CooperativeCommunications inWirelessSystems Elza Erkip - - PowerPoint PPT Presentation

cooperative communications in wireless systems
SMART_READER_LITE
LIVE PREVIEW

CooperativeCommunications inWirelessSystems Elza Erkip - - PowerPoint PPT Presentation

CooperativeCommunications inWirelessSystems Elza Erkip WirelessNetworks Ad-hoc Network Backbone Network Wireless Cellular Wireless Cellular WirelessLAN Wireless LAN Outline Relaychannel


slide-1
SLIDE 1

CooperativeCommunications inWirelessSystems

Elza Erkip

slide-2
SLIDE 2

WirelessNetworks

Backbone Network

Wireless LAN Wireless Cellular

WirelessLAN Wireless Cellular Ad-hoc Network

slide-3
SLIDE 3

Outline

  • Relaychannel

– Relayinginwireless:Userelaysfordiversity – Traditionalmultihop versusmultihop withdiversity

  • Conceptofcooperation:“Virtual”antennaarray
  • Informationtheoreticmodelandanalysis
  • Cooperativecodes
slide-4
SLIDE 4

RelayChannel

  • IntroducedbyVen der Meulen (1971)
  • CoverandElGamal (1979):Innerandouterboundsonthe

capacity;capacityofdegradedrelaychannel

  • Multiplerelays:

– Schein &Gallager – Gupta&Kumar – Xie &Kumar – Reznik,Kulkarni &Verdu – Gastpar &Vetterli – Gastpar,Kramer&Gupta

S R D

slide-5
SLIDE 5

Multi-hopWirelessCommunications

Source Relay Destination d d § Directtransmission:

§ Relayedtransmission:

( )

α α α

Pd d P P

total

2 2 = =

α

Pd P

total

2 =

  • Multi-hopreducespowerrequirementsandinterference
  • Currenttrends:

– Onlyconsiderpathloss – Destinationonlyprocessessignalfromrelay

  • Whataboutfading?Usemulti-hopfordiversity.
slide-6
SLIDE 6

DiversityinMultihop

  • Onesource/destination,tworelays
  • Pathlossandfading
  • Uncoded BPSK,bitbybitprocessing

– Relaysamplifyreceivedsignalandforward

  • Direct
  • Viarelays

– Totalpowerdividedequallyamongsourceandrelays

S S S S R1 R2 distance

s

R1 R2 D

time time

slide-7
SLIDE 7

EffectofMultihop Diversity

2 4 6 8 10 12 10

  • 6

10

  • 5

10

  • 4

10

  • 3

10

  • 2

10

  • 1

10 SNR(dB) BER Directtx. Multihoptx. Diversitytx. Tx.div.3

slide-8
SLIDE 8

UserCooperationDiversity

  • Diversitythroughcooperationofmobiles

(Sendonaris,Erkip,Aazhang)

  • Twomainideas

– Userelaystoprovidediversity – Collaborativescheme:Bothmobileshelpeachother (“partners”)

slide-9
SLIDE 9
  • Mobileantennasareomnidirectional

– Signalstransmittedtowardsthedestinationcanbe“overheard” atthepartner – Partnersprocessthisoverheardinformationandre-transmit towardsthedestination

  • Totalresources(power,bandwidth)aresameasnon-cooperativecase

– Destinationprocessessignalsfrombothmobiles

  • Spatialdiversitythroughpartner’santenna,a“virtual”

antennaarray

– Inter-userchannelisnoisy!

UserCooperation

slide-10
SLIDE 10

Approach

  • Informationtheory(Sendonaris,Erkip,Aazhang)

– Ageneralmodelillustratingthatcooperation resultsinthroughputanddiversitygains – Someidealizedassumptions

  • Codedesign(Stefanov,Erkip)

– Morerealisticassumptions – Channelcodesthatexploitusercooperationgains

slide-11
SLIDE 11

InformationTheoreticModel

1

E

2

E

1

X

1

Y

2

W

1

W Z

2

Y

1

Z Y

10

K

12

K

2

Z

21

K

20

K

2

X

User1 User2 D

  • Kij :Fadingamplitudes
  • Receiverstrackthefadingparameters,transmittersdon’t
slide-12
SLIDE 12

CapacityAnalysis

  • Assumptions

– Partnersreceiveandtransmitatthesametime

(sameassumptionasclassicalrelayofCover&ElGamal)

– Transmittershavephaseinformation:Coherent combiningispossible

  • Modelissimilartomultipleaccesschannel

withgeneralizedfeedback

– DMCandGaussian(Carleial,Willems et.al.)

  • Achievableregionunderfading
slide-13
SLIDE 13

TransmitSignalStructure

  • InformationW1 =(W10,W12)
  • SignalX1 =(X10,X12,U1)

– X10:SendW10 atrateR10 todestination(D) – X12:SendW12 atrateR12 tomobile2(alsoheardatD) – U1:Cooperativesignalbasedon(W12,W12)totheBS

  • Powerallocation:P1=P10+P12+PU1
slide-14
SLIDE 14

AchievableRateRegionwithCooperation

} , | ) , {(

21 20 2 12 10 1 2 1

R R R R R R R R + = + =

2 , 1 E

2

=

  • <

k N P K C R

k k k

2 , 1 E

2 ~ ~ 2~ ~

=

  • +

< k N P K P K C R

k k k k k k k k k k

  • +

< + N P K P K C R R

20 2 20 10 2 10 20 10

E

  • +

+ < + + + N P P K K P K P K C R R R R

U U 2 1 20 10 2 2 20 1 2 10 21 12 20 10

2 E

2 21 20 2 1 12 10 1

,

U U

P P P P P P P P + + = + + =

Theclosureof where forsomepowerassignment

slide-15
SLIDE 15

AchievableRateRegion(Symmetric)

slide-16
SLIDE 16

AchievableRateRegion(Asymmetric)

slide-17
SLIDE 17

ProbabilityofOutage

slide-18
SLIDE 18

BenefitsofCooperation

  • Higherdatarates
  • Reducedoutage

Providesanincentiveforrelaying

  • Increasedbatterylife
  • Extendedcoverageforcellular
  • Worksevenwheninter-userchannelisbadorone

userisclosertothebasestation Forbothmobiles

slide-19
SLIDE 19

RealisticConstraints

  • Hardtotransmit/receiveatthesametime

– Withoutcooperation – Withcooperation N N N/2 N/2 N/2 N/2

M1 M2forM1 M2 M1forM2 M1

M2

slide-20
SLIDE 20

CooperationUsingTime-Sharing

(Laneman,Tse,Wornell)

  • Worksforsame/separatereceivers
  • Outageprobabilityanalysisforslowfading
  • Partnercan

– Amplifyandforward

  • Amplifiesnoise,twoleveldiversity

– Decodeandforward

  • Perfectdecodingisastrictconstraint,oneleveldiversity

– Adaptivedecodeandforward

  • Partnerdecodesonlywhenitcan,twoleveldiversity
  • Channelcodesforcooperation?
slide-21
SLIDE 21

ChannelCodingforCooperation

(Stefanov,Erkip)

  • Withcooperation

N/2 N/2 N/2 N/2 M1 M2forM1 M2 M1forM2

  • Tocooperate,M1onlytransmitshalfofitscodedbits.
  • ThisisreceivedatM2andatthedestination.
  • M2decodesM1’sinformation,re-encodesandtransmitsremaininghalf
  • fthecodedbits.
  • M2forwardsonlywhenithascorrectinformation(checkedbyCRC),otherwise

M1continues.

  • ProtocolsimilartoHunter&Nosratinia.
slide-22
SLIDE 22

BlockFadingModel

  • Quasi-staticchannels

Blockfadingwhencooperationtakesplace

  • Canusecodesdesignedforblockfadingchannels

(Knopp &Humblet)

– Maximizediversityandcodingadvantages

  • Additionalconstraints

– Firsthalfofthecodehastobegoodinthequasi-staticinter- userchannel – Codehastobegoodinthequasi-staticchannelwhen cooperationdoesnottakeplace

N/2 N/2 M1 M2forM1

slide-23
SLIDE 23

CooperativeCodingPerformanceAnalysis

  • Frameerrorratewiththeabovecooperationprotocol

– Pf

C :FERforthecooperativeprotocol

– Pf

in :FERforinter-userchannel

– Pf

BF :FERfortheblockfadingchannelresultingfromcooperation

– Pf

QS :FERforthedirect(non-cooperative)channel

– SNRi:ReceivedSNRforuseriatthedestination,i=1,2 – SNRin:ReceivedSNRintheinter-userchannel – K:Codeparameters

1 2 1

) 1 ( SNR K SNR K SNR SNR K P P P P P P P P

QS in in BF QS f in f BF f QS f in f BF f in f C f

+ ≤ + ≤ + − =

slide-24
SLIDE 24

CooperativeCoding(Similarusers)

slide-25
SLIDE 25

CooperativeCoding(Asymmetriccase)

slide-26
SLIDE 26

EffectofCooperationonRouting

– Traditionalrouting

  • Onlypathlossisconsidered:ChooseM3asrelay

– Cooperativediversity

  • Formedium-highSNR,M2resultsinlowererrorrate
  • Goodinter-userchanneldominates

M1 M3 M2 D

slide-27
SLIDE 27

RoutingwithCooperativeDiversity

slide-28
SLIDE 28

OtherRelatedProjectsatPoly

  • Cooperativespace-timecoding(Stefanov,

Erkip)

– Cooperativecodesformobileswithmultipleantennas – Asymmetric:Cooperationamongmobileswithdifferent numberofantennas(overlaycodes)

  • Choiceofpartners:Geometryofcooperation

(Lin,Erkip,Stefanov)

  • Diversityinrelayingprotocols (Yuksel,Erkip)
  • Visitourwebpage:eeweb.poly.edu/~elza
slide-29
SLIDE 29

References

1. A.Sendonaris,E.Erkip andB.Aazhang,“Usercooperationdiversity-partI: Systemdescription.”Toappear,IEEETransactionsonCommunications. 2. A.Sendonaris,E.Erkip andB.Aazhang,“Usercooperationdiversity-partII: Implementationaspectsandperformanceanalysis.”Toappear,IEEE TransactionsonCommunications. 3. A.Stefanov andE.Erkip,“Cooperativecodingforwirelessnetworks.”In ProceedingsofIEEEConferenceonMobileandWirelessCommunications Networks,Stockholm,Sweden,September2002. 4. A.Stefanov andE.Erkip,“Ontheperformanceanalysisofcooperativespace- timecodedsystems.”ToappearIEEEWirelessCommunicationsand NetworkingConference(WCNC2003), NewOrleans,Lousiana,March2003. 5. A.Stefanov andE.Erkip,“Cooperativespace-timecodingforwireless networks.”Toappear,2003InformationTheoryWorkshop,Paris,France,April 2003.