport A9157dest IP port B80
source IPport B80dest IPport A9157
host IP address
A
host IP address
C
server IP
address B
network
P3
source IPport C5775dest IPport B80
source IPport C9157dest IPport B80
P4
server socket also port 80threaded server
Transport Layer 3-14
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-15
UDP User Datagram Protocol [RFC 768] no frills bare bones
transport protocol for ldquobest effortrdquo service UDP segments may be lost delivered out-of-
order connectionless
no sender-receiver handshaking
each UDP segment handled independently
UDP uses streaming
multimedia apps (loss tolerant rate sensitive)
DNS SNMP
reliable transfer over UDP add reliability at
application layer application-specific
error recovery
Transport Layer 3-16
UDP segment header
source port dest port
32 bits
applicationdata
(payload)
UDP segment format
length checksum
length in bytes of UDP segment
including header
no connection establishment (which can add delay)
simple no connection state at sender receiver
small header size no congestion control
UDP can blast away as fast as desired
why is there a UDP
Transport Layer 3-17
UDP checksum
sender treat segment contents
including header fields as sequence of 16-bit integers
checksum addition (onersquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (flipped bits) in segments
Transport Layer 3-18
Internet checksum exampleexample add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Note when adding numbers a carryout from the most significant bit needs to be added to the
result
Q1 Sockets and multiplexing TCP uses more information in packet
headers in order to demultiplex packets compared to UDPA TrueB False
Transport Layer 3-19
Q2 Sockets UDP Suppose we use UDP instead of TCP under
HTTP for designing a web server where all requests and responses fit in a single packet Suppose a 100 clients are simultaneously communicating with this web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-20
Q3 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-21
Q4 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server Do all of the sockets at the server have the same server-side port numberA YesB No
Transport Layer 3-22
Q5 UDP checksums Letrsquos denote a UDP packet as (checksum
data) ignoring other fields for this question Suppose a sender sends (0010 1110) and the receiver receives (00111110) Which of the following is true of the receiverA Thinks the packet is corrupted and
discards the packetB Thinks only the checksum is corrupted
and delivers the correct data to the application
C Can possibly conclude that nothing is wrong with the packet
D A and CTransport Layer 3-23
Transport Layer 3-24
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-25
TCP Overview RFCs 79311221323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) inits sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order
byte steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
Transport Layer 3-26
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement number
receive window
Urg data pointerchecksum
FSRPAUheadlen
notused
options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 3-27
TCP seq numbers ACKssequence numbers
byte stream ldquonumberrdquo of first byte in segmentrsquos data
acknowledgementsseq of next byte expected from other side
cumulative ACKQ how receiver handles out-of-order segmentsA TCP spec doesnrsquot say - up to implementor
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent ACKed
sent not-yet ACKed(ldquoin-flightrdquo)
usablebut not yet sent
not usable
window size N
sender sequence number space
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
outgoing segment from sender
Transport Layer 3-28
TCP seq numbers ACKs
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt oflsquoCrsquo echoesback lsquoCrsquo
simple telnet scenario
Host BHost A
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43 data = lsquoCrsquo
Seq=43 ACK=80
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-4
Transport vs network layer network layer
logical communication between hosts
transport layer logical communication between processes relies on and
enhances network layer services
12 kids in Annrsquos house sending letters to 12 kids in Billrsquos house
hosts = houses processes = kids app messages =
letters in envelopes transport protocol =
Ann and Bill who demux to in-house siblings
network-layer protocol = postal service
household analogy
Transport Layer 3-5
Internet transport-layer protocols
reliable in-order delivery (TCP) congestion control flow control connection setup
unreliable unordered delivery UDP no-frills extension of
ldquobest-effortrdquo IP services not
available delay guarantees bandwidth
guarantees
application
transportnetworkdata linkphysical
application
transportnetworkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
logical end-end transport
Transport Layer 3-6
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-7
Multiplexingdemultiplexing
process
socket
use header info to deliverreceived segments to correct socket
demultiplexing at receiverhandle data from multiplesockets add transport header (later used for demultiplexing)
multiplexing at sender
transport
application
physical
link
network
P2P1
transport
application
physical
link
network
P4
transport
application
physical
link
network
P3
Transport Layer 3-8
How demultiplexing works
host receives IP datagrams each datagram has source and
destination IP address each datagram carries one
transport-layer segment each segment has source and
destination port number host uses IP addresses amp port
numbers to direct segment to right socket
source port dest port
32 bits
applicationdata
(payload)
other header fields
TCPUDP segment format
Transport Layer 3-9
Connectionless demultiplexing
recall created socket has host-local port
DatagramSocket mySocket1 = new DatagramSocket(12534)
when host receives UDP segment checks destination IP
and port in segment directs UDP segment
to socket bound to that (IPport)
recall when creating datagram to send into UDP socket must specify
destination IP address destination port IP datagrams with same dest (IP port) but different source IP addresses andor source port numbers will be directed to same socket
Transport Layer 3-10
Connectionless demux example
DatagramSocket serverSocket = new DatagramSocket
(6428)
transport
application
physical
link
network
P3transport
application
physical
link
network
P1
transport
application
physical
link
network
P4
DatagramSocket mySocket1 = new DatagramSocket (5775)
DatagramSocket mySocket2 = new DatagramSocket (9157)
source port 9157dest port 6428
source port 6428dest port 9157
source port dest port
source port dest port
Transport Layer 3-11
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
demux receiver uses all four values to direct segment to right socket
server host has many simultaneous TCP sockets each socket identified
by its own 4-tuple web servers have
different socket each client non-persistent HTTP
will have different socket for each request
Transport Layer 3-12
Connection-oriented demux example
transport
application
physical
link
network
P3transport
app
physical
link
P4
transport
application
physical
link
network
P2
source IPport A9157dest IP port B80
source IPport B80dest IPport A9157
host IP address
A
host IP address
C
network
P6P5P3
source IPport C5775dest IPport B80
source IPport C9157dest IPport B80
three segments all destined to IP address B dest port 80 are demultiplexed to different sockets
server IP
address B
server socket also port 80
Transport Layer 3-13
Connection-oriented demux example
transport
application
physical
link
network
P3transport
app
physical
link
transport
application
physical
link
network
P2
source IPport A9157dest IP port B80
source IPport B80dest IPport A9157
host IP address
A
host IP address
C
server IP
address B
network
P3
source IPport C5775dest IPport B80
source IPport C9157dest IPport B80
P4
server socket also port 80threaded server
Transport Layer 3-14
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-15
UDP User Datagram Protocol [RFC 768] no frills bare bones
transport protocol for ldquobest effortrdquo service UDP segments may be lost delivered out-of-
order connectionless
no sender-receiver handshaking
each UDP segment handled independently
UDP uses streaming
multimedia apps (loss tolerant rate sensitive)
DNS SNMP
reliable transfer over UDP add reliability at
application layer application-specific
error recovery
Transport Layer 3-16
UDP segment header
source port dest port
32 bits
applicationdata
(payload)
UDP segment format
length checksum
length in bytes of UDP segment
including header
no connection establishment (which can add delay)
simple no connection state at sender receiver
small header size no congestion control
UDP can blast away as fast as desired
why is there a UDP
Transport Layer 3-17
UDP checksum
sender treat segment contents
including header fields as sequence of 16-bit integers
checksum addition (onersquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (flipped bits) in segments
Transport Layer 3-18
Internet checksum exampleexample add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Note when adding numbers a carryout from the most significant bit needs to be added to the
result
Q1 Sockets and multiplexing TCP uses more information in packet
headers in order to demultiplex packets compared to UDPA TrueB False
Transport Layer 3-19
Q2 Sockets UDP Suppose we use UDP instead of TCP under
HTTP for designing a web server where all requests and responses fit in a single packet Suppose a 100 clients are simultaneously communicating with this web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-20
Q3 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-21
Q4 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server Do all of the sockets at the server have the same server-side port numberA YesB No
Transport Layer 3-22
Q5 UDP checksums Letrsquos denote a UDP packet as (checksum
data) ignoring other fields for this question Suppose a sender sends (0010 1110) and the receiver receives (00111110) Which of the following is true of the receiverA Thinks the packet is corrupted and
discards the packetB Thinks only the checksum is corrupted
and delivers the correct data to the application
C Can possibly conclude that nothing is wrong with the packet
D A and CTransport Layer 3-23
Transport Layer 3-24
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-25
TCP Overview RFCs 79311221323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) inits sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order
byte steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
Transport Layer 3-26
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement number
receive window
Urg data pointerchecksum
FSRPAUheadlen
notused
options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 3-27
TCP seq numbers ACKssequence numbers
byte stream ldquonumberrdquo of first byte in segmentrsquos data
acknowledgementsseq of next byte expected from other side
cumulative ACKQ how receiver handles out-of-order segmentsA TCP spec doesnrsquot say - up to implementor
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent ACKed
sent not-yet ACKed(ldquoin-flightrdquo)
usablebut not yet sent
not usable
window size N
sender sequence number space
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
outgoing segment from sender
Transport Layer 3-28
TCP seq numbers ACKs
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt oflsquoCrsquo echoesback lsquoCrsquo
simple telnet scenario
Host BHost A
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43 data = lsquoCrsquo
Seq=43 ACK=80
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-5
Internet transport-layer protocols
reliable in-order delivery (TCP) congestion control flow control connection setup
unreliable unordered delivery UDP no-frills extension of
ldquobest-effortrdquo IP services not
available delay guarantees bandwidth
guarantees
application
transportnetworkdata linkphysical
application
transportnetworkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
logical end-end transport
Transport Layer 3-6
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-7
Multiplexingdemultiplexing
process
socket
use header info to deliverreceived segments to correct socket
demultiplexing at receiverhandle data from multiplesockets add transport header (later used for demultiplexing)
multiplexing at sender
transport
application
physical
link
network
P2P1
transport
application
physical
link
network
P4
transport
application
physical
link
network
P3
Transport Layer 3-8
How demultiplexing works
host receives IP datagrams each datagram has source and
destination IP address each datagram carries one
transport-layer segment each segment has source and
destination port number host uses IP addresses amp port
numbers to direct segment to right socket
source port dest port
32 bits
applicationdata
(payload)
other header fields
TCPUDP segment format
Transport Layer 3-9
Connectionless demultiplexing
recall created socket has host-local port
DatagramSocket mySocket1 = new DatagramSocket(12534)
when host receives UDP segment checks destination IP
and port in segment directs UDP segment
to socket bound to that (IPport)
recall when creating datagram to send into UDP socket must specify
destination IP address destination port IP datagrams with same dest (IP port) but different source IP addresses andor source port numbers will be directed to same socket
Transport Layer 3-10
Connectionless demux example
DatagramSocket serverSocket = new DatagramSocket
(6428)
transport
application
physical
link
network
P3transport
application
physical
link
network
P1
transport
application
physical
link
network
P4
DatagramSocket mySocket1 = new DatagramSocket (5775)
DatagramSocket mySocket2 = new DatagramSocket (9157)
source port 9157dest port 6428
source port 6428dest port 9157
source port dest port
source port dest port
Transport Layer 3-11
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
demux receiver uses all four values to direct segment to right socket
server host has many simultaneous TCP sockets each socket identified
by its own 4-tuple web servers have
different socket each client non-persistent HTTP
will have different socket for each request
Transport Layer 3-12
Connection-oriented demux example
transport
application
physical
link
network
P3transport
app
physical
link
P4
transport
application
physical
link
network
P2
source IPport A9157dest IP port B80
source IPport B80dest IPport A9157
host IP address
A
host IP address
C
network
P6P5P3
source IPport C5775dest IPport B80
source IPport C9157dest IPport B80
three segments all destined to IP address B dest port 80 are demultiplexed to different sockets
server IP
address B
server socket also port 80
Transport Layer 3-13
Connection-oriented demux example
transport
application
physical
link
network
P3transport
app
physical
link
transport
application
physical
link
network
P2
source IPport A9157dest IP port B80
source IPport B80dest IPport A9157
host IP address
A
host IP address
C
server IP
address B
network
P3
source IPport C5775dest IPport B80
source IPport C9157dest IPport B80
P4
server socket also port 80threaded server
Transport Layer 3-14
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-15
UDP User Datagram Protocol [RFC 768] no frills bare bones
transport protocol for ldquobest effortrdquo service UDP segments may be lost delivered out-of-
order connectionless
no sender-receiver handshaking
each UDP segment handled independently
UDP uses streaming
multimedia apps (loss tolerant rate sensitive)
DNS SNMP
reliable transfer over UDP add reliability at
application layer application-specific
error recovery
Transport Layer 3-16
UDP segment header
source port dest port
32 bits
applicationdata
(payload)
UDP segment format
length checksum
length in bytes of UDP segment
including header
no connection establishment (which can add delay)
simple no connection state at sender receiver
small header size no congestion control
UDP can blast away as fast as desired
why is there a UDP
Transport Layer 3-17
UDP checksum
sender treat segment contents
including header fields as sequence of 16-bit integers
checksum addition (onersquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (flipped bits) in segments
Transport Layer 3-18
Internet checksum exampleexample add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Note when adding numbers a carryout from the most significant bit needs to be added to the
result
Q1 Sockets and multiplexing TCP uses more information in packet
headers in order to demultiplex packets compared to UDPA TrueB False
Transport Layer 3-19
Q2 Sockets UDP Suppose we use UDP instead of TCP under
HTTP for designing a web server where all requests and responses fit in a single packet Suppose a 100 clients are simultaneously communicating with this web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-20
Q3 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-21
Q4 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server Do all of the sockets at the server have the same server-side port numberA YesB No
Transport Layer 3-22
Q5 UDP checksums Letrsquos denote a UDP packet as (checksum
data) ignoring other fields for this question Suppose a sender sends (0010 1110) and the receiver receives (00111110) Which of the following is true of the receiverA Thinks the packet is corrupted and
discards the packetB Thinks only the checksum is corrupted
and delivers the correct data to the application
C Can possibly conclude that nothing is wrong with the packet
D A and CTransport Layer 3-23
Transport Layer 3-24
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-25
TCP Overview RFCs 79311221323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) inits sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order
byte steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
Transport Layer 3-26
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement number
receive window
Urg data pointerchecksum
FSRPAUheadlen
notused
options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 3-27
TCP seq numbers ACKssequence numbers
byte stream ldquonumberrdquo of first byte in segmentrsquos data
acknowledgementsseq of next byte expected from other side
cumulative ACKQ how receiver handles out-of-order segmentsA TCP spec doesnrsquot say - up to implementor
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent ACKed
sent not-yet ACKed(ldquoin-flightrdquo)
usablebut not yet sent
not usable
window size N
sender sequence number space
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
outgoing segment from sender
Transport Layer 3-28
TCP seq numbers ACKs
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt oflsquoCrsquo echoesback lsquoCrsquo
simple telnet scenario
Host BHost A
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43 data = lsquoCrsquo
Seq=43 ACK=80
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-6
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-7
Multiplexingdemultiplexing
process
socket
use header info to deliverreceived segments to correct socket
demultiplexing at receiverhandle data from multiplesockets add transport header (later used for demultiplexing)
multiplexing at sender
transport
application
physical
link
network
P2P1
transport
application
physical
link
network
P4
transport
application
physical
link
network
P3
Transport Layer 3-8
How demultiplexing works
host receives IP datagrams each datagram has source and
destination IP address each datagram carries one
transport-layer segment each segment has source and
destination port number host uses IP addresses amp port
numbers to direct segment to right socket
source port dest port
32 bits
applicationdata
(payload)
other header fields
TCPUDP segment format
Transport Layer 3-9
Connectionless demultiplexing
recall created socket has host-local port
DatagramSocket mySocket1 = new DatagramSocket(12534)
when host receives UDP segment checks destination IP
and port in segment directs UDP segment
to socket bound to that (IPport)
recall when creating datagram to send into UDP socket must specify
destination IP address destination port IP datagrams with same dest (IP port) but different source IP addresses andor source port numbers will be directed to same socket
Transport Layer 3-10
Connectionless demux example
DatagramSocket serverSocket = new DatagramSocket
(6428)
transport
application
physical
link
network
P3transport
application
physical
link
network
P1
transport
application
physical
link
network
P4
DatagramSocket mySocket1 = new DatagramSocket (5775)
DatagramSocket mySocket2 = new DatagramSocket (9157)
source port 9157dest port 6428
source port 6428dest port 9157
source port dest port
source port dest port
Transport Layer 3-11
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
demux receiver uses all four values to direct segment to right socket
server host has many simultaneous TCP sockets each socket identified
by its own 4-tuple web servers have
different socket each client non-persistent HTTP
will have different socket for each request
Transport Layer 3-12
Connection-oriented demux example
transport
application
physical
link
network
P3transport
app
physical
link
P4
transport
application
physical
link
network
P2
source IPport A9157dest IP port B80
source IPport B80dest IPport A9157
host IP address
A
host IP address
C
network
P6P5P3
source IPport C5775dest IPport B80
source IPport C9157dest IPport B80
three segments all destined to IP address B dest port 80 are demultiplexed to different sockets
server IP
address B
server socket also port 80
Transport Layer 3-13
Connection-oriented demux example
transport
application
physical
link
network
P3transport
app
physical
link
transport
application
physical
link
network
P2
source IPport A9157dest IP port B80
source IPport B80dest IPport A9157
host IP address
A
host IP address
C
server IP
address B
network
P3
source IPport C5775dest IPport B80
source IPport C9157dest IPport B80
P4
server socket also port 80threaded server
Transport Layer 3-14
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-15
UDP User Datagram Protocol [RFC 768] no frills bare bones
transport protocol for ldquobest effortrdquo service UDP segments may be lost delivered out-of-
order connectionless
no sender-receiver handshaking
each UDP segment handled independently
UDP uses streaming
multimedia apps (loss tolerant rate sensitive)
DNS SNMP
reliable transfer over UDP add reliability at
application layer application-specific
error recovery
Transport Layer 3-16
UDP segment header
source port dest port
32 bits
applicationdata
(payload)
UDP segment format
length checksum
length in bytes of UDP segment
including header
no connection establishment (which can add delay)
simple no connection state at sender receiver
small header size no congestion control
UDP can blast away as fast as desired
why is there a UDP
Transport Layer 3-17
UDP checksum
sender treat segment contents
including header fields as sequence of 16-bit integers
checksum addition (onersquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (flipped bits) in segments
Transport Layer 3-18
Internet checksum exampleexample add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Note when adding numbers a carryout from the most significant bit needs to be added to the
result
Q1 Sockets and multiplexing TCP uses more information in packet
headers in order to demultiplex packets compared to UDPA TrueB False
Transport Layer 3-19
Q2 Sockets UDP Suppose we use UDP instead of TCP under
HTTP for designing a web server where all requests and responses fit in a single packet Suppose a 100 clients are simultaneously communicating with this web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-20
Q3 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-21
Q4 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server Do all of the sockets at the server have the same server-side port numberA YesB No
Transport Layer 3-22
Q5 UDP checksums Letrsquos denote a UDP packet as (checksum
data) ignoring other fields for this question Suppose a sender sends (0010 1110) and the receiver receives (00111110) Which of the following is true of the receiverA Thinks the packet is corrupted and
discards the packetB Thinks only the checksum is corrupted
and delivers the correct data to the application
C Can possibly conclude that nothing is wrong with the packet
D A and CTransport Layer 3-23
Transport Layer 3-24
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-25
TCP Overview RFCs 79311221323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) inits sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order
byte steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
Transport Layer 3-26
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement number
receive window
Urg data pointerchecksum
FSRPAUheadlen
notused
options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 3-27
TCP seq numbers ACKssequence numbers
byte stream ldquonumberrdquo of first byte in segmentrsquos data
acknowledgementsseq of next byte expected from other side
cumulative ACKQ how receiver handles out-of-order segmentsA TCP spec doesnrsquot say - up to implementor
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent ACKed
sent not-yet ACKed(ldquoin-flightrdquo)
usablebut not yet sent
not usable
window size N
sender sequence number space
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
outgoing segment from sender
Transport Layer 3-28
TCP seq numbers ACKs
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt oflsquoCrsquo echoesback lsquoCrsquo
simple telnet scenario
Host BHost A
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43 data = lsquoCrsquo
Seq=43 ACK=80
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-7
Multiplexingdemultiplexing
process
socket
use header info to deliverreceived segments to correct socket
demultiplexing at receiverhandle data from multiplesockets add transport header (later used for demultiplexing)
multiplexing at sender
transport
application
physical
link
network
P2P1
transport
application
physical
link
network
P4
transport
application
physical
link
network
P3
Transport Layer 3-8
How demultiplexing works
host receives IP datagrams each datagram has source and
destination IP address each datagram carries one
transport-layer segment each segment has source and
destination port number host uses IP addresses amp port
numbers to direct segment to right socket
source port dest port
32 bits
applicationdata
(payload)
other header fields
TCPUDP segment format
Transport Layer 3-9
Connectionless demultiplexing
recall created socket has host-local port
DatagramSocket mySocket1 = new DatagramSocket(12534)
when host receives UDP segment checks destination IP
and port in segment directs UDP segment
to socket bound to that (IPport)
recall when creating datagram to send into UDP socket must specify
destination IP address destination port IP datagrams with same dest (IP port) but different source IP addresses andor source port numbers will be directed to same socket
Transport Layer 3-10
Connectionless demux example
DatagramSocket serverSocket = new DatagramSocket
(6428)
transport
application
physical
link
network
P3transport
application
physical
link
network
P1
transport
application
physical
link
network
P4
DatagramSocket mySocket1 = new DatagramSocket (5775)
DatagramSocket mySocket2 = new DatagramSocket (9157)
source port 9157dest port 6428
source port 6428dest port 9157
source port dest port
source port dest port
Transport Layer 3-11
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
demux receiver uses all four values to direct segment to right socket
server host has many simultaneous TCP sockets each socket identified
by its own 4-tuple web servers have
different socket each client non-persistent HTTP
will have different socket for each request
Transport Layer 3-12
Connection-oriented demux example
transport
application
physical
link
network
P3transport
app
physical
link
P4
transport
application
physical
link
network
P2
source IPport A9157dest IP port B80
source IPport B80dest IPport A9157
host IP address
A
host IP address
C
network
P6P5P3
source IPport C5775dest IPport B80
source IPport C9157dest IPport B80
three segments all destined to IP address B dest port 80 are demultiplexed to different sockets
server IP
address B
server socket also port 80
Transport Layer 3-13
Connection-oriented demux example
transport
application
physical
link
network
P3transport
app
physical
link
transport
application
physical
link
network
P2
source IPport A9157dest IP port B80
source IPport B80dest IPport A9157
host IP address
A
host IP address
C
server IP
address B
network
P3
source IPport C5775dest IPport B80
source IPport C9157dest IPport B80
P4
server socket also port 80threaded server
Transport Layer 3-14
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-15
UDP User Datagram Protocol [RFC 768] no frills bare bones
transport protocol for ldquobest effortrdquo service UDP segments may be lost delivered out-of-
order connectionless
no sender-receiver handshaking
each UDP segment handled independently
UDP uses streaming
multimedia apps (loss tolerant rate sensitive)
DNS SNMP
reliable transfer over UDP add reliability at
application layer application-specific
error recovery
Transport Layer 3-16
UDP segment header
source port dest port
32 bits
applicationdata
(payload)
UDP segment format
length checksum
length in bytes of UDP segment
including header
no connection establishment (which can add delay)
simple no connection state at sender receiver
small header size no congestion control
UDP can blast away as fast as desired
why is there a UDP
Transport Layer 3-17
UDP checksum
sender treat segment contents
including header fields as sequence of 16-bit integers
checksum addition (onersquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (flipped bits) in segments
Transport Layer 3-18
Internet checksum exampleexample add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Note when adding numbers a carryout from the most significant bit needs to be added to the
result
Q1 Sockets and multiplexing TCP uses more information in packet
headers in order to demultiplex packets compared to UDPA TrueB False
Transport Layer 3-19
Q2 Sockets UDP Suppose we use UDP instead of TCP under
HTTP for designing a web server where all requests and responses fit in a single packet Suppose a 100 clients are simultaneously communicating with this web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-20
Q3 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-21
Q4 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server Do all of the sockets at the server have the same server-side port numberA YesB No
Transport Layer 3-22
Q5 UDP checksums Letrsquos denote a UDP packet as (checksum
data) ignoring other fields for this question Suppose a sender sends (0010 1110) and the receiver receives (00111110) Which of the following is true of the receiverA Thinks the packet is corrupted and
discards the packetB Thinks only the checksum is corrupted
and delivers the correct data to the application
C Can possibly conclude that nothing is wrong with the packet
D A and CTransport Layer 3-23
Transport Layer 3-24
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-25
TCP Overview RFCs 79311221323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) inits sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order
byte steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
Transport Layer 3-26
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement number
receive window
Urg data pointerchecksum
FSRPAUheadlen
notused
options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 3-27
TCP seq numbers ACKssequence numbers
byte stream ldquonumberrdquo of first byte in segmentrsquos data
acknowledgementsseq of next byte expected from other side
cumulative ACKQ how receiver handles out-of-order segmentsA TCP spec doesnrsquot say - up to implementor
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent ACKed
sent not-yet ACKed(ldquoin-flightrdquo)
usablebut not yet sent
not usable
window size N
sender sequence number space
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
outgoing segment from sender
Transport Layer 3-28
TCP seq numbers ACKs
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt oflsquoCrsquo echoesback lsquoCrsquo
simple telnet scenario
Host BHost A
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43 data = lsquoCrsquo
Seq=43 ACK=80
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-8
How demultiplexing works
host receives IP datagrams each datagram has source and
destination IP address each datagram carries one
transport-layer segment each segment has source and
destination port number host uses IP addresses amp port
numbers to direct segment to right socket
source port dest port
32 bits
applicationdata
(payload)
other header fields
TCPUDP segment format
Transport Layer 3-9
Connectionless demultiplexing
recall created socket has host-local port
DatagramSocket mySocket1 = new DatagramSocket(12534)
when host receives UDP segment checks destination IP
and port in segment directs UDP segment
to socket bound to that (IPport)
recall when creating datagram to send into UDP socket must specify
destination IP address destination port IP datagrams with same dest (IP port) but different source IP addresses andor source port numbers will be directed to same socket
Transport Layer 3-10
Connectionless demux example
DatagramSocket serverSocket = new DatagramSocket
(6428)
transport
application
physical
link
network
P3transport
application
physical
link
network
P1
transport
application
physical
link
network
P4
DatagramSocket mySocket1 = new DatagramSocket (5775)
DatagramSocket mySocket2 = new DatagramSocket (9157)
source port 9157dest port 6428
source port 6428dest port 9157
source port dest port
source port dest port
Transport Layer 3-11
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
demux receiver uses all four values to direct segment to right socket
server host has many simultaneous TCP sockets each socket identified
by its own 4-tuple web servers have
different socket each client non-persistent HTTP
will have different socket for each request
Transport Layer 3-12
Connection-oriented demux example
transport
application
physical
link
network
P3transport
app
physical
link
P4
transport
application
physical
link
network
P2
source IPport A9157dest IP port B80
source IPport B80dest IPport A9157
host IP address
A
host IP address
C
network
P6P5P3
source IPport C5775dest IPport B80
source IPport C9157dest IPport B80
three segments all destined to IP address B dest port 80 are demultiplexed to different sockets
server IP
address B
server socket also port 80
Transport Layer 3-13
Connection-oriented demux example
transport
application
physical
link
network
P3transport
app
physical
link
transport
application
physical
link
network
P2
source IPport A9157dest IP port B80
source IPport B80dest IPport A9157
host IP address
A
host IP address
C
server IP
address B
network
P3
source IPport C5775dest IPport B80
source IPport C9157dest IPport B80
P4
server socket also port 80threaded server
Transport Layer 3-14
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-15
UDP User Datagram Protocol [RFC 768] no frills bare bones
transport protocol for ldquobest effortrdquo service UDP segments may be lost delivered out-of-
order connectionless
no sender-receiver handshaking
each UDP segment handled independently
UDP uses streaming
multimedia apps (loss tolerant rate sensitive)
DNS SNMP
reliable transfer over UDP add reliability at
application layer application-specific
error recovery
Transport Layer 3-16
UDP segment header
source port dest port
32 bits
applicationdata
(payload)
UDP segment format
length checksum
length in bytes of UDP segment
including header
no connection establishment (which can add delay)
simple no connection state at sender receiver
small header size no congestion control
UDP can blast away as fast as desired
why is there a UDP
Transport Layer 3-17
UDP checksum
sender treat segment contents
including header fields as sequence of 16-bit integers
checksum addition (onersquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (flipped bits) in segments
Transport Layer 3-18
Internet checksum exampleexample add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Note when adding numbers a carryout from the most significant bit needs to be added to the
result
Q1 Sockets and multiplexing TCP uses more information in packet
headers in order to demultiplex packets compared to UDPA TrueB False
Transport Layer 3-19
Q2 Sockets UDP Suppose we use UDP instead of TCP under
HTTP for designing a web server where all requests and responses fit in a single packet Suppose a 100 clients are simultaneously communicating with this web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-20
Q3 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-21
Q4 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server Do all of the sockets at the server have the same server-side port numberA YesB No
Transport Layer 3-22
Q5 UDP checksums Letrsquos denote a UDP packet as (checksum
data) ignoring other fields for this question Suppose a sender sends (0010 1110) and the receiver receives (00111110) Which of the following is true of the receiverA Thinks the packet is corrupted and
discards the packetB Thinks only the checksum is corrupted
and delivers the correct data to the application
C Can possibly conclude that nothing is wrong with the packet
D A and CTransport Layer 3-23
Transport Layer 3-24
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-25
TCP Overview RFCs 79311221323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) inits sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order
byte steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
Transport Layer 3-26
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement number
receive window
Urg data pointerchecksum
FSRPAUheadlen
notused
options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 3-27
TCP seq numbers ACKssequence numbers
byte stream ldquonumberrdquo of first byte in segmentrsquos data
acknowledgementsseq of next byte expected from other side
cumulative ACKQ how receiver handles out-of-order segmentsA TCP spec doesnrsquot say - up to implementor
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent ACKed
sent not-yet ACKed(ldquoin-flightrdquo)
usablebut not yet sent
not usable
window size N
sender sequence number space
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
outgoing segment from sender
Transport Layer 3-28
TCP seq numbers ACKs
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt oflsquoCrsquo echoesback lsquoCrsquo
simple telnet scenario
Host BHost A
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43 data = lsquoCrsquo
Seq=43 ACK=80
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-9
Connectionless demultiplexing
recall created socket has host-local port
DatagramSocket mySocket1 = new DatagramSocket(12534)
when host receives UDP segment checks destination IP
and port in segment directs UDP segment
to socket bound to that (IPport)
recall when creating datagram to send into UDP socket must specify
destination IP address destination port IP datagrams with same dest (IP port) but different source IP addresses andor source port numbers will be directed to same socket
Transport Layer 3-10
Connectionless demux example
DatagramSocket serverSocket = new DatagramSocket
(6428)
transport
application
physical
link
network
P3transport
application
physical
link
network
P1
transport
application
physical
link
network
P4
DatagramSocket mySocket1 = new DatagramSocket (5775)
DatagramSocket mySocket2 = new DatagramSocket (9157)
source port 9157dest port 6428
source port 6428dest port 9157
source port dest port
source port dest port
Transport Layer 3-11
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
demux receiver uses all four values to direct segment to right socket
server host has many simultaneous TCP sockets each socket identified
by its own 4-tuple web servers have
different socket each client non-persistent HTTP
will have different socket for each request
Transport Layer 3-12
Connection-oriented demux example
transport
application
physical
link
network
P3transport
app
physical
link
P4
transport
application
physical
link
network
P2
source IPport A9157dest IP port B80
source IPport B80dest IPport A9157
host IP address
A
host IP address
C
network
P6P5P3
source IPport C5775dest IPport B80
source IPport C9157dest IPport B80
three segments all destined to IP address B dest port 80 are demultiplexed to different sockets
server IP
address B
server socket also port 80
Transport Layer 3-13
Connection-oriented demux example
transport
application
physical
link
network
P3transport
app
physical
link
transport
application
physical
link
network
P2
source IPport A9157dest IP port B80
source IPport B80dest IPport A9157
host IP address
A
host IP address
C
server IP
address B
network
P3
source IPport C5775dest IPport B80
source IPport C9157dest IPport B80
P4
server socket also port 80threaded server
Transport Layer 3-14
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-15
UDP User Datagram Protocol [RFC 768] no frills bare bones
transport protocol for ldquobest effortrdquo service UDP segments may be lost delivered out-of-
order connectionless
no sender-receiver handshaking
each UDP segment handled independently
UDP uses streaming
multimedia apps (loss tolerant rate sensitive)
DNS SNMP
reliable transfer over UDP add reliability at
application layer application-specific
error recovery
Transport Layer 3-16
UDP segment header
source port dest port
32 bits
applicationdata
(payload)
UDP segment format
length checksum
length in bytes of UDP segment
including header
no connection establishment (which can add delay)
simple no connection state at sender receiver
small header size no congestion control
UDP can blast away as fast as desired
why is there a UDP
Transport Layer 3-17
UDP checksum
sender treat segment contents
including header fields as sequence of 16-bit integers
checksum addition (onersquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (flipped bits) in segments
Transport Layer 3-18
Internet checksum exampleexample add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Note when adding numbers a carryout from the most significant bit needs to be added to the
result
Q1 Sockets and multiplexing TCP uses more information in packet
headers in order to demultiplex packets compared to UDPA TrueB False
Transport Layer 3-19
Q2 Sockets UDP Suppose we use UDP instead of TCP under
HTTP for designing a web server where all requests and responses fit in a single packet Suppose a 100 clients are simultaneously communicating with this web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-20
Q3 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-21
Q4 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server Do all of the sockets at the server have the same server-side port numberA YesB No
Transport Layer 3-22
Q5 UDP checksums Letrsquos denote a UDP packet as (checksum
data) ignoring other fields for this question Suppose a sender sends (0010 1110) and the receiver receives (00111110) Which of the following is true of the receiverA Thinks the packet is corrupted and
discards the packetB Thinks only the checksum is corrupted
and delivers the correct data to the application
C Can possibly conclude that nothing is wrong with the packet
D A and CTransport Layer 3-23
Transport Layer 3-24
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-25
TCP Overview RFCs 79311221323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) inits sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order
byte steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
Transport Layer 3-26
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement number
receive window
Urg data pointerchecksum
FSRPAUheadlen
notused
options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 3-27
TCP seq numbers ACKssequence numbers
byte stream ldquonumberrdquo of first byte in segmentrsquos data
acknowledgementsseq of next byte expected from other side
cumulative ACKQ how receiver handles out-of-order segmentsA TCP spec doesnrsquot say - up to implementor
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent ACKed
sent not-yet ACKed(ldquoin-flightrdquo)
usablebut not yet sent
not usable
window size N
sender sequence number space
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
outgoing segment from sender
Transport Layer 3-28
TCP seq numbers ACKs
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt oflsquoCrsquo echoesback lsquoCrsquo
simple telnet scenario
Host BHost A
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43 data = lsquoCrsquo
Seq=43 ACK=80
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-10
Connectionless demux example
DatagramSocket serverSocket = new DatagramSocket
(6428)
transport
application
physical
link
network
P3transport
application
physical
link
network
P1
transport
application
physical
link
network
P4
DatagramSocket mySocket1 = new DatagramSocket (5775)
DatagramSocket mySocket2 = new DatagramSocket (9157)
source port 9157dest port 6428
source port 6428dest port 9157
source port dest port
source port dest port
Transport Layer 3-11
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
demux receiver uses all four values to direct segment to right socket
server host has many simultaneous TCP sockets each socket identified
by its own 4-tuple web servers have
different socket each client non-persistent HTTP
will have different socket for each request
Transport Layer 3-12
Connection-oriented demux example
transport
application
physical
link
network
P3transport
app
physical
link
P4
transport
application
physical
link
network
P2
source IPport A9157dest IP port B80
source IPport B80dest IPport A9157
host IP address
A
host IP address
C
network
P6P5P3
source IPport C5775dest IPport B80
source IPport C9157dest IPport B80
three segments all destined to IP address B dest port 80 are demultiplexed to different sockets
server IP
address B
server socket also port 80
Transport Layer 3-13
Connection-oriented demux example
transport
application
physical
link
network
P3transport
app
physical
link
transport
application
physical
link
network
P2
source IPport A9157dest IP port B80
source IPport B80dest IPport A9157
host IP address
A
host IP address
C
server IP
address B
network
P3
source IPport C5775dest IPport B80
source IPport C9157dest IPport B80
P4
server socket also port 80threaded server
Transport Layer 3-14
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-15
UDP User Datagram Protocol [RFC 768] no frills bare bones
transport protocol for ldquobest effortrdquo service UDP segments may be lost delivered out-of-
order connectionless
no sender-receiver handshaking
each UDP segment handled independently
UDP uses streaming
multimedia apps (loss tolerant rate sensitive)
DNS SNMP
reliable transfer over UDP add reliability at
application layer application-specific
error recovery
Transport Layer 3-16
UDP segment header
source port dest port
32 bits
applicationdata
(payload)
UDP segment format
length checksum
length in bytes of UDP segment
including header
no connection establishment (which can add delay)
simple no connection state at sender receiver
small header size no congestion control
UDP can blast away as fast as desired
why is there a UDP
Transport Layer 3-17
UDP checksum
sender treat segment contents
including header fields as sequence of 16-bit integers
checksum addition (onersquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (flipped bits) in segments
Transport Layer 3-18
Internet checksum exampleexample add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Note when adding numbers a carryout from the most significant bit needs to be added to the
result
Q1 Sockets and multiplexing TCP uses more information in packet
headers in order to demultiplex packets compared to UDPA TrueB False
Transport Layer 3-19
Q2 Sockets UDP Suppose we use UDP instead of TCP under
HTTP for designing a web server where all requests and responses fit in a single packet Suppose a 100 clients are simultaneously communicating with this web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-20
Q3 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-21
Q4 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server Do all of the sockets at the server have the same server-side port numberA YesB No
Transport Layer 3-22
Q5 UDP checksums Letrsquos denote a UDP packet as (checksum
data) ignoring other fields for this question Suppose a sender sends (0010 1110) and the receiver receives (00111110) Which of the following is true of the receiverA Thinks the packet is corrupted and
discards the packetB Thinks only the checksum is corrupted
and delivers the correct data to the application
C Can possibly conclude that nothing is wrong with the packet
D A and CTransport Layer 3-23
Transport Layer 3-24
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-25
TCP Overview RFCs 79311221323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) inits sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order
byte steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
Transport Layer 3-26
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement number
receive window
Urg data pointerchecksum
FSRPAUheadlen
notused
options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 3-27
TCP seq numbers ACKssequence numbers
byte stream ldquonumberrdquo of first byte in segmentrsquos data
acknowledgementsseq of next byte expected from other side
cumulative ACKQ how receiver handles out-of-order segmentsA TCP spec doesnrsquot say - up to implementor
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent ACKed
sent not-yet ACKed(ldquoin-flightrdquo)
usablebut not yet sent
not usable
window size N
sender sequence number space
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
outgoing segment from sender
Transport Layer 3-28
TCP seq numbers ACKs
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt oflsquoCrsquo echoesback lsquoCrsquo
simple telnet scenario
Host BHost A
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43 data = lsquoCrsquo
Seq=43 ACK=80
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-11
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
demux receiver uses all four values to direct segment to right socket
server host has many simultaneous TCP sockets each socket identified
by its own 4-tuple web servers have
different socket each client non-persistent HTTP
will have different socket for each request
Transport Layer 3-12
Connection-oriented demux example
transport
application
physical
link
network
P3transport
app
physical
link
P4
transport
application
physical
link
network
P2
source IPport A9157dest IP port B80
source IPport B80dest IPport A9157
host IP address
A
host IP address
C
network
P6P5P3
source IPport C5775dest IPport B80
source IPport C9157dest IPport B80
three segments all destined to IP address B dest port 80 are demultiplexed to different sockets
server IP
address B
server socket also port 80
Transport Layer 3-13
Connection-oriented demux example
transport
application
physical
link
network
P3transport
app
physical
link
transport
application
physical
link
network
P2
source IPport A9157dest IP port B80
source IPport B80dest IPport A9157
host IP address
A
host IP address
C
server IP
address B
network
P3
source IPport C5775dest IPport B80
source IPport C9157dest IPport B80
P4
server socket also port 80threaded server
Transport Layer 3-14
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-15
UDP User Datagram Protocol [RFC 768] no frills bare bones
transport protocol for ldquobest effortrdquo service UDP segments may be lost delivered out-of-
order connectionless
no sender-receiver handshaking
each UDP segment handled independently
UDP uses streaming
multimedia apps (loss tolerant rate sensitive)
DNS SNMP
reliable transfer over UDP add reliability at
application layer application-specific
error recovery
Transport Layer 3-16
UDP segment header
source port dest port
32 bits
applicationdata
(payload)
UDP segment format
length checksum
length in bytes of UDP segment
including header
no connection establishment (which can add delay)
simple no connection state at sender receiver
small header size no congestion control
UDP can blast away as fast as desired
why is there a UDP
Transport Layer 3-17
UDP checksum
sender treat segment contents
including header fields as sequence of 16-bit integers
checksum addition (onersquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (flipped bits) in segments
Transport Layer 3-18
Internet checksum exampleexample add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Note when adding numbers a carryout from the most significant bit needs to be added to the
result
Q1 Sockets and multiplexing TCP uses more information in packet
headers in order to demultiplex packets compared to UDPA TrueB False
Transport Layer 3-19
Q2 Sockets UDP Suppose we use UDP instead of TCP under
HTTP for designing a web server where all requests and responses fit in a single packet Suppose a 100 clients are simultaneously communicating with this web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-20
Q3 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-21
Q4 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server Do all of the sockets at the server have the same server-side port numberA YesB No
Transport Layer 3-22
Q5 UDP checksums Letrsquos denote a UDP packet as (checksum
data) ignoring other fields for this question Suppose a sender sends (0010 1110) and the receiver receives (00111110) Which of the following is true of the receiverA Thinks the packet is corrupted and
discards the packetB Thinks only the checksum is corrupted
and delivers the correct data to the application
C Can possibly conclude that nothing is wrong with the packet
D A and CTransport Layer 3-23
Transport Layer 3-24
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-25
TCP Overview RFCs 79311221323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) inits sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order
byte steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
Transport Layer 3-26
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement number
receive window
Urg data pointerchecksum
FSRPAUheadlen
notused
options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 3-27
TCP seq numbers ACKssequence numbers
byte stream ldquonumberrdquo of first byte in segmentrsquos data
acknowledgementsseq of next byte expected from other side
cumulative ACKQ how receiver handles out-of-order segmentsA TCP spec doesnrsquot say - up to implementor
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent ACKed
sent not-yet ACKed(ldquoin-flightrdquo)
usablebut not yet sent
not usable
window size N
sender sequence number space
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
outgoing segment from sender
Transport Layer 3-28
TCP seq numbers ACKs
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt oflsquoCrsquo echoesback lsquoCrsquo
simple telnet scenario
Host BHost A
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43 data = lsquoCrsquo
Seq=43 ACK=80
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-12
Connection-oriented demux example
transport
application
physical
link
network
P3transport
app
physical
link
P4
transport
application
physical
link
network
P2
source IPport A9157dest IP port B80
source IPport B80dest IPport A9157
host IP address
A
host IP address
C
network
P6P5P3
source IPport C5775dest IPport B80
source IPport C9157dest IPport B80
three segments all destined to IP address B dest port 80 are demultiplexed to different sockets
server IP
address B
server socket also port 80
Transport Layer 3-13
Connection-oriented demux example
transport
application
physical
link
network
P3transport
app
physical
link
transport
application
physical
link
network
P2
source IPport A9157dest IP port B80
source IPport B80dest IPport A9157
host IP address
A
host IP address
C
server IP
address B
network
P3
source IPport C5775dest IPport B80
source IPport C9157dest IPport B80
P4
server socket also port 80threaded server
Transport Layer 3-14
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-15
UDP User Datagram Protocol [RFC 768] no frills bare bones
transport protocol for ldquobest effortrdquo service UDP segments may be lost delivered out-of-
order connectionless
no sender-receiver handshaking
each UDP segment handled independently
UDP uses streaming
multimedia apps (loss tolerant rate sensitive)
DNS SNMP
reliable transfer over UDP add reliability at
application layer application-specific
error recovery
Transport Layer 3-16
UDP segment header
source port dest port
32 bits
applicationdata
(payload)
UDP segment format
length checksum
length in bytes of UDP segment
including header
no connection establishment (which can add delay)
simple no connection state at sender receiver
small header size no congestion control
UDP can blast away as fast as desired
why is there a UDP
Transport Layer 3-17
UDP checksum
sender treat segment contents
including header fields as sequence of 16-bit integers
checksum addition (onersquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (flipped bits) in segments
Transport Layer 3-18
Internet checksum exampleexample add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Note when adding numbers a carryout from the most significant bit needs to be added to the
result
Q1 Sockets and multiplexing TCP uses more information in packet
headers in order to demultiplex packets compared to UDPA TrueB False
Transport Layer 3-19
Q2 Sockets UDP Suppose we use UDP instead of TCP under
HTTP for designing a web server where all requests and responses fit in a single packet Suppose a 100 clients are simultaneously communicating with this web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-20
Q3 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-21
Q4 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server Do all of the sockets at the server have the same server-side port numberA YesB No
Transport Layer 3-22
Q5 UDP checksums Letrsquos denote a UDP packet as (checksum
data) ignoring other fields for this question Suppose a sender sends (0010 1110) and the receiver receives (00111110) Which of the following is true of the receiverA Thinks the packet is corrupted and
discards the packetB Thinks only the checksum is corrupted
and delivers the correct data to the application
C Can possibly conclude that nothing is wrong with the packet
D A and CTransport Layer 3-23
Transport Layer 3-24
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-25
TCP Overview RFCs 79311221323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) inits sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order
byte steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
Transport Layer 3-26
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement number
receive window
Urg data pointerchecksum
FSRPAUheadlen
notused
options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 3-27
TCP seq numbers ACKssequence numbers
byte stream ldquonumberrdquo of first byte in segmentrsquos data
acknowledgementsseq of next byte expected from other side
cumulative ACKQ how receiver handles out-of-order segmentsA TCP spec doesnrsquot say - up to implementor
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent ACKed
sent not-yet ACKed(ldquoin-flightrdquo)
usablebut not yet sent
not usable
window size N
sender sequence number space
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
outgoing segment from sender
Transport Layer 3-28
TCP seq numbers ACKs
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt oflsquoCrsquo echoesback lsquoCrsquo
simple telnet scenario
Host BHost A
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43 data = lsquoCrsquo
Seq=43 ACK=80
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-13
Connection-oriented demux example
transport
application
physical
link
network
P3transport
app
physical
link
transport
application
physical
link
network
P2
source IPport A9157dest IP port B80
source IPport B80dest IPport A9157
host IP address
A
host IP address
C
server IP
address B
network
P3
source IPport C5775dest IPport B80
source IPport C9157dest IPport B80
P4
server socket also port 80threaded server
Transport Layer 3-14
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-15
UDP User Datagram Protocol [RFC 768] no frills bare bones
transport protocol for ldquobest effortrdquo service UDP segments may be lost delivered out-of-
order connectionless
no sender-receiver handshaking
each UDP segment handled independently
UDP uses streaming
multimedia apps (loss tolerant rate sensitive)
DNS SNMP
reliable transfer over UDP add reliability at
application layer application-specific
error recovery
Transport Layer 3-16
UDP segment header
source port dest port
32 bits
applicationdata
(payload)
UDP segment format
length checksum
length in bytes of UDP segment
including header
no connection establishment (which can add delay)
simple no connection state at sender receiver
small header size no congestion control
UDP can blast away as fast as desired
why is there a UDP
Transport Layer 3-17
UDP checksum
sender treat segment contents
including header fields as sequence of 16-bit integers
checksum addition (onersquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (flipped bits) in segments
Transport Layer 3-18
Internet checksum exampleexample add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Note when adding numbers a carryout from the most significant bit needs to be added to the
result
Q1 Sockets and multiplexing TCP uses more information in packet
headers in order to demultiplex packets compared to UDPA TrueB False
Transport Layer 3-19
Q2 Sockets UDP Suppose we use UDP instead of TCP under
HTTP for designing a web server where all requests and responses fit in a single packet Suppose a 100 clients are simultaneously communicating with this web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-20
Q3 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-21
Q4 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server Do all of the sockets at the server have the same server-side port numberA YesB No
Transport Layer 3-22
Q5 UDP checksums Letrsquos denote a UDP packet as (checksum
data) ignoring other fields for this question Suppose a sender sends (0010 1110) and the receiver receives (00111110) Which of the following is true of the receiverA Thinks the packet is corrupted and
discards the packetB Thinks only the checksum is corrupted
and delivers the correct data to the application
C Can possibly conclude that nothing is wrong with the packet
D A and CTransport Layer 3-23
Transport Layer 3-24
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-25
TCP Overview RFCs 79311221323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) inits sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order
byte steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
Transport Layer 3-26
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement number
receive window
Urg data pointerchecksum
FSRPAUheadlen
notused
options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 3-27
TCP seq numbers ACKssequence numbers
byte stream ldquonumberrdquo of first byte in segmentrsquos data
acknowledgementsseq of next byte expected from other side
cumulative ACKQ how receiver handles out-of-order segmentsA TCP spec doesnrsquot say - up to implementor
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent ACKed
sent not-yet ACKed(ldquoin-flightrdquo)
usablebut not yet sent
not usable
window size N
sender sequence number space
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
outgoing segment from sender
Transport Layer 3-28
TCP seq numbers ACKs
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt oflsquoCrsquo echoesback lsquoCrsquo
simple telnet scenario
Host BHost A
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43 data = lsquoCrsquo
Seq=43 ACK=80
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-14
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-15
UDP User Datagram Protocol [RFC 768] no frills bare bones
transport protocol for ldquobest effortrdquo service UDP segments may be lost delivered out-of-
order connectionless
no sender-receiver handshaking
each UDP segment handled independently
UDP uses streaming
multimedia apps (loss tolerant rate sensitive)
DNS SNMP
reliable transfer over UDP add reliability at
application layer application-specific
error recovery
Transport Layer 3-16
UDP segment header
source port dest port
32 bits
applicationdata
(payload)
UDP segment format
length checksum
length in bytes of UDP segment
including header
no connection establishment (which can add delay)
simple no connection state at sender receiver
small header size no congestion control
UDP can blast away as fast as desired
why is there a UDP
Transport Layer 3-17
UDP checksum
sender treat segment contents
including header fields as sequence of 16-bit integers
checksum addition (onersquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (flipped bits) in segments
Transport Layer 3-18
Internet checksum exampleexample add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Note when adding numbers a carryout from the most significant bit needs to be added to the
result
Q1 Sockets and multiplexing TCP uses more information in packet
headers in order to demultiplex packets compared to UDPA TrueB False
Transport Layer 3-19
Q2 Sockets UDP Suppose we use UDP instead of TCP under
HTTP for designing a web server where all requests and responses fit in a single packet Suppose a 100 clients are simultaneously communicating with this web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-20
Q3 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-21
Q4 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server Do all of the sockets at the server have the same server-side port numberA YesB No
Transport Layer 3-22
Q5 UDP checksums Letrsquos denote a UDP packet as (checksum
data) ignoring other fields for this question Suppose a sender sends (0010 1110) and the receiver receives (00111110) Which of the following is true of the receiverA Thinks the packet is corrupted and
discards the packetB Thinks only the checksum is corrupted
and delivers the correct data to the application
C Can possibly conclude that nothing is wrong with the packet
D A and CTransport Layer 3-23
Transport Layer 3-24
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-25
TCP Overview RFCs 79311221323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) inits sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order
byte steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
Transport Layer 3-26
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement number
receive window
Urg data pointerchecksum
FSRPAUheadlen
notused
options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 3-27
TCP seq numbers ACKssequence numbers
byte stream ldquonumberrdquo of first byte in segmentrsquos data
acknowledgementsseq of next byte expected from other side
cumulative ACKQ how receiver handles out-of-order segmentsA TCP spec doesnrsquot say - up to implementor
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent ACKed
sent not-yet ACKed(ldquoin-flightrdquo)
usablebut not yet sent
not usable
window size N
sender sequence number space
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
outgoing segment from sender
Transport Layer 3-28
TCP seq numbers ACKs
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt oflsquoCrsquo echoesback lsquoCrsquo
simple telnet scenario
Host BHost A
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43 data = lsquoCrsquo
Seq=43 ACK=80
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-15
UDP User Datagram Protocol [RFC 768] no frills bare bones
transport protocol for ldquobest effortrdquo service UDP segments may be lost delivered out-of-
order connectionless
no sender-receiver handshaking
each UDP segment handled independently
UDP uses streaming
multimedia apps (loss tolerant rate sensitive)
DNS SNMP
reliable transfer over UDP add reliability at
application layer application-specific
error recovery
Transport Layer 3-16
UDP segment header
source port dest port
32 bits
applicationdata
(payload)
UDP segment format
length checksum
length in bytes of UDP segment
including header
no connection establishment (which can add delay)
simple no connection state at sender receiver
small header size no congestion control
UDP can blast away as fast as desired
why is there a UDP
Transport Layer 3-17
UDP checksum
sender treat segment contents
including header fields as sequence of 16-bit integers
checksum addition (onersquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (flipped bits) in segments
Transport Layer 3-18
Internet checksum exampleexample add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Note when adding numbers a carryout from the most significant bit needs to be added to the
result
Q1 Sockets and multiplexing TCP uses more information in packet
headers in order to demultiplex packets compared to UDPA TrueB False
Transport Layer 3-19
Q2 Sockets UDP Suppose we use UDP instead of TCP under
HTTP for designing a web server where all requests and responses fit in a single packet Suppose a 100 clients are simultaneously communicating with this web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-20
Q3 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-21
Q4 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server Do all of the sockets at the server have the same server-side port numberA YesB No
Transport Layer 3-22
Q5 UDP checksums Letrsquos denote a UDP packet as (checksum
data) ignoring other fields for this question Suppose a sender sends (0010 1110) and the receiver receives (00111110) Which of the following is true of the receiverA Thinks the packet is corrupted and
discards the packetB Thinks only the checksum is corrupted
and delivers the correct data to the application
C Can possibly conclude that nothing is wrong with the packet
D A and CTransport Layer 3-23
Transport Layer 3-24
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-25
TCP Overview RFCs 79311221323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) inits sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order
byte steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
Transport Layer 3-26
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement number
receive window
Urg data pointerchecksum
FSRPAUheadlen
notused
options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 3-27
TCP seq numbers ACKssequence numbers
byte stream ldquonumberrdquo of first byte in segmentrsquos data
acknowledgementsseq of next byte expected from other side
cumulative ACKQ how receiver handles out-of-order segmentsA TCP spec doesnrsquot say - up to implementor
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent ACKed
sent not-yet ACKed(ldquoin-flightrdquo)
usablebut not yet sent
not usable
window size N
sender sequence number space
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
outgoing segment from sender
Transport Layer 3-28
TCP seq numbers ACKs
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt oflsquoCrsquo echoesback lsquoCrsquo
simple telnet scenario
Host BHost A
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43 data = lsquoCrsquo
Seq=43 ACK=80
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-16
UDP segment header
source port dest port
32 bits
applicationdata
(payload)
UDP segment format
length checksum
length in bytes of UDP segment
including header
no connection establishment (which can add delay)
simple no connection state at sender receiver
small header size no congestion control
UDP can blast away as fast as desired
why is there a UDP
Transport Layer 3-17
UDP checksum
sender treat segment contents
including header fields as sequence of 16-bit integers
checksum addition (onersquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (flipped bits) in segments
Transport Layer 3-18
Internet checksum exampleexample add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Note when adding numbers a carryout from the most significant bit needs to be added to the
result
Q1 Sockets and multiplexing TCP uses more information in packet
headers in order to demultiplex packets compared to UDPA TrueB False
Transport Layer 3-19
Q2 Sockets UDP Suppose we use UDP instead of TCP under
HTTP for designing a web server where all requests and responses fit in a single packet Suppose a 100 clients are simultaneously communicating with this web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-20
Q3 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-21
Q4 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server Do all of the sockets at the server have the same server-side port numberA YesB No
Transport Layer 3-22
Q5 UDP checksums Letrsquos denote a UDP packet as (checksum
data) ignoring other fields for this question Suppose a sender sends (0010 1110) and the receiver receives (00111110) Which of the following is true of the receiverA Thinks the packet is corrupted and
discards the packetB Thinks only the checksum is corrupted
and delivers the correct data to the application
C Can possibly conclude that nothing is wrong with the packet
D A and CTransport Layer 3-23
Transport Layer 3-24
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-25
TCP Overview RFCs 79311221323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) inits sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order
byte steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
Transport Layer 3-26
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement number
receive window
Urg data pointerchecksum
FSRPAUheadlen
notused
options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 3-27
TCP seq numbers ACKssequence numbers
byte stream ldquonumberrdquo of first byte in segmentrsquos data
acknowledgementsseq of next byte expected from other side
cumulative ACKQ how receiver handles out-of-order segmentsA TCP spec doesnrsquot say - up to implementor
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent ACKed
sent not-yet ACKed(ldquoin-flightrdquo)
usablebut not yet sent
not usable
window size N
sender sequence number space
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
outgoing segment from sender
Transport Layer 3-28
TCP seq numbers ACKs
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt oflsquoCrsquo echoesback lsquoCrsquo
simple telnet scenario
Host BHost A
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43 data = lsquoCrsquo
Seq=43 ACK=80
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-17
UDP checksum
sender treat segment contents
including header fields as sequence of 16-bit integers
checksum addition (onersquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (flipped bits) in segments
Transport Layer 3-18
Internet checksum exampleexample add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Note when adding numbers a carryout from the most significant bit needs to be added to the
result
Q1 Sockets and multiplexing TCP uses more information in packet
headers in order to demultiplex packets compared to UDPA TrueB False
Transport Layer 3-19
Q2 Sockets UDP Suppose we use UDP instead of TCP under
HTTP for designing a web server where all requests and responses fit in a single packet Suppose a 100 clients are simultaneously communicating with this web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-20
Q3 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-21
Q4 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server Do all of the sockets at the server have the same server-side port numberA YesB No
Transport Layer 3-22
Q5 UDP checksums Letrsquos denote a UDP packet as (checksum
data) ignoring other fields for this question Suppose a sender sends (0010 1110) and the receiver receives (00111110) Which of the following is true of the receiverA Thinks the packet is corrupted and
discards the packetB Thinks only the checksum is corrupted
and delivers the correct data to the application
C Can possibly conclude that nothing is wrong with the packet
D A and CTransport Layer 3-23
Transport Layer 3-24
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-25
TCP Overview RFCs 79311221323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) inits sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order
byte steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
Transport Layer 3-26
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement number
receive window
Urg data pointerchecksum
FSRPAUheadlen
notused
options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 3-27
TCP seq numbers ACKssequence numbers
byte stream ldquonumberrdquo of first byte in segmentrsquos data
acknowledgementsseq of next byte expected from other side
cumulative ACKQ how receiver handles out-of-order segmentsA TCP spec doesnrsquot say - up to implementor
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent ACKed
sent not-yet ACKed(ldquoin-flightrdquo)
usablebut not yet sent
not usable
window size N
sender sequence number space
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
outgoing segment from sender
Transport Layer 3-28
TCP seq numbers ACKs
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt oflsquoCrsquo echoesback lsquoCrsquo
simple telnet scenario
Host BHost A
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43 data = lsquoCrsquo
Seq=43 ACK=80
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-18
Internet checksum exampleexample add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Note when adding numbers a carryout from the most significant bit needs to be added to the
result
Q1 Sockets and multiplexing TCP uses more information in packet
headers in order to demultiplex packets compared to UDPA TrueB False
Transport Layer 3-19
Q2 Sockets UDP Suppose we use UDP instead of TCP under
HTTP for designing a web server where all requests and responses fit in a single packet Suppose a 100 clients are simultaneously communicating with this web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-20
Q3 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-21
Q4 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server Do all of the sockets at the server have the same server-side port numberA YesB No
Transport Layer 3-22
Q5 UDP checksums Letrsquos denote a UDP packet as (checksum
data) ignoring other fields for this question Suppose a sender sends (0010 1110) and the receiver receives (00111110) Which of the following is true of the receiverA Thinks the packet is corrupted and
discards the packetB Thinks only the checksum is corrupted
and delivers the correct data to the application
C Can possibly conclude that nothing is wrong with the packet
D A and CTransport Layer 3-23
Transport Layer 3-24
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-25
TCP Overview RFCs 79311221323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) inits sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order
byte steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
Transport Layer 3-26
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement number
receive window
Urg data pointerchecksum
FSRPAUheadlen
notused
options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 3-27
TCP seq numbers ACKssequence numbers
byte stream ldquonumberrdquo of first byte in segmentrsquos data
acknowledgementsseq of next byte expected from other side
cumulative ACKQ how receiver handles out-of-order segmentsA TCP spec doesnrsquot say - up to implementor
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent ACKed
sent not-yet ACKed(ldquoin-flightrdquo)
usablebut not yet sent
not usable
window size N
sender sequence number space
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
outgoing segment from sender
Transport Layer 3-28
TCP seq numbers ACKs
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt oflsquoCrsquo echoesback lsquoCrsquo
simple telnet scenario
Host BHost A
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43 data = lsquoCrsquo
Seq=43 ACK=80
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Q1 Sockets and multiplexing TCP uses more information in packet
headers in order to demultiplex packets compared to UDPA TrueB False
Transport Layer 3-19
Q2 Sockets UDP Suppose we use UDP instead of TCP under
HTTP for designing a web server where all requests and responses fit in a single packet Suppose a 100 clients are simultaneously communicating with this web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-20
Q3 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-21
Q4 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server Do all of the sockets at the server have the same server-side port numberA YesB No
Transport Layer 3-22
Q5 UDP checksums Letrsquos denote a UDP packet as (checksum
data) ignoring other fields for this question Suppose a sender sends (0010 1110) and the receiver receives (00111110) Which of the following is true of the receiverA Thinks the packet is corrupted and
discards the packetB Thinks only the checksum is corrupted
and delivers the correct data to the application
C Can possibly conclude that nothing is wrong with the packet
D A and CTransport Layer 3-23
Transport Layer 3-24
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-25
TCP Overview RFCs 79311221323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) inits sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order
byte steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
Transport Layer 3-26
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement number
receive window
Urg data pointerchecksum
FSRPAUheadlen
notused
options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 3-27
TCP seq numbers ACKssequence numbers
byte stream ldquonumberrdquo of first byte in segmentrsquos data
acknowledgementsseq of next byte expected from other side
cumulative ACKQ how receiver handles out-of-order segmentsA TCP spec doesnrsquot say - up to implementor
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent ACKed
sent not-yet ACKed(ldquoin-flightrdquo)
usablebut not yet sent
not usable
window size N
sender sequence number space
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
outgoing segment from sender
Transport Layer 3-28
TCP seq numbers ACKs
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt oflsquoCrsquo echoesback lsquoCrsquo
simple telnet scenario
Host BHost A
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43 data = lsquoCrsquo
Seq=43 ACK=80
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Q2 Sockets UDP Suppose we use UDP instead of TCP under
HTTP for designing a web server where all requests and responses fit in a single packet Suppose a 100 clients are simultaneously communicating with this web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-20
Q3 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-21
Q4 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server Do all of the sockets at the server have the same server-side port numberA YesB No
Transport Layer 3-22
Q5 UDP checksums Letrsquos denote a UDP packet as (checksum
data) ignoring other fields for this question Suppose a sender sends (0010 1110) and the receiver receives (00111110) Which of the following is true of the receiverA Thinks the packet is corrupted and
discards the packetB Thinks only the checksum is corrupted
and delivers the correct data to the application
C Can possibly conclude that nothing is wrong with the packet
D A and CTransport Layer 3-23
Transport Layer 3-24
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-25
TCP Overview RFCs 79311221323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) inits sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order
byte steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
Transport Layer 3-26
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement number
receive window
Urg data pointerchecksum
FSRPAUheadlen
notused
options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 3-27
TCP seq numbers ACKssequence numbers
byte stream ldquonumberrdquo of first byte in segmentrsquos data
acknowledgementsseq of next byte expected from other side
cumulative ACKQ how receiver handles out-of-order segmentsA TCP spec doesnrsquot say - up to implementor
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent ACKed
sent not-yet ACKed(ldquoin-flightrdquo)
usablebut not yet sent
not usable
window size N
sender sequence number space
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
outgoing segment from sender
Transport Layer 3-28
TCP seq numbers ACKs
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt oflsquoCrsquo echoesback lsquoCrsquo
simple telnet scenario
Host BHost A
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43 data = lsquoCrsquo
Seq=43 ACK=80
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Q3 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server How many sockets are respectively at the server and at each clientA 11B 21C 2002D 1001E 101 1
Transport Layer 3-21
Q4 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server Do all of the sockets at the server have the same server-side port numberA YesB No
Transport Layer 3-22
Q5 UDP checksums Letrsquos denote a UDP packet as (checksum
data) ignoring other fields for this question Suppose a sender sends (0010 1110) and the receiver receives (00111110) Which of the following is true of the receiverA Thinks the packet is corrupted and
discards the packetB Thinks only the checksum is corrupted
and delivers the correct data to the application
C Can possibly conclude that nothing is wrong with the packet
D A and CTransport Layer 3-23
Transport Layer 3-24
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-25
TCP Overview RFCs 79311221323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) inits sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order
byte steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
Transport Layer 3-26
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement number
receive window
Urg data pointerchecksum
FSRPAUheadlen
notused
options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 3-27
TCP seq numbers ACKssequence numbers
byte stream ldquonumberrdquo of first byte in segmentrsquos data
acknowledgementsseq of next byte expected from other side
cumulative ACKQ how receiver handles out-of-order segmentsA TCP spec doesnrsquot say - up to implementor
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent ACKed
sent not-yet ACKed(ldquoin-flightrdquo)
usablebut not yet sent
not usable
window size N
sender sequence number space
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
outgoing segment from sender
Transport Layer 3-28
TCP seq numbers ACKs
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt oflsquoCrsquo echoesback lsquoCrsquo
simple telnet scenario
Host BHost A
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43 data = lsquoCrsquo
Seq=43 ACK=80
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Q4 Sockets TCP
Suppose a 100 clients are simultaneously communicating with (a traditional HTTPTCP) web server Do all of the sockets at the server have the same server-side port numberA YesB No
Transport Layer 3-22
Q5 UDP checksums Letrsquos denote a UDP packet as (checksum
data) ignoring other fields for this question Suppose a sender sends (0010 1110) and the receiver receives (00111110) Which of the following is true of the receiverA Thinks the packet is corrupted and
discards the packetB Thinks only the checksum is corrupted
and delivers the correct data to the application
C Can possibly conclude that nothing is wrong with the packet
D A and CTransport Layer 3-23
Transport Layer 3-24
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-25
TCP Overview RFCs 79311221323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) inits sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order
byte steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
Transport Layer 3-26
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement number
receive window
Urg data pointerchecksum
FSRPAUheadlen
notused
options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 3-27
TCP seq numbers ACKssequence numbers
byte stream ldquonumberrdquo of first byte in segmentrsquos data
acknowledgementsseq of next byte expected from other side
cumulative ACKQ how receiver handles out-of-order segmentsA TCP spec doesnrsquot say - up to implementor
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent ACKed
sent not-yet ACKed(ldquoin-flightrdquo)
usablebut not yet sent
not usable
window size N
sender sequence number space
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
outgoing segment from sender
Transport Layer 3-28
TCP seq numbers ACKs
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt oflsquoCrsquo echoesback lsquoCrsquo
simple telnet scenario
Host BHost A
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43 data = lsquoCrsquo
Seq=43 ACK=80
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Q5 UDP checksums Letrsquos denote a UDP packet as (checksum
data) ignoring other fields for this question Suppose a sender sends (0010 1110) and the receiver receives (00111110) Which of the following is true of the receiverA Thinks the packet is corrupted and
discards the packetB Thinks only the checksum is corrupted
and delivers the correct data to the application
C Can possibly conclude that nothing is wrong with the packet
D A and CTransport Layer 3-23
Transport Layer 3-24
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-25
TCP Overview RFCs 79311221323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) inits sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order
byte steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
Transport Layer 3-26
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement number
receive window
Urg data pointerchecksum
FSRPAUheadlen
notused
options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 3-27
TCP seq numbers ACKssequence numbers
byte stream ldquonumberrdquo of first byte in segmentrsquos data
acknowledgementsseq of next byte expected from other side
cumulative ACKQ how receiver handles out-of-order segmentsA TCP spec doesnrsquot say - up to implementor
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent ACKed
sent not-yet ACKed(ldquoin-flightrdquo)
usablebut not yet sent
not usable
window size N
sender sequence number space
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
outgoing segment from sender
Transport Layer 3-28
TCP seq numbers ACKs
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt oflsquoCrsquo echoesback lsquoCrsquo
simple telnet scenario
Host BHost A
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43 data = lsquoCrsquo
Seq=43 ACK=80
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-24
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-25
TCP Overview RFCs 79311221323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) inits sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order
byte steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
Transport Layer 3-26
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement number
receive window
Urg data pointerchecksum
FSRPAUheadlen
notused
options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 3-27
TCP seq numbers ACKssequence numbers
byte stream ldquonumberrdquo of first byte in segmentrsquos data
acknowledgementsseq of next byte expected from other side
cumulative ACKQ how receiver handles out-of-order segmentsA TCP spec doesnrsquot say - up to implementor
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent ACKed
sent not-yet ACKed(ldquoin-flightrdquo)
usablebut not yet sent
not usable
window size N
sender sequence number space
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
outgoing segment from sender
Transport Layer 3-28
TCP seq numbers ACKs
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt oflsquoCrsquo echoesback lsquoCrsquo
simple telnet scenario
Host BHost A
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43 data = lsquoCrsquo
Seq=43 ACK=80
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-25
TCP Overview RFCs 79311221323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) inits sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order
byte steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
Transport Layer 3-26
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement number
receive window
Urg data pointerchecksum
FSRPAUheadlen
notused
options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 3-27
TCP seq numbers ACKssequence numbers
byte stream ldquonumberrdquo of first byte in segmentrsquos data
acknowledgementsseq of next byte expected from other side
cumulative ACKQ how receiver handles out-of-order segmentsA TCP spec doesnrsquot say - up to implementor
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent ACKed
sent not-yet ACKed(ldquoin-flightrdquo)
usablebut not yet sent
not usable
window size N
sender sequence number space
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
outgoing segment from sender
Transport Layer 3-28
TCP seq numbers ACKs
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt oflsquoCrsquo echoesback lsquoCrsquo
simple telnet scenario
Host BHost A
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43 data = lsquoCrsquo
Seq=43 ACK=80
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-26
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement number
receive window
Urg data pointerchecksum
FSRPAUheadlen
notused
options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 3-27
TCP seq numbers ACKssequence numbers
byte stream ldquonumberrdquo of first byte in segmentrsquos data
acknowledgementsseq of next byte expected from other side
cumulative ACKQ how receiver handles out-of-order segmentsA TCP spec doesnrsquot say - up to implementor
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent ACKed
sent not-yet ACKed(ldquoin-flightrdquo)
usablebut not yet sent
not usable
window size N
sender sequence number space
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
outgoing segment from sender
Transport Layer 3-28
TCP seq numbers ACKs
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt oflsquoCrsquo echoesback lsquoCrsquo
simple telnet scenario
Host BHost A
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43 data = lsquoCrsquo
Seq=43 ACK=80
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-27
TCP seq numbers ACKssequence numbers
byte stream ldquonumberrdquo of first byte in segmentrsquos data
acknowledgementsseq of next byte expected from other side
cumulative ACKQ how receiver handles out-of-order segmentsA TCP spec doesnrsquot say - up to implementor
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
incoming segment to sender
A
sent ACKed
sent not-yet ACKed(ldquoin-flightrdquo)
usablebut not yet sent
not usable
window size N
sender sequence number space
source port dest port
sequence number
acknowledgement number
checksum
rwnd
urg pointer
outgoing segment from sender
Transport Layer 3-28
TCP seq numbers ACKs
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt oflsquoCrsquo echoesback lsquoCrsquo
simple telnet scenario
Host BHost A
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43 data = lsquoCrsquo
Seq=43 ACK=80
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-28
TCP seq numbers ACKs
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt oflsquoCrsquo echoesback lsquoCrsquo
simple telnet scenario
Host BHost A
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43 data = lsquoCrsquo
Seq=43 ACK=80
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-29
TCP round trip time timeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short
premature timeout unnecessary retransmissions
too long slow reaction to segment loss
Q how to estimate RTT
SampleRTT measured time from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several
recent measurements not just current SampleRTT
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-30
RTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
EstimatedRTT = (1- )EstimatedRTT + SampleRTT exponential weighted moving average influence of past sample decreases
exponentially fast typical value = 0125
TCP round trip time timeout
RTT
(mill
iseco
nds)
RTT gaiacsumassedu to fantasiaeurecomfr
sampleRTT
EstimatedRTT
time (seconds)
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-31
timeout interval EstimatedRTT plus ldquosafety marginrdquo large variation in EstimatedRTT -gt larger safety margin
estimate SampleRTT deviation from EstimatedRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
TCP round trip time timeout
(typically = 025)
TimeoutInterval = EstimatedRTT + 4DevRTT
estimated RTT ldquosafety marginrdquo
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-32
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-33
TCP reliable data transfer TCP creates rdt
service on top of IPrsquos unreliable service pipelined segments cumulative acks
bull selective acks often supported as an option
single retransmission timer
retransmissions triggered by timeout events duplicate acks
letrsquos initially consider simplified TCP sender ignore duplicate acks ignore flow control
congestion control
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-34
TCP sender eventsdata rcvd from app create segment
with seq (= byte-stream number of first data byte in segment)
start timer if not already running (for oldest unacked segment) TimeOutInterval =
smoothed_RTT + 4deviation_RTT
timeout retransmit segment
that caused timeout restart timer ack rcvd if ack acknowledges
previously unacked segments update what is
known to be ACKed (re-)start timer if still
unacked segments
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-35
TCP sender (simplified)
waitfor
event
NextSeqNum = InitialSeqNumSendBase = InitialSeqNum
create segment seq NextSeqNumpass segment to IP (ie ldquosendrdquo)NextSeqNum = NextSeqNum + length(data) if (timer currently not running) start timer
data received from application above
retransmit not-yet-acked segment with smallest seq start timer
timeout
if (y gt SendBase) SendBase = y SendBasendash1 last cumulatively ACKed byte if (there are currently not-yet-acked segments) (re-)start timer else stop timer
ACK received with ACK field value y
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-36
TCP retransmission scenarios
lost ACK scenario
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8 bytes of data
Xtim
eo
ut
ACK=100
premature timeout
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=92 8bytes of data
tim
eo
ut
ACK=120
Seq=100 20 bytes of data
ACK=120
SendBase=100
SendBase=120
SendBase=120
SendBase=92
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-37
TCP retransmission scenarios
X
cumulative ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
Seq=120 15 bytes of data
tim
eo
ut
Seq=100 20 bytes of data
ACK=120
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-38
TCP ACK generation [RFC 1122 RFC
2581]
event at receiver
arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
arrival of in-order segment withexpected seq One other segment has ACK pending
arrival of out-of-order segmenthigher-than-expect seq Gap detected
arrival of segment that partially or completely fills gap
TCP receiver action
delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
immediately send single cumulative ACK ACKing both in-order segments
immediately send duplicate ACK indicating seq of next expected byte
immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-39
TCP fast retransmit time-out period
often relatively long long delay before
resending lost packet detect lost
segments via duplicate ACKs sender often sends
many segments back-to-back
if segment is lost there will likely be many duplicate ACKs
if sender receives 3 ACKs for same data(ldquotriple duplicate ACKsrdquo) resend unacked segment with smallest seq
likely that unacked segment lost so donrsquot wait for timeout
TCP fast retransmit
(ldquotriple duplicate ACKsrdquo)
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-40
X
fast retransmit after sender receipt of triple duplicate ACK
Host BHost A
Seq=92 8 bytes of data
ACK=100
tim
eo
ut ACK=100
ACK=100
ACK=100
TCP fast retransmit
Seq=100 20 bytes of data
Seq=100 20 bytes of data
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-41
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-42
TCP flow controlapplication
process
TCP socketreceiver buffers
TCPcode
IPcode
application
OS
receiver protocol stack
application may remove data from
TCP socket buffers hellip
hellip slower than TCP
receiver is delivering(sender is sending)
from sender
receiver controls sender so sender wonrsquot overflow receiverrsquos buffer by transmitting too much too fast
flow control
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-43
TCP flow control
buffered data
free buffer spacerwnd
RcvBuffer
TCP segment payloads
to application process
receiver ldquoadvertisesrdquo free buffer space by including rwnd value in TCP header of receiver-to-sender segments RcvBuffer size can be
set via socket options most operating systems
auto-adjust RcvBuffer sender limits amount
of unacked (ldquoin-flightrdquo) data to receiverrsquos rwnd value to ensure receive buffer will not overflow
receiver-side buffering
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-44
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-45
Connection Managementbefore exchanging data senderreceiver
ldquohandshakerdquo agree to establish connection (each knowing the
other willing to establish connection) agree on connection parameters
connection state ESTABconnection variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
connection state ESTABconnection Variables
seq client-to-server server-to-clientrcvBuffer size at serverclient
application
network
Socket clientSocket = newSocket(hostnameport
number)
Socket connectionSocket = welcomeSocketaccept()
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-46
Q will 2-way handshake always work in network
variable delays retransmitted messages
(eg req_conn(x)) due to message loss
message reordering canrsquot ldquoseerdquo other side
2-way handshake
Letrsquos talk
OKESTAB
ESTAB
choose xreq_conn(x)
ESTAB
ESTABacc_conn(x)
Agreeing to establish a connection
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-47
Agreeing to establish a connection
2-way handshake failure scenarios
retransmitreq_conn(
x)
ESTAB
req_conn(x)
half open connection(no client)
client terminat
es
serverforgets x
connection x completes
retransmitreq_conn(
x)
ESTAB
req_conn(x)
data(x+1)
retransmitdata(x+1)
acceptdata(x+1)
choose xreq_conn(x)
ESTAB
ESTAB
acc_conn(x)
client terminat
es
ESTAB
choose xreq_conn(x)
ESTAB
acc_conn(x)
data(x+1) acceptdata(x+1)
connection x completes server
forgets x
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-48
TCP 3-way handshake
SYNbit=1 Seq=x
choose init seq num xsend TCP SYN msg
ESTAB
SYNbit=1 Seq=yACKbit=1 ACKnum=x+1
choose init seq num ysend TCP SYNACKmsg acking SYN
ACKbit=1 ACKnum=y+1
received SYNACK(x) indicates server is livesend ACK for SYNACK
this segment may contain client-to-server data
received ACK(y) indicates client is live
SYNSENT
ESTAB
SYN RCVD
client state
LISTEN
server state
LISTEN
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-49
TCP 3-way handshake FSM
closed
listen
SYNrcvd
SYNsent
ESTAB
Socket clientSocket = newSocket(hostnameport
number)
SYN(seq=x)
Socket connectionSocket = welcomeSocketaccept()
SYN(x)
SYNACK(seq=yACKnum=x+1)create new socket for
communication back to client
SYNACK(seq=yACKnum=x+1)
ACK(ACKnum=y+1)ACK(ACKnum=y+1)
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-50
TCP closing a connection client server each close their side of
connection send TCP segment with FIN bit = 1
respond to received FIN with ACK on receiving FIN ACK can be combined with
own FIN simultaneous FIN exchanges can be
handled
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-51
FIN_WAIT_2
CLOSE_WAIT
FINbit=1 seq=y
ACKbit=1 ACKnum=y+1
ACKbit=1 ACKnum=x+1 wait for server
close
can stillsend data
can no longersend data
LAST_ACK
CLOSED
TIMED_WAIT
timed wait for 2max
segment lifetime
CLOSED
TCP closing a connection
FIN_WAIT_1 FINbit=1 seq=xcan no longersend but can receive data
clientSocketclose()
client state server state
ESTABESTAB
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
TCP Overall state machine
Transport Layer 3-52
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-53
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-54
congestion informally ldquotoo many sources sending
too much data too fast for network to handlerdquo
different from flow control manifestations
lost packets (buffer overflow at routers)
long delays (queueing in router buffers)
a top-10 problem
Principles of congestion control
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-55
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
output link capacity R
no retransmission
maximum per-connection throughput R2
unlimited shared output link buffers
Host A
original data in
Host B
throughputout
R2
R2
out
in R2d
ela
yin
large delays as arrival rate in approaches capacity
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-56
one router finite buffers sender retransmission of timed-out packet
app-layer input = app-layer outputin = out
transport-layer input includes retransmissions rsquoin
ge in
finite shared output link buffers
Host A
in original data
Host B
outin original data plus
retransmitted data
Causescosts of congestion scenario 2
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-57
idealization perfect knowledge
sender sends only when router buffers available
finite shared output link buffers
in original dataoutin original data plus
retransmitted data
copy
free buffer space
R2
R2
out
in
Causescosts of congestion scenario 2
Host B
A
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-58
in original dataoutin original data plus
retransmitted data
copy
no buffer space
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
Causescosts of congestion scenario 2
A
Host B
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-59
in original dataoutin original data plus
retransmitted data
free buffer space
Causescosts of congestion scenario 2
Idealization known loss packets can be lost dropped at router due to full buffers
sender only resends if packet known to be lost
R2
R2in
out
when sending at R2 some packets are retransmissions but asymptotic goodput is still R2 (why)
A
Host B
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-60
A
in outincopy
free buffer space
timeout
R2
R2in
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
Host B
Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Causescosts of congestion scenario 2
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-61
R2
out
when sending at R2 some packets are retransmissions including duplicated that are delivered
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple
copies of pkt decreasing goodput
R2in
Causescosts of congestion scenario 2 Realistic duplicates packets can be lost
dropped at router due to full buffers
sender times out prematurely sending two copies both of which are delivered
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-62
four senders multihop paths timeoutretransmit
Q what happens as in and rsquo
in increase
finite shared output link buffers
Host A out
Causescosts of congestion scenario 3
Host B
Host CHost D
in original data
in original data plus
retransmitted data
A as red rsquoin increases all
arriving blue pkts at upper queue are dropped blue throughput 0
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-63
another ldquocostrdquo of congestion when packet dropped any ldquoupstream
bandwidth used for that packet wasted
Causescosts of congestion scenario 3
C2
C2
ou
t
inrsquo
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-64
Approaches towards congestion controltwo broad approaches towards congestion
controlend-end
congestion control
no explicit feedback from network
congestion inferred from end-system observed loss delay
approach taken by TCP
network-assisted congestion control
routers provide feedback to end systemssingle bit indicating congestion (SNA DECbit TCPIP ECN ATM)
explicit rate for sender to send at
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-65
Case study ATM ABR congestion control
ABR available bit rate
ldquoelastic servicerdquo if senderrsquos path
ldquounderloadedrdquo sender should
use available bandwidth
if senderrsquos path congested sender throttled
to minimum guaranteed rate
RM (resource management) cells
sent by sender interspersed with data cells
bits in RM cell set by switches (ldquonetwork-assistedrdquo) NI bit no increase in rate
(mild congestion) CI bit congestion
indication RM cells returned to sender
by receiver with bits intact
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-66
Case study ATM ABR congestion control
two-byte ER (explicit rate) field in RM cell congested switch may lower ER value in cell sendersrsquo send rate thus max supportable rate on path
EFCI bit in data cells set to 1 in congested switch if data cell preceding RM cell has EFCI set receiver
sets CI bit in returned RM cell
RM cell data cell
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-67
Transport Layer Outline
1 transport-layer services
2 multiplexing and demultiplexing
3 connectionless transport UDP
4 connection-oriented transport TCP segment structure reliable data
transfer flow control connection
management5 principles of
congestion control6 TCP congestion
control
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-68
TCP congestion control additive increase multiplicative decrease
approach sender increases transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase cwnd by 1
MSS every RTT until loss detected multiplicative decrease cut cwnd in half
after loss
cwnd
TC
P s
ende
r co
nges
tion
win
dow
siz
e
AIMD saw toothbehavior probing
for bandwidth
additively increase window size helliphellip until loss occurs (then cut window in half)
time
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-69
TCP congestion control window
sender limits transmission
cwnd is dynamic function of perceived congestion
TCP sending rate roughly send
cwnd bytes wait RTT for ACKS then send more bytes
last byteACKed sent not-
yet ACKed(ldquoin-flightrdquo)
last byte sent
cwnd
LastByteSent -LastByteAcked
lt cwnd
sender sequence number space
rate ~~cwnd
RTTbytessec
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-70
TCP Slow Start when connection
begins increase rate exponentially until first loss event initially cwnd = 1 MSS double cwnd every
RTT done by incrementing cwnd upon every ACK
summary initial rate is slow but ramps up exponentially fast
Host A
one segment
RT
T
Host B
time
two segments
four segments
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-71
TCP detecting reacting to loss
loss indicated by timeout cwnd set to 1 MSS window then grows exponentially (as in slow start) to threshold
then grows linearly loss indicated by 3 duplicate ACKs TCP RENO
dup ACKs indicate network capable of delivering some segments cwnd is cut in half window then grows linearly
TCP Tahoe always sets cwnd to 1 (timeout or 3 duplicate acks)
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-72
Q when should the exponential increase switch to linear
A when cwnd gets to 12 of its value before timeout
Implementation variable ssthresh on loss event ssthresh is set to 12 of cwnd just before loss event
TCP slow start cong avoidance
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-73
Summary TCP Congestion Control
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd gt ssthresh
congestionavoidance
cwnd = cwnd + MSS (MSScwnd)dupACKcount = 0
transmit new segment(s) as allowed
new ACK
dupACKcount++
duplicate ACK
fastrecovery
cwnd = cwnd + MSStransmit new segment(s) as allowed
duplicate ACK
ssthresh= cwnd2cwnd = ssthresh + 3
retransmit missing segment
dupACKcount == 3
timeoutssthresh = cwnd2cwnd = 1 dupACKcount = 0retransmit missing segment
ssthresh= cwnd2cwnd = ssthresh + 3retransmit missing segment
dupACKcount == 3cwnd = ssthreshdupACKcount = 0
New ACK
slow start
timeoutssthresh = cwnd2
cwnd = 1 MSSdupACKcount = 0
retransmit missing segment
cwnd = cwnd+MSSdupACKcount = 0transmit new segment(s) as allowed
new ACKdupACKcount++
duplicate ACK
cwnd = 1 MSS
ssthresh = 64 KBdupACKcount = 0
NewACK
NewACK
NewACK
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-74
TCP throughput Simplistic model avg TCP thruput as function of window
size RTT ignore slow start assume always data to send
W window size (measured in bytes) where loss occurs avg window size ( in-flight bytes) is frac34 W avg throughput is 34W per RTT
W
W2
avg TCP thruput = 34W
RTTbytessec
In practice W not known or fixed so this model is too simplistic to be useful
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-75
TCP throughput More practical model Throughput in terms of segment loss
probability L round-trip time T and maximum segment size M [Mathis et al 1997]
TCP throughput = 122 MT L
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-76
TCP futures TCP over ldquolong fat pipesrdquo example 1500 byte segments 100ms
RTT want 10 Gbps throughput requires W = 83333 in-flight segments
as per the throughput formula
to achieve 10 Gbps throughput need a loss rate of L = 210-10 ndash an unrealistically small loss rate
new versions of TCP for high-speed
TCP throughput = 122 MSSRTT L
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
TCP throughput wrap-up
Assume sender window cwnd receiver window rwnd bottleneck capacity C round-trip time T path loss rate L maximum segment size MSS Then Instantaneous TCP throughput =
bull min(C cwndTrwndT) Steady-state TCP throughput =
bull min(C 122M(TradicL))
Transport Layer 3-77
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-78
fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP Fairness
TCP connection 2
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-79
Why is TCP fairtwo competing sessions additive increase gives slope of 1 as throughout
increases multiplicative decrease decreases throughput
proportionally R
R
equal bandwidth share
Connection 1 throughput
Con
nect
ion
2 th
roug
h pu t
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-
Transport Layer 3-80
Fairness (more)Fairness and UDP multimedia apps
often do not use TCP rate throttling by
congestion control can hurt streaming quality
instead use UDP send audiovideo
at constant rate tolerate packet loss
Fairness parallel TCP connections
application can open many parallel connections between two hosts
web browsers do this eg link of rate R with 9
existing connections new app asks for 1 TCP gets
R10 new app asks for 11 TCPs gets
R2
- Transport Layer UDP and TCP
- Transport Layer Outline
- Transport services and protocols
- Transport vs network layer
- Internet transport-layer protocols
- Slide 6
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux example
- Connection-oriented demux
- Connection-oriented demux example
- Slide 13
- Slide 14
- UDP User Datagram Protocol [RFC 768]
- UDP segment header
- UDP checksum
- Internet checksum example
- Q1 Sockets and multiplexing
- Q2 Sockets UDP
- Q3 Sockets TCP
- Q4 Sockets TCP
- Q5 UDP checksums
- Slide 24
- TCP Overview RFCs 79311221323 2018 2581
- TCP segment structure
- TCP seq numbers ACKs
- Slide 28
- TCP round trip time timeout
- Slide 30
- Slide 31
- Slide 32
- TCP reliable data transfer
- TCP sender events
- TCP sender (simplified)
- TCP retransmission scenarios
- Slide 37
- TCP ACK generation [RFC 1122 RFC 2581]
- TCP fast retransmit
- Slide 40
- Slide 41
- TCP flow control
- Slide 43
- Slide 44
- Connection Management
- Agreeing to establish a connection
- Slide 47
- TCP 3-way handshake
- TCP 3-way handshake FSM
- TCP closing a connection
- Slide 51
- TCP Overall state machine
- Slide 53
- Principles of congestion control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Slide 57
- Slide 58
- Slide 59
- Slide 60
- Slide 61
- Causescosts of congestion scenario 3
- Slide 63
- Approaches towards congestion control
- Case study ATM ABR congestion control
- Slide 66
- Slide 67
- TCP congestion control additive increase multiplicative decrease
- TCP congestion control window
- TCP Slow Start
- TCP detecting reacting to loss
- TCP slow start cong avoidance
- Summary TCP Congestion Control
- TCP throughput Simplistic model
- TCP throughput More practical model
- TCP futures TCP over ldquolong fat pipesrdquo
- TCP throughput wrap-up
- TCP Fairness
- Why is TCP fair
- Fairness (more)
-