Katz, Stoica F04

EECS 122: Introduction to Computer Networks

TCP Variations

Computer Science Division

Department of Electrical Engineering and Computer Sciences

University of California, Berkeley

Berkeley, CA 94720-1776

2Katz, Stoica F04

Today’s Lecture: 11

Network (IP)

Application

Transport

Link

Physical

2

7, 8, 9

10,11

17, 18, 19

14, 15, 16

21, 22, 23

25

6

3Katz, Stoica F04

Outline

TCP congestion control- Quick Review- TCP flavors- Equation-based congestion control- Impact of losses- Cheating

Router-based support- RED- ECN- Fair Queueing- XCP

4Katz, Stoica F04

Quick Review

Slow-Start: cwnd++ upon every new ACK

Congestion avoidance: AIMD if cwnd > ssthresh- ACK: cwnd = cwnd + 1/cwnd

- Drop: ssthresh =cwnd/2 and cwnd=1

Fast Recovery:- duplicate ACKS: cwnd=cwnd/2

- Timeout: cwnd=1

5Katz, Stoica F04

TCP Flavors

TCP-Tahoe- cwnd =1 whenever drop is detected

TCP-Reno- cwnd =1 on timeout

- cwnd = cwnd/2 on dupack

TCP-newReno- TCP-Reno + improved fast recovery

TCP-Vegas, TCP-SACK

6Katz, Stoica F04

TCP Vegas

Improved timeout mechanism

Decrease cwnd only for losses sent at current rate- avoids reducing rate twice

Congestion avoidance phase:- compare Actual rate (A) to Expected rate (E)

- if E-A > , decrease cwnd linearly

- if E-A < , increase cwnd linearly

- rate measurements ~ delay measurements

- see textbook for details!

7Katz, Stoica F04

TCP-SACK

SACK = Selective Acknowledgements

ACK packets identify exactly which packets have arrived

Makes recovery from multiple losses much easier

8Katz, Stoica F04

Standards?

How can all these algorithms coexist?

Don’t we need a single, uniform standard?

What happens if I’m using Reno and you are using Tahoe, and we try to communicate?

9Katz, Stoica F04

Equation-Based CC

Simple scenario- assume a drop every k’th RTT (for some large k)

- w, w+1, w+2, ...w+k-1 DROP (w+k-1)/2, (w+k-1)/2+1,...

Observations:- In steady state: w= (w+k-1)/2 so w=k-1

- Average window: 1.5(k-1)

- Total packets between drops: 1.5k(k-1)

- Drop probability: p = 1/[1.5k(k-1)]

Throughput: T ~ (1/RTT)*sqrt(3/2p)

10Katz, Stoica F04

Equation-Based CC

Idea:- Forget complicated increase/decrease algorithms

- Use this equation T(p) directly!

Approach:- measure drop rate (don’t need ACKs for this)

- send drop rate p to source

- source sends at rate T(p)

Good for streaming audio/video that can’t tolerate the high variability of TCP’s sending rate

11Katz, Stoica F04

Question!

Why use the TCP equation?

Why not use any equation for T(p)?

12Katz, Stoica F04

Cheating

Three main ways to cheat:- increasing cwnd faster than 1 per RTT

- using large initial cwnd

- Opening many connections

13Katz, Stoica F04

Increasing cwnd Faster

A Bx

D Ey

Limit rates:x = 2y

C

x

y

x increases by 2 per RTTy increases by 1 per RTT

14Katz, Stoica F04

Increasing cwnd Faster

A Bx

D Ey

0

10

20

30

40

50

60

1 28 55 82 109

136

163

190

217

244

271

298

325

352

379

406

433

460

487

15Katz, Stoica F04

Larger Initial cwnd

A Bx

D Ey

x starts SS with cwnd = 4y starts SS with cwnd = 1

16Katz, Stoica F04

Open Many Connections

A Bx

D Ey

Assume • A starts 10 connections to B• D starts 1 connection to E• Each connection gets about the same throughput

Then A gets 10 times more throughput than D

17Katz, Stoica F04

Cheating and Game Theory

A Bx

D Ey

22, 22 10, 35

35, 10 15, 15

(x, y)A

Increases by 1

Increases by 5

D Increases by 1 Increases by 5

Individual incentives: cheating paysSocial incentives: better off without cheating

Classic PD: resolution depends on accountability

Too aggressiveLossesThroughput falls

18Katz, Stoica F04

Lossy Links

TCP assumes that all losses are due to congestion

What happens when the link is lossy?

Recall that Tput ~ 1/sqrt(p) where p is loss prob.

This applies even for non-congestion losses

19Katz, Stoica F04

Example

0

10

20

30

40

50

60

1 26 51 76 101 126 151 176 201 226 251 276 301 326 351 376 401 426 451 476

p = 0

p = 1%

p = 10%

Katz, Stoica F04

What can routers do to help?

21Katz, Stoica F04

Paradox

Routers are in middle of action

But traditional routers are verypassive in terms of congestioncontrol

- FIFO

- Drop-tail

22Katz, Stoica F04

FIFO: First-In First-Out

Maintain a queue to store all packets Send packet at the head of the queue

Queued packetsArriving packet

Next to transmit

23Katz, Stoica F04

Tail-drop Buffer Management

Drop packets only when buffer is full Drop arriving packet

Arriving packet

Next to transmit

Drop

24Katz, Stoica F04

Ways Routers Can Help

Packet scheduling: non-FIFO scheduling

Packet dropping: - not drop-tail

- not only when buffer is full

Congestion signaling

25Katz, Stoica F04

Question!

Why not use infinite buffers?- no packet drops!

26Katz, Stoica F04

The Buffer Size Quandary

Small buffers:- often drop packets due to bursts

- but have small delays

Large buffers:- reduce number of packet drops (due to bursts)

- but increase delays

Can we have the best of both worlds?

27Katz, Stoica F04

Random Early Detection (RED)

Basic premise:- router should signal congestion when the queue first

starts building up (by dropping a packet)

- but router should give flows time to reduce their sending rates before dropping more packets

Therefore, packet drops should be:- early: don’t wait for queue to overflow

- random: don’t drop all packets in burst, but space drops out

28Katz, Stoica F04

RED

FIFO scheduling Buffer management:

- Probabilistically discard packets

- Probability is computed as a function of average queue length (why average?)

Discard Probability

AverageQueue Length

0

1

min_th max_th queue_len

29Katz, Stoica F04

RED (cont’d)

min_th – minimum threshold max_th – maximum threshold avg_len – average queue length

- avg_len = (1-w)*avg_len + w*sample_len

Discard Probability

AverageQueue Length

0

1

min_th max_th queue_len

30Katz, Stoica F04

RED (cont’d)

If (avg_len < min_th) enqueue packet If (avg_len > max_th) drop packet If (avg_len >= min_th and avg_len < max_th)

enqueue packet with probability P

Discard Probability (P)

AverageQueue Length

0

1

min_th max_th queue_len

31Katz, Stoica F04

RED (cont’d)

P = max_P*(avg_len – min_th)/(max_th – min_th) Improvements to spread the drops (see textbook)

Discard Probability

AverageQueue Length

0

1

min_th max_th queue_len

avg_len

P

max_P

32Katz, Stoica F04

Average vs Instantaneous Queue

33Katz, Stoica F04

RED Advantages

High network utilization with low delays

Average queue length small, but capable of absorbing large bursts

Many refinements to basic algorithm make it more adaptive (requires less tuning)

34Katz, Stoica F04

Explicit Congestion Notification

Rather than drop packets to signal congestion, router can send an explicit signal

Explicit congestion notification (ECN):- instead of optionally dropping packet, router sets a bit in

the packet header

- If data packet has bit set, then ACK has ECN bit set

Backward compatibility:- bit in header indicates if host implements ECN

- note that not all routers need to implement ECN

35Katz, Stoica F04

Picture

WW/2

A B

36Katz, Stoica F04

ECN Advantages

No need for retransmitting optionally dropped packets

No confusion between congestion losses and corruption losses

37Katz, Stoica F04

Remaining Problem

Internet vulnerable to CC cheaters!

Single CC standard can’t satisfy all applications- EBCC might answer this point

Goal:- make Internet invulnerable to cheaters- allow end users to use whatever congestion control

they want

How?

38Katz, Stoica F04

One Approach: Nagle (1987)

Round-robin among different flows- one queue per flow

39Katz, Stoica F04

Round-Robin Discussion

Advantages: protection among flows- Misbehaving flows will not affect the performance of well-

behaving flows

• Misbehaving flow – a flow that does not implement any congestion control

- FIFO does not have such a property

Disadvantages:- More complex than FIFO: per flow queue/state

- Biased toward large packets – a flow receives service proportional to the number of packets

40Katz, Stoica F04

Solution?

Bit-by-bit round robin Can you do this in practice? No, packets cannot be preempted (why?) …we can only approximate it

41Katz, Stoica F04

Fair Queueing (FQ)

Define a fluid flow system: a system in which flows are served bit-by-bit

Then serve packets in the increasing order of their deadlines

Advantages- Each flow will receive exactly its fair rate

Note:- FQ achieves max-min fairness

42Katz, Stoica F04

Max-Min Fairness

Denote- C – link capacity- N – number of flows- ri – arrival rate

Max-min fair rate computation:1. compute C/N

2. if there are flows i such that ri <= C/N, update C and N

3. if no, f = C/N; terminate4. go to 1

A flow can receive at most the fair rate, i.e., min(f, ri)

NCrtsi ii

rCC/.

43Katz, Stoica F04

Example

C = 10; r1 = 8, r2 = 6, r3 = 2; N = 3 C/3 = 3.33 C = C – r3 = 8; N = 2 C/2 = 4; f = 4

8

6

244

2

f = 4: min(8, 4) = 4 min(6, 4) = 4 min(2, 4) = 2

10

44Katz, Stoica F04

Implementing Fair Queueing

Idea: serve packets in the order in which they would have finished transmission in the fluid flow system

45Katz, Stoica F04

Example

1 2 3 4 5

1 2 3 4

1 23

1 24

3 45

5 6

1 2 1 3 2 3 4 4

5 6

55 6

Flow 1(arrival traffic)

Flow 2(arrival traffic)

Servicein fluid flow

system

Packetsystem

time

time

time

time

46Katz, Stoica F04

System Virtual Time: V(t) Measure service, instead of time V(t) slope – rate at which every active flow receives service

- C – link capacity- N(t) – number of active flows in fluid flow system at time t

1 23

1 24

3 45

5 6Servicein fluid flow

systemtime

time

V(t)

)(

)(

tN

C

t

tV

47Katz, Stoica F04

Fair Queueing Implementation

Define- - finishing time of packet k of flow i (in system virtual

time reference system)

- - arrival time of packet k of flow i

- - length of packet k of flow i

The finishing time of packet k+1 of flow i is

kiL

kia

kiF

111 )),(max( ki

ki

ki

ki LFaVF

48Katz, Stoica F04

FQ Advantages

FQ protect well-behaved flows from ill-behaved flows Example: 1 UDP (10 Mbps) and 31 TCP’s sharing a 10 Mbps

link

FQ

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

1 4 7 10 13 16 19 22 25 28 31Flow Number

Thro

ughp

ut(M

bps)

RED

0

1

2

3

4

5

6

7

8

9

10

1 4 7 10 13 16 19 22 25 28 31Flow Number

Thro

ughp

ut(M

bps)

49Katz, Stoica F04

Alternative Implementations of Max-Min

Deficit round-robin Core-stateless fair queueing

- label packets with rate- drop according to rates- check at ingress to make sure rates are truthful

Approximate fair dropping- keep small sample of previous packets- estimate rates based on these- apply dropping as above- wins because few large flows

• per-elephant state, not per-mouse state

RED-PD: not max-min, but punishes big cheaters

50Katz, Stoica F04

Big Picture

FQ does not eliminate congestion it just manages the congestion

You need both end-host congestion control and router support for congestion control

- end-host congestion control to adapt

- router congestion control to protect/isolate

51Katz, Stoica F04

Explicit Rate Signaling (XCP)

Each packet contains: cwnd, RTT, feedback field

Routers indicate to flows whether to increase or decrease:

- give explicit rates for increase/decrease amounts

- feedback is carried back to source in ACK

Separation of concerns:- aggregate load

- allocation among flows

52Katz, Stoica F04

XCP (continued)

Aggregate:- measures spare capacity and avg queue size

- computes desired aggregate change: D=aRS-bQ

Allocation:- uses AIMD

- positive feedback is same for all flows

- negative feedback is proportional to current rate

- when D=0, reshuffle bandwidth

- all changes normalized by RTT

• want equal rates, not equal windows

53Katz, Stoica F04

XCP (continued)

Challenge: - how to give per-flow feedback without per-flow state?

- do you keep track of which flows you’ve signaled and which you haven’t?

Solution:- figure out desired change

- divide from expected number of packets from flow in time interval

- give each packet share of rate adjustment

- flow totals up all rate adjustment