Part IV – Wireline Multiuser Examples
description
Transcript of Part IV – Wireline Multiuser Examples
1 Multiuser WondersMultiuser Wonders
Part IV – Wireline Multiuser ExamplesPart IV – Wireline Multiuser Examples
March 31, 2001April 26 2001
September 9, 2001September 9, 2001
Prof. John M. CioffiProf. John M. Cioffi
Dept of EEDept of EE
Stanford UniversityStanford University
[email protected]@stanford.edu
2 Multiuser WondersMultiuser Wonders
Parts 3 and 4: Outline/ScheduleParts 3 and 4: Outline/Schedule
2:00-2:452:00-2:45 MU TheoryMU Theory 2:45-3:302:45-3:30 channels for wirelinechannels for wireline 3:30-4:00 3:30-4:00 CoffeeCoffee 4:00-4:304:00-4:30 DSL and Ethernet archDSL and Ethernet arch 4:30-5:154:30-5:15 Multiuser improvementsMultiuser improvements
3 Multiuser WondersMultiuser Wonders
Outline – Part 4Outline – Part 4
DSL and Ethernet ApplicationsDSL and Ethernet Applications Evolution of DSL and DSM vs SMEvolution of DSL and DSM vs SM
UnbundlingUnbundling Towards ethernetTowards ethernet
Multiuser Spectrum BalancingMultiuser Spectrum Balancing Vectored VDSLVectored VDSL Vectored EthernetVectored Ethernet
4 Multiuser WondersMultiuser Wonders
DSL in 2001DSL in 2001
Data, some voice beginning, to 1.5/.5 MbpsData, some voice beginning, to 1.5/.5 Mbps Conservative deployments (9 Million lines)Conservative deployments (9 Million lines)
Telephone company officeTelephone company office
ServiceProvider
split
POTSNetworkdigitaldigital
analoganalog
analoganalog
Splitter ?
0-4 miles0-4 miles
customer premisescustomer premises
GGAATTEEWWAAYYSS DSLAM
Voice
Muxor
Demux
ADSLmodem
ADSLmodem
.
.
ADSLmodem
InternetServiceProvider
digitaldigital
ADSL
5 Multiuser WondersMultiuser Wonders
DSL Future ApplicationsDSL Future Applications
Multiline voice (VoDSL)Multiline voice (VoDSL) Video (internet, packet based)Video (internet, packet based)
Not likely “traditional” TVNot likely “traditional” TV Small Businesses – less asymmetricSmall Businesses – less asymmetric
filt
Enet.
802.11
VoDSLinterface
Home/business wiring
A/VDSLA/VDSL
Enet
100BT100BTGigabit EGigabit E
wireless (802.11) orwireless (802.11) orother distribution possibleother distribution possiblewithin CPEwithin CPE
ToToserviceservice
providerprovider routerrouter
6 Multiuser WondersMultiuser Wonders
Part Fiber, Part DSL = VDSLPart Fiber, Part DSL = VDSL
Only 12,000 of 10Only 12,000 of 1077 businesses connected by fiber all the way – essentially 0 residences businesses connected by fiber all the way – essentially 0 residences Rather ONU’s/remote terminals, incrementally with timeRather ONU’s/remote terminals, incrementally with time
How long is fiber? Twisted-Pair?How long is fiber? Twisted-Pair? Incremental trade-off with time/demandIncremental trade-off with time/demand
.1- 2 km
Split
VDSL
ONU
fiber
To 100 MbpsTo 100 Mbps
Split
POTSC-Bank
VDSL
LT
7 Multiuser WondersMultiuser Wonders
Ethernet – 10BTEthernet – 10BT
STAR wiring – collision detection not really usedSTAR wiring – collision detection not really used Hub is electronic version of a “coax-wired” connectionHub is electronic version of a “coax-wired” connection All lines on Hub share the 10 MbpsAll lines on Hub share the 10 Mbps
100’s millions deployed on copper100’s millions deployed on copper Manchester codingManchester coding Category 5 tp (20 dB less xtalk than phone lines- cat 3, 24 gauge)Category 5 tp (20 dB less xtalk than phone lines- cat 3, 24 gauge) Each line is 2 pairs (so 2 of 4 in cat 5)Each line is 2 pairs (so 2 of 4 in cat 5)
100 meters (systems run longer)100 meters (systems run longer) Result of delay specification related to days in which HUB was a coax.Result of delay specification related to days in which HUB was a coax.
HubHub(yellow(yellowcoax)coax)
InternetInternetRouterRouter
ServerServer
8 Multiuser WondersMultiuser Wonders
Introduction of BridgesIntroduction of Bridges
Little higher level than Hub, which is an electronic Little higher level than Hub, which is an electronic version of a “yellow coax”version of a “yellow coax”
Allows each line’s 10 Mbps to be differentAllows each line’s 10 Mbps to be different Dedicated connection to each userDedicated connection to each user
HubHub
InternetInternetRouterRouter
bridgebridgeServerServer
HubHub
9 Multiuser WondersMultiuser Wonders
Ethernet – 100BTEthernet – 100BT
Each user gets 100 Mbps, dedicated linkEach user gets 100 Mbps, dedicated link MLT3 code (4B5B to ternary) – 125 MHzMLT3 code (4B5B to ternary) – 125 MHz 100 meter range – back compatible with 10BT100 meter range – back compatible with 10BT
Data sent when packets availableData sent when packets available Delay spec really not needed on link anymoreDelay spec really not needed on link anymore Remains a 2-wire duplex situationRemains a 2-wire duplex situation Lead suppliers do 170m rangeLead suppliers do 170m range
BridgeBridge(router)(router)
ServerServer
InternetInternet
10 Multiuser WondersMultiuser Wonders
Ethernet – 1000BTEthernet – 1000BT
Continuous, 250 Mbps on each of 4 lines for 1 Gbps totalContinuous, 250 Mbps on each of 4 lines for 1 Gbps total Continuous onlyContinuous only All 4 wires in cat 5 used by a single user in duplex, echo-cancelled fashionAll 4 wires in cat 5 used by a single user in duplex, echo-cancelled fashion 125 MHz 5-level (effective 4, or 2 bits) PAM on each line125 MHz 5-level (effective 4, or 2 bits) PAM on each line
100 meters – physical layer constraint (lead suppliers do 160 m)100 meters – physical layer constraint (lead suppliers do 160 m)
RouterRouterServerServer
InternetInternet
44
11 Multiuser WondersMultiuser Wonders
Ethernet Faster?Ethernet Faster?
10,000BT – vectored multiuser, 100 10,000BT – vectored multiuser, 100 metersmeters
RouterRouterServerServer
InternetInternet
44
VDMTVDMT
VDMTVDMT
VDMTVDMT VDMTVDMT
VDMTVDMTVDMTVDMT
VDMTVDMTVDMTVDMT
VDMTVDMT
12 Multiuser WondersMultiuser Wonders
Ethernet – longer (EFM)Ethernet – longer (EFM)
VDSL=VDMTVDSL=VDMT 100BT (4 wires) > 1 km100BT (4 wires) > 1 km Single wire – very high speeds, Single wire – very high speeds,
Distance/rate tradeoffDistance/rate tradeoff Vectored/spectrally balanced – number of pairs Vectored/spectrally balanced – number of pairs
allocated, etc. allocated, etc.
RouterRouterDSLAMDSLAM
ServerServer
InternetInternet
44
VDMTVDMT
VDMTVDMT
VDMTVDMT VDMTVDMT
VDMTVDMTVDMTVDMT
VDMTVDMTVDMTVDMT
VDMTVDMT
Sites may nowSites may nowbe at homes/businessesbe at homes/businessesat end of telco linesat end of telco lines
13 Multiuser WondersMultiuser Wonders
Outline – Part 4Outline – Part 4
DSL and Ethernet ApplicationsDSL and Ethernet Applications Evolution of DSL and DSM vs SMEvolution of DSL and DSM vs SM
UnbundlingUnbundling Towards ethernetTowards ethernet
Multiuser Spectrum BalancingMultiuser Spectrum Balancing Vectored VDSLVectored VDSL Vectored EthernetVectored Ethernet
14 Multiuser WondersMultiuser Wonders
Unbundling in DSLUnbundling in DSL
Different service providers can “rent” lines to Different service providers can “rent” lines to customerscustomers Consequent emissions between lines (crosstalk)Consequent emissions between lines (crosstalk)
Gets worse with wider bandwidths (shorter lines)Gets worse with wider bandwidths (shorter lines) Can be reduced with asymmetric transmissionCan be reduced with asymmetric transmission
ILEC – often wants asymmetricILEC – often wants asymmetric CLEC – often wants symmetricCLEC – often wants symmetric
ILECILEC
CLECCLEC
15 Multiuser WondersMultiuser Wonders
Spectrum ManagementSpectrum Management
Regulate the spectra of the different service Regulate the spectra of the different service providers within the cableproviders within the cable Minimize radiation between cablesMinimize radiation between cables Balance interest of symmetric/asymmetric (CLEC/ILEC)Balance interest of symmetric/asymmetric (CLEC/ILEC)
Fixed spectra defined for each type of DSL Fixed spectra defined for each type of DSL service, no matter where it is used in the world service, no matter where it is used in the world STATICSTATIC spectrum management spectrum management
CO-based SM before FCC for approvalCO-based SM before FCC for approval ONU/fiber-feed case still under studyONU/fiber-feed case still under study
16 Multiuser WondersMultiuser Wonders
Network
fiberCentral
Office SP #1 LTSP #1..
twisted pair, DSL
Fiber, FTTH
SP #2 SP #2..
multiple fibersin this area lead
to multiple fibersto home
DSL Line-Unbundled Evolution?DSL Line-Unbundled Evolution?
Fiber for each service provider?Fiber for each service provider? Space at LT for each service provider?Space at LT for each service provider?
17 Multiuser WondersMultiuser Wonders
HFC – Hybrid Fiber CoaxHFC – Hybrid Fiber Coax
CableCableHead endHead end
satellitesatellite
splitsplitfiberfiber
Fiber from head end to first splitFiber from head end to first split More bandwidth, easier bidirectionallyMore bandwidth, easier bidirectionally Coax still shared among 100’s of users (500)Coax still shared among 100’s of users (500) Multiple content providers, unbundling?Multiple content providers, unbundling?
Fiber or coaxFiber or coax
coaxcoax
50 MHz50 MHz500 MHz500 MHz
5-40 MHz5-40 MHz
??
18 Multiuser WondersMultiuser Wonders
Packet UnbundlingPacket Unbundling
Single fiber/LT controlled by one service providerSingle fiber/LT controlled by one service provider SBC example: FCC allowed in Sept 2000 at LT onlySBC example: FCC allowed in Sept 2000 at LT only
Services unbundled at packet levelServices unbundled at packet level MAC in LT to control crosstalk problemsMAC in LT to control crosstalk problems
Dynamic spectrum managementDynamic spectrum management
CentralCentralOfficeOffice
Service 1Service 1
LTLTfiberfiber
twisted pair, DSLtwisted pair, DSL
Fiber, FTTHFiber, FTTH
Service 2Service 2
xtalkxtalk
19 Multiuser WondersMultiuser Wonders
The Main Technology IssueThe Main Technology Issue(a prime app for MU)(a prime app for MU)
Crosstalk – interference between linesCrosstalk – interference between lines Increases with frequency (data rate)Increases with frequency (data rate) Largest source of noise and performance lossLargest source of noise and performance loss Requires spectrum managementRequires spectrum management
Cross-section of cableAlso, space
20 Multiuser WondersMultiuser Wonders
xtalk coupling “resistance”xtalk coupling “resistance”
105
106
107
108
20
25
30
35
40
45
50
55
60
65
Freq (MHz)
cros
stal
k co
uplin
g (d
B)
21 Multiuser WondersMultiuser Wonders
Static Spectrum ManagementStatic Spectrum Management
Up (green) does not like down (yellow)Up (green) does not like down (yellow) Avoid overlap of yellow with green = static SMAvoid overlap of yellow with green = static SM
Fix allocations to compromise ILEC/CLECFix allocations to compromise ILEC/CLEC but xtalk/lines are different as are customers in each but xtalk/lines are different as are customers in each
locationlocation
ILECILEC
CLECCLEC
22 Multiuser WondersMultiuser Wonders
DYNAMIC SM = MU MethodsDYNAMIC SM = MU Methods
Line spectra/signals varied according to situationLine spectra/signals varied according to situation Data rates/symmetries of customersData rates/symmetries of customers Relative line lengthsRelative line lengths Topology of cable loopsTopology of cable loops
Large Improvements PossibleLarge Improvements Possible Migration Strategy to DSL/DSM of future Migration Strategy to DSL/DSM of future
necessarynecessary Line to packet unbundlingLine to packet unbundling Static to dynamic spectrum managementStatic to dynamic spectrum management
23 Multiuser WondersMultiuser Wonders
DSL Evolution GoalDSL Evolution Goal
ADSL VDSL VDSL- EFM VEFM
VDSL Evolution
6 10
25
100
0.65
25
100
0
20
40
60
80
100
120
4 km 2 km 1 km .3 km
TP Length
Dat
a R
ate
(Mbp
s)
Down
Up
24 Multiuser WondersMultiuser Wonders
Loop Plant of a Service ProviderLoop Plant of a Service Provider
Convenient pointsConvenient points RT, SAI (distribution node), pedestalRT, SAI (distribution node), pedestal
CentralOffice
MainDistributingFrame
equipment
SAI
feeder
distribution
pedestals
drop wire
Inside wire
Telephone Loop Plantcustomerpremises
20,000 to 1,500 to 200 to 4 to160,000 4,000 800 12 number of lines present at a site
22,000 feet 9,000 feet 3,000 feet 500 ft wire length to customer (90th percentile)
NID
25 Multiuser WondersMultiuser Wonders
Dynamic Spectrum Management StepsDynamic Spectrum Management Steps
Spectrum balancing – line unbundling Spectrum balancing – line unbundling For existing “line unbundling” situation where 3For existing “line unbundling” situation where 3rdrd party can make party can make
recommendations to individual lines, all or somerecommendations to individual lines, all or some Evolution to packet unbundling where situations are mixed pac/lineEvolution to packet unbundling where situations are mixed pac/line Implementation possible with current ADSL, VDSL systemsImplementation possible with current ADSL, VDSL systems
Vectoring – packet unbundlingVectoring – packet unbundling Lines coordinated at ONU side (or CO side) in terms of signals Lines coordinated at ONU side (or CO side) in terms of signals
placed on the line (packet unbunding)placed on the line (packet unbunding) Highest possible performance levels, shortest linesHighest possible performance levels, shortest lines Add-on (backward compatible) with ADSL, VDSLAdd-on (backward compatible) with ADSL, VDSL
26 Multiuser WondersMultiuser Wonders
Towards EthernetTowards Ethernet
DSLAMsDSLAMs Move toward customerMove toward customer Speeds go up and DSM usedSpeeds go up and DSM used Ethernet reused on top of DSLEthernet reused on top of DSL
Multiuser methodsMultiuser methods Spectrum balancingSpectrum balancing VectoringVectoring Combination of pairs to get 10, 100, or Combination of pairs to get 10, 100, or
10001000
27 Multiuser WondersMultiuser Wonders
Outline – Part 4Outline – Part 4
DSL and Ethernet ApplicationsDSL and Ethernet Applications Evolution of DSL and DSM vs SMEvolution of DSL and DSM vs SM
UnbundlingUnbundling Towards ethernetTowards ethernet
Multiuser Spectrum BalancingMultiuser Spectrum Balancing Vectored VDSLVectored VDSL Vectored EthernetVectored Ethernet
28 Multiuser WondersMultiuser Wonders
Spectrum BalancingSpectrum Balancing
ILEC
CLEC
Lines report informationLines report information Helps with deployment, problem isolationHelps with deployment, problem isolation Can be used for DSMCan be used for DSM
Cent main recommends line spectraCent main recommends line spectra
Central DSM/DSL MaintenanceCentral DSM/DSL Maintenance
29 Multiuser WondersMultiuser Wonders
Need FiberPremium ServiceOrMaintenance necessary
Translation to Telco TermsTranslation to Telco Terms
Problem?
Data Rate?Cheap service
time of daytime of day as availableas available guaranteedguaranteed
30 Multiuser WondersMultiuser Wonders
DSLAMDSLAM MDFMDF&/or&/or
splitter(s)splitter(s)
ADSLADSL
ADSLADSL
ADSLADSL
Network MaintenanceNetwork Maintenance
POTSPOTSSwitchSwitch
MLTMLT
Telco’s want to knowTelco’s want to know prevent problemsprevent problems make $ make $
>$>$1B/rboc-ptt/yr1B/rboc-ptt/yr ! ! Spec ManagementSpec Management
DSLTDSLTADSLADSL
DynamicDynamicSpectrumSpectrum
ManagementManagement
31 Multiuser WondersMultiuser Wonders
Rate REGIONSRate REGIONS
Plot of all possible rates of linesPlot of all possible rates of lines UpstreamUpstream DownstreamDownstream
Any point in region is possible, but each with different spectraAny point in region is possible, but each with different spectra Varies for each cable and loop topologyVaries for each cable and loop topology Varies for each combination of desired (allowed) ratesVaries for each combination of desired (allowed) rates
RRshortshort
RRlonglong
Spectral pair 1Spectral pair 1
Spectral pair 2Spectral pair 2
32 Multiuser WondersMultiuser Wonders
Simple Example –PBOSimple Example –PBO
4 lines at 3000’ (7.8 Mbps upstream)4 lines at 3000’ (7.8 Mbps upstream) Line at 500, 1000, 1500, 2000, 2500’Line at 500, 1000, 1500, 2000, 2500’ Compare against best SSMCompare against best SSM
Very little coordination (power of line, rate)Very little coordination (power of line, rate)
500500 14.0 28.514.0 28.510001000 11.5 24.011.5 24.015001500 10.0 19.510.0 19.520002000 9.0 15.59.0 15.52500 8.5 11.02500 8.5 11.0
Ref length(SSM) iterwater.Ref length(SSM) iterwater.
33 Multiuser WondersMultiuser Wonders
So far, Static SMSo far, Static SM
PlansPlans 998 (USA) – more asymmetric998 (USA) – more asymmetric
997 (Europe) – more symmetric997 (Europe) – more symmetric
Flex planFlex plan Number of bands programmableNumber of bands programmable Start/stop frequencies programmableStart/stop frequencies programmable USA VDSL Standard – part 3 (allows DSM)USA VDSL Standard – part 3 (allows DSM)
34 Multiuser WondersMultiuser Wonders
998 with spectrum balancing998 with spectrum balancing
4 6 8 10 12 14 16 18 20 22 241
2
3
4
5
6
7
8
9Rate Region: 3000ft loops vs 1000ft loops: Upstream, Plan 998
Mbps
Mbp
s
0 5 10 15 20 25 30 35 40 45 5020
21
22
23
24
25
26Rate Region: 3000ft loops vs 1000ft loops: Downstream, Plan 998
MbpsM
bps
Line1500 ft versus 3000 ftLine1500 ft versus 3000 ft 26/6 on 3000’ while 30/22 on 1500’26/6 on 3000’ while 30/22 on 1500’ Static SM only 18/1.5 and 6/6Static SM only 18/1.5 and 6/6
DownUp
35 Multiuser WondersMultiuser Wonders
flex with spectrum balancingflex with spectrum balancing
Line1500 ft versus 3000 ft.Line1500 ft versus 3000 ft. 26/13 Mbps on 3000’, 52/26 Mbps on 1500’26/13 Mbps on 3000’, 52/26 Mbps on 1500’
5 10 15 20 25 30 352
4
6
8
10
12
14
16Rate Region: 3000ft loops vs 1000ft loops: Upstream, Flexible Plan
Mbps
Mbp
s
0 10 20 30 40 50 6020
21
22
23
24
25
26
27Rate Region: 3000ft loops vs 1000ft loops: Downstream, Flexible Plan
Mbps
Mbp
s
DownUp
36 Multiuser WondersMultiuser Wonders
VDSL and ADSLVDSL and ADSL
1 1.5 2 2.5 3 3.5 4 4.5 5 5.520
21
22
23
24
25
26
27
28
29
30
12000ft ADSL downstream rate (Mbps)
3000
ft V
DS
L do
wns
trea
m ra
te (M
bps)
ADSL vs VDSL: co-location at CP
1 1.5 2 2.5 3 3.5 4 4.5 5 5.59
10
11
12
13
14
15
16
17
18
12000ft ADSL downstream rate (Mbps)
4500
ft V
DS
L do
wns
trea
m ra
te (M
bps)
ADSL vs VDSL: co-location at CP
Yellow on short line acts like greenYellow on short line acts like green 9000’ ADSL (fig 6)9000’ ADSL (fig 6) 26/3 possible on 3000’ VDSL while ADSL runs 5/.5.26/3 possible on 3000’ VDSL while ADSL runs 5/.5.
Down 3000’ Down 4500’
37 Multiuser WondersMultiuser Wonders
VDSL and ADSL with Hdsl/IdsnVDSL and ADSL with Hdsl/Idsn
9000’ ADSL9000’ ADSL
Down 3000’ Down 4500’
0 0.5 1 1.5 2 2.5 320
21
22
23
24
25
26
27
28
29
30
12000ft ADSL downstream rate (Mbps)
3000
ft V
DS
L do
wns
trea
m ra
te (M
bps)
ADSL vs VDSL: co-location at CP
0 0.5 1 1.5 2 2.5 38
9
10
11
12
13
14
15
16
17
18
12000ft ADSL downstream rate (Mbps)
4500
ft V
DS
L do
wns
trea
m ra
te (M
bps)
ADSL vs VDSL: co-location at CP
38 Multiuser WondersMultiuser Wonders
Outline – Part 4Outline – Part 4
DSL and Ethernet ApplicationsDSL and Ethernet Applications Evolution of DSL and DSM vs SMEvolution of DSL and DSM vs SM
UnbundlingUnbundling Towards ethernetTowards ethernet
Multiuser Spectrum BalancingMultiuser Spectrum Balancing Vectored VDSLVectored VDSL Vectored EthernetVectored Ethernet
39 Multiuser WondersMultiuser Wonders
Vectoring ResultsVectoring Results
500 1000 1500 2000 2500 3000 3500 4000 450010
20
30
40
50
60
70
80
90
Loop length (ft)
Da
ta r
ate
(M
bp
s)
Downs tream with Nois e A
997
998
V ec tored-V DS L
DownDown Up (see correction)Up (see correction)
500 1000 1500 2000 2500 3000 3500 4000 45000
10
20
30
40
50
60
70
80
Loop length (ft)
Dat
a ra
te (
Mbp
s)
Upstream with Noise A
997 998 Vectored-VDSL
40 Multiuser WondersMultiuser Wonders
Rate RegionsRate Regions
Last slide had all lines same lengthLast slide had all lines same length Rate Region allows tradeoffs between lines (1500, 1000)Rate Region allows tradeoffs between lines (1500, 1000)
0 50 100 1500
50
100
150
Data rate of user 1 (M bps )
Da
ta r
ate
of
us
er
2 (
Mb
ps
)
Rate regions (user 1 at 1500 ft, user 2 at 1000 ft)
No Data-Coordination
V ec tored-V DS L
41 Multiuser WondersMultiuser Wonders
How much better can we do?How much better can we do?(ADSL) (ADSL)
0 2000 4000 6000 8000 10000 12000 14000 160000
1
2
3
4
5
6
7x 10
7To
tal r
ate
Length (ft)
ADSL evolution
current cancel self-xtalkcancel all xtalk increase bit cap improve coding
42 Multiuser WondersMultiuser Wonders
Ultimate VDSLUltimate VDSL
0 2000 4000 6000 8000 10000 12000 14000 160000
0.5
1
1.5
2
2.5
3x 10
8
Tota
l rat
e
Length (ft)
VDSL evolution
current cancel self-xtalkcancel all xtalk increase bit cap improve coding ultra-DSL
43 Multiuser WondersMultiuser Wonders
Outline – Part 4Outline – Part 4
DSL and Ethernet ApplicationsDSL and Ethernet Applications Evolution of DSL and DSM vs SMEvolution of DSL and DSM vs SM
UnbundlingUnbundling Towards ethernetTowards ethernet
Multiuser Spectrum BalancingMultiuser Spectrum Balancing Vectored VDSLVectored VDSL Vectored EthernetVectored Ethernet
44 Multiuser WondersMultiuser Wonders
EFM ExamplesEFM Examples
50 100 150 200 250 300 350 400 450 5000
50
100
150
200
250
300
350
Wire Length (m)
Da
ta R
ate
(Mb
ps)
Data Rates achievable over a CAT-5 private network
Vectoring within quad-TDDVectoring within quad Vectoring within bundle
45 Multiuser WondersMultiuser Wonders
EFM RangeEFM Range
2 lines, 100BT at 1 km2 lines, 100BT at 1 km 10BT at 1 km on one line easily10BT at 1 km on one line easily 4 lines, 100BT at 2 km4 lines, 100BT at 2 km
500 1000 15000
20
40
60
80
100
120
140
160
180
200
Wire Length (m)
Da
ta R
ate
(M
bps)
Data Rates achievable over a CAT-5 private network
Vectoring within bundle with 20dBm power Vectoring within bundle with 11.5dBm powerVectoring within quad with 20 dBm power Vectoring within quad with 11.5 dBm power
46 Multiuser WondersMultiuser Wonders
Ethernet ExamplesEthernet Examples
50 100 150 200 250 300 350 400 450 5000
2
4
6
8
10
12
14
16
18
20
Wire Length (m)
Da
ta R
ate
(G
bp
s)Data Rates achievable over a CAT-5 private network
VDSL Quad - 25 MHz Quad - 50 MHz Quad - 100 MHzQuad - 200 MHzQuad - 400 MHz
47 Multiuser WondersMultiuser Wonders
Copper has more bw than fiber?Copper has more bw than fiber?
50 line bundle in last segment of phone network50 line bundle in last segment of phone network 50 lines (200 Mbps/line) = 10 Gbps50 lines (200 Mbps/line) = 10 Gbps FTTH shares 2.5 Gbps among several homes in PON FTTH shares 2.5 Gbps among several homes in PON
architecturearchitecture Get bandwidth up in fiber connections to and Get bandwidth up in fiber connections to and
within network within network Copper in last mile has more BW than system can Copper in last mile has more BW than system can
handlehandle 100BT/100 Mbps to everyone, everywhere a 100BT/100 Mbps to everyone, everywhere a
phone line goes, is possible in the next decade.phone line goes, is possible in the next decade.
48 Multiuser WondersMultiuser Wonders
ConclusionsConclusions
Enormous wireline opportunity for multiuserEnormous wireline opportunity for multiuser Gains may be even larger than for wirelessGains may be even larger than for wireless Relatively stationary environmentRelatively stationary environment
The real broadbandThe real broadband At least 100 BT to everyone anywhere over a At least 100 BT to everyone anywhere over a
twisted pairtwisted pair DataData Voice, voice, voiceVoice, voice, voice VideoVideo
Welcome to the broadband age in this century Welcome to the broadband age in this century
49 Multiuser WondersMultiuser Wonders
Parts 3 and 4: Outline/ScheduleParts 3 and 4: Outline/Schedule
2:00-2:452:00-2:45 MU TheoryMU Theory 2:45-3:302:45-3:30 channels for wirelinechannels for wireline 3:30-4:00 3:30-4:00 CoffeeCoffee 4:00-4:304:00-4:30 DSL and Ethernet archDSL and Ethernet arch 4:30-5:154:30-5:15 Multiuser improvementsMultiuser improvements
50 Multiuser WondersMultiuser Wonders
ReferencesReferences1)1) T. Starr, J. Cioffi, and P. Silverman, T. Starr, J. Cioffi, and P. Silverman, Understanding Digital Subscriber Line TechnologyUnderstanding Digital Subscriber Line Technology, Prentice-Hall, Upper , Prentice-Hall, Upper
Saddle River, NJ, 1999.Saddle River, NJ, 1999.2)2) S. Verdu, S. Verdu, Multiuser DetectionMultiuser Detection, Cambridge Press, UK, 1998., Cambridge Press, UK, 1998.3)3) C. Aldana and J. Cioffi, “Channel Tracking for MISO Systems using EM Algorithm,” ICC 2001, Helsinki, Fin, C. Aldana and J. Cioffi, “Channel Tracking for MISO Systems using EM Algorithm,” ICC 2001, Helsinki, Fin,
http://cafe.stanford.edu/people/cioffi/dsm/channelpap/icc2001.pdfhttp://cafe.stanford.edu/people/cioffi/dsm/channelpap/icc2001.pdf4)4) C. Zeng, C. Aldana, A. Salvekar, and J. Cioffi, “Crosstalk Identification in xDSL Systems,” C. Zeng, C. Aldana, A. Salvekar, and J. Cioffi, “Crosstalk Identification in xDSL Systems,”
http://cafe.stanford.edu/people/cioffi/dsm/channelpap/jsac01.pdfhttp://cafe.stanford.edu/people/cioffi/dsm/channelpap/jsac01.pdf, August 2001 IEEE JSAC., August 2001 IEEE JSAC.5)5) G. Ginis and J.M. Cioffi, "Vectored-DMT: A FEXT CancellingModulation Scheme for Coordinating Users," ICC G. Ginis and J.M. Cioffi, "Vectored-DMT: A FEXT CancellingModulation Scheme for Coordinating Users," ICC
2001, Helsinki, Finland,pp. 305-309.2001, Helsinki, Finland,pp. 305-309.6)6) A. Duel-Hallen, "Equalizers for Multiple Input/Multiple Output Channelsand PAM Systems with A. Duel-Hallen, "Equalizers for Multiple Input/Multiple Output Channelsand PAM Systems with
Cyclostationary Input Sequences," IEEE J. Sel. AreasCommun., vol. 10, no. 3, pp. 630-639, April 1992.Cyclostationary Input Sequences," IEEE J. Sel. AreasCommun., vol. 10, no. 3, pp. 630-639, April 1992.(Generalizes infinite length MMSE-DFE's derived with spectralfactorization to MIMO case.)(Generalizes infinite length MMSE-DFE's derived with spectralfactorization to MIMO case.)
7)7) A. Duel-Hallen, "Decorrelating Decision-Feedback Multiuser Detector forSynchronous Code-Division Multiple-A. Duel-Hallen, "Decorrelating Decision-Feedback Multiuser Detector forSynchronous Code-Division Multiple-Access Channel," IEEE Trans. Commun.,vol. 41, no. 2, pp.285-290, Feb. 1993.(Zero forcing DFE solution.)Access Channel," IEEE Trans. Commun.,vol. 41, no. 2, pp.285-290, Feb. 1993.(Zero forcing DFE solution.)
8)8) J. Yang and S. Roy, ``Joint Transmitter-Receiver Optimization forMulti-Input Multi-Output Systems with J. Yang and S. Roy, ``Joint Transmitter-Receiver Optimization forMulti-Input Multi-Output Systems with Decision Feedback,'' IEEETransactions on Information Theory, vol. 40, no. 5, pp. 1334-1347,September Decision Feedback,'' IEEETransactions on Information Theory, vol. 40, no. 5, pp. 1334-1347,September 1994.(Showed that minimizing the decision feedback error is equivalent toachieving the mutual information.)1994.(Showed that minimizing the decision feedback error is equivalent toachieving the mutual information.)
9)9) G. J. Foschini, G. D. Golden, R. A. Valenzuela and P. W. Wolniansky,``Simplified Processing for High G. J. Foschini, G. D. Golden, R. A. Valenzuela and P. W. Wolniansky,``Simplified Processing for High Spectral Efficiency WirelessCommunication Employing Multi-Element Arrays,'' IEEE Journal onSelected Spectral Efficiency WirelessCommunication Employing Multi-Element Arrays,'' IEEE Journal onSelected Areas in Communications, vol. 17, no. 11, pp. 1841-1852,November 1999.(A zero-forcing GDFE combined Areas in Communications, vol. 17, no. 11, pp. 1841-1852,November 1999.(A zero-forcing GDFE combined with ordering.)with ordering.)
10)10) M. K. Varanasi, ``Decision Feedback Multiuser Detection: A SystematicApproach,'' IEEE Transactions on M. K. Varanasi, ``Decision Feedback Multiuser Detection: A SystematicApproach,'' IEEE Transactions on Information Theory, vol. 45, no. 1,pp. 219-240, January 1999.(Asymptotic analysis (high SNR) of decision Information Theory, vol. 45, no. 1,pp. 219-240, January 1999.(Asymptotic analysis (high SNR) of decision feedback and issues ofordering.feedback and issues ofordering.
51 Multiuser WondersMultiuser Wonders
More ReferencesMore References
11)11) N. Al-Dhahir and A. H. Sayed, ``The Finite-Length Multi-Input Multi-OutputMMSE-DFE,'' N. Al-Dhahir and A. H. Sayed, ``The Finite-Length Multi-Input Multi-OutputMMSE-DFE,'' IEEE Transactions on Signal Processing, vol. 48, no. 10,pp. 2921-2936, October 2000.IEEE Transactions on Signal Processing, vol. 48, no. 10,pp. 2921-2936, October 2000.
12)12) M. L. Honig, P. Crespo, K. Steiglitz, "Suppression of Near- and Far-EndCrosstalk by M. L. Honig, P. Crespo, K. Steiglitz, "Suppression of Near- and Far-EndCrosstalk by Linear Pre- and Post-Filtering," IEEE JSAC, vol. 10, no. 3,April 1992, pp. 614-629.Linear Pre- and Post-Filtering," IEEE JSAC, vol. 10, no. 3,April 1992, pp. 614-629.(MIMO MMSE linear equalizers.)(MIMO MMSE linear equalizers.)
13)13) A. Sendonaris, V. V. Veeravalli, "Joint Signaling Strategies forApproaching the Capacity A. Sendonaris, V. V. Veeravalli, "Joint Signaling Strategies forApproaching the Capacity of Twisted-Pair Channels," IEEE Tran. Commun.,vol. 46, no. 5, May 1998, pp. 673-685.of Twisted-Pair Channels," IEEE Tran. Commun.,vol. 46, no. 5, May 1998, pp. 673-685.
14)14) R.S. Cheng and S. Verdu, "Gaussian multiaccess channels with ISI: Capacityregion and R.S. Cheng and S. Verdu, "Gaussian multiaccess channels with ISI: Capacityregion and multiuser water-filling". IEEE Trans. Info. Th. IT-39, pp773-783, May 1993.multiuser water-filling". IEEE Trans. Info. Th. IT-39, pp773-783, May 1993.
15)15) W. Yu, W.Rhee, S. Boyd, and J. Cioffi, "Iterative Water-filling for Vector Multiple Access W. Yu, W.Rhee, S. Boyd, and J. Cioffi, "Iterative Water-filling for Vector Multiple Access Channel," IEEE International Symposium on InformationTheory 2001.Channel," IEEE International Symposium on InformationTheory 2001.
16)16) W. Yu, G. Ginis,and J. Cioffi, "An Adaptive Multiuser Power Control Algorithm for VDSL," W. Yu, G. Ginis,and J. Cioffi, "An Adaptive Multiuser Power Control Algorithm for VDSL," Submitted to JSAC. Also T1E1.4-2001/200R3Submitted to JSAC. Also T1E1.4-2001/200R3
17)17) W. Yu, G. Ginis, J. Cioffi, “optimum solution of broadcast communications problem,” in W. Yu, G. Ginis, J. Cioffi, “optimum solution of broadcast communications problem,” in preparation, 2001, preparation, 2001, [email protected]@dsl.stanford.edu..
18)18) K. Cheong, J. Choi, and J. Cioffi, “Multiuser Interference Canceler via Iterative Decoding K. Cheong, J. Choi, and J. Cioffi, “Multiuser Interference Canceler via Iterative Decoding for DSL Applications,” IEEE JSAC, Feb 2002, to appear, see also August 1999, ITU for DSL Applications,” IEEE JSAC, Feb 2002, to appear, see also August 1999, ITU contribution, SG15/Q4-NG-085.contribution, SG15/Q4-NG-085.