Iterative Frequency-Domain Detection for Single CarrierDetection for Single-Carrier

Modulations

Rui Dinis

1

SummarySummary

1 Context and motivation1. Context and motivation2. OFDM and SC-FDE modulations3 h i3. IB-DFE techniques4. Applications

– MIMO systems– Network diversity schemesy– BS cooperation

5 Conclusions

2

5. Conclusions

Equalization TechniquesEqualization Techniques• Frequency selective channels (time-dispersive)q y ( p )

– Inter-Symbol Interference (ISI)– Performance degradation

• MLSE receiver– Optimum receiver– Efficient implementation: Viterbi equalizer – Exponential complexity

C i l li• Conventional equalizers– Time-domain implementation

C l it li ith L

3

– Complexity linear with LC

– Suitable for moderately dispersive channels

Block Transmission TechniquesBlock Transmission Techniques• Block transmission techniquesq

– Frequency-domain receiver implementation– Low complexity FFT-based receivers – Suitable for severely time-dispersive channels

• OFDM techniques– Multicarrier modulation– CP-assisted block transmission

Hi h i h i ti i t– Higher carrier synchronization requirements– High envelope fluctuations

=> Amplification difficulties 2

3

4

=> Amplification difficulties

0 0.2 0.4 0.6 0.8 10

1

2

t/T

|s(t)|

SC-FDE Modulations

• CP-assisted block transmissionCP assisted block transmission• Lower envelope fluctuations• FDE structure• FDE structure

kS kY nyDFT X IDFT

ns

1/ ZFH

kF

• FDE coefficients *

2

1/ , ZF

, MMSE/

k

kk

HHF

5

21/ kSNR H

SC-FDE vs OFDM

kS ns ny kY kS ˆkS

ns ny kY kS ˆns ns

• Similar overall complexitySC FDE ith i l t itt > li k• SC-FDE with simpler transmitter => uplink

• OFDM with simpler receiver => downlink=> Hybrid OFDM/SC-FDE systems (e.g., LTE)

6

y y ( g , )

SC-FDE Performance

MFB (Matched Filter Bound):MFB (Matched Filter Bound):

1 2

, 00

12k

Nbb MFB kH k

EP E Q HN N

10-1

100

_ _ : OFDM____ : SC-FDE (ZF)

10-2

10

BER

__o__ : SC-FDE (MMSE) : MFB

10-3

B

70 5 10 15 2010

-4

Eb/N0(dB)

Performance Constraints f SC FDEfor SC-FDE

• Linear FDE => Noise enhancement, residual ISILinear FDE Noise enhancement, residual ISI• Better performance with nonlinear equalizers • Conventional DFE• Conventional DFE

– Good performance/complexity tradeoffs– Error propagationp p g

y sny

nsns

8Can we make a DFE for SC-FDE?

IB-DFE Designg kS kY ny ˆns ns

kF ˆkS ˆns

• Iterative FDEO ti i ffi i t

kB

• Optimum receiver coefficients

*

221

1/ 1k

k k k kHF B F H

SNR H

with selected to ensure 110

1Nk kN k

F H

* 2

21/ 1 kSNR H

and9

* 2ˆ / | | 1 2 (for QPSK)n n n bE s s E s P

IB-DFE Performance

100

10-1

____: IB-DFE : MFB

10-2

BER

(+): Iter. 1(*): Iter. 2(): Iter. 3( ) It 4

10-3

(o): Iter. 4

0 2 4 6 8 10 12 14 1610

-4

10BER for iterations 1 (linear FDE), 2, 3 and 4Eb/N0(dB)

IB-DFE for General C t ll tiConstellations

kF

kYX IDFT

kS

+ -

s

ns

kS

LLRcomput.

( )mn

Dec.

A ’

• Required operations

x kB DFT ns k Average’s

comput.

– Computation of bits’ LLRs– Computation of average bit valuesp g– Computation of average symbol values– Computations of block reliabilities

11

p

Generalized Mapping Rulepp g

• Constellation symbols expressed as function of theConstellation symbols expressed as function of the corresponding bits as:

1

)(1

)3()2()1()2()1( ,bgbgbbgbgbggs mM

im

.oftionrepresentabinarytheis,,..,,

,...

,0,1,2,1

1043210

i

bgbgbbgbgbggs

iiii

mn

iinnnnnn

• The log-likehood ratio of the mth bit of the nth transmitted symbol is by:

s ed sy bo s by:

,

exp

explog~0Pr

~1Prlog

2

)(1

2

2

~

2

~

)(

)()(

n

mn

ss

Sss

m

nm

nmn s

s

12

exp0Pr)(

022

m

n

Snn s

ly.respective,1or0whereofsubsetsareand )(1

)(0 n

mm

Computation of IB-DFE P tParameters

1 1

)()(

)(

output FDE thetodconditionebit values AverageMm

i

0 0

)()( ,

2tanh

i m

mnin

nmn

imbgsb

estimates theofy reliabilitBlock 1 1

)(2M

m im

2tanhwhere,

ˆ )()(

12

0 0

)(

2

*,

mnm

nMi m

mni

n

nn

g

g

sEssE

im

13

0i

ig

Turbo FDE

ns ns

ns ns

14

BER Performance

100

10-1

- - - : Turbo FDE____: IB-DFE (soft) : MFB

10-2

BER

(+): Iter. 1(*): Iter. 2(o): Iter. 4

10-3

-3 -2 -1 0 1 2 3 4 5 6 710

-4

15Coded BER performance for IB-DFE and Turbo FDEEb/N0(dB)

MIMO SystemsMIMO Systems

• P transmitters and RP multiple receiversP transmitters and RP multiple receivers• rth Received signal

( ) ( ) ( )P

r eq r r TY S H N Y H S N

• Detection of a given signal

, ,1

k k p k p k k k k kp

Y S H N

Y H S N

T TS F Y B SDetection of a given signal• Optimum receiver coefficients

, , ,k p k p k k p kS F Y B S

12

2, 2

H HNk p k p k R k

S

F H I P H I H

16, ,k p k k p p B H F e

1 2diag( , ,..., )P P

Network Diversity Techniquesy q

• Noise and/or collisionsNoise and/or collisions => Corrupted packets

C i l h• Conventional schemes– Discard corrupted packets + retransmissions

=> Poor performance• Packet Combining (PC) and Multi-Packet g ( )

Detection (MPD)– Take advantage of information in corrupted packets

17

g p p– Improved performance with retransmissions

Packet Combining (Diversity C bi i )Combining)

• Type-II H-ARQ schemeType II H ARQ scheme• Code-combining based on repetition codes

=> Reduced complexity=> Reduced complexity

SC-FDE with Diversity CombiningCombining

Receiver structure

19

IB-DFE Optimization

• Feedforward coefficients:

p

Feedforward coefficients:

( )*

( )22 ( )1/ 1

rr k

k R r

HFSNR H

• selected to have

11/ 1 kr

SNR H

1N L1 ( ) ( )10 1

1N L r rk kN k r

F H

• Feedback coefficients:

( ) ( )1

1R r rk k kB F H

20

1k k kr

Coping with Fixed Channelsp g

• Ideal conditions: uncorrelated channels (UC)Ideal conditions: uncorrelated channels (UC)• Coping with fixed channels: shifted packet (SP)

2.5

3

N

1.5

2|H

k|

0 5

1

210 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 10

0.5

k/N

PER Performance

22Linear FDE with PC (1, 2, 3, 4 retransmissions)

PER Performance

23IB-DFE with PC (4 iterations; 1, 2, 3, 4 retransmissions)

Gooput Performancep

24Rmax=5 retransmissions and saturated load

NDMA (Network Diversity Multiple Access)Multiple Access)

llCollisions

25

IB-DFE-Based Receiver for NDMA

26

PER Performance

27Turbo IB-DFE (4 iterations)

Q Simultaneous UsersQ

28

Coping with Q>Qmaxp g Q Qmax

• Unsolvable collisionsUnsolvable collisions when Q>Qmax

• Backoff techniques to qreduce the number of transmitters

• Constant backoff (CB)pk = p

• Exponential backoff (EB)pk = pk+1

29

• Optimum p* Saturated throughput with J=8 users and Qmax=4

Throughput for Qmax=1, 2 4 and Jg p Qmax ,

30Eb/N0=11dB Eb/N0=1dB

Base Station CooperationBase Station Cooperation

• Point-to-point linksPoint to point links– Capacity: Shannon limit– Performance close to capacityPerformance close to capacity

• Cellular system (multipoint-to-multipoint)– Different frequencies for different cells– Different frequencies for different cells– Poor spectral efficiency

• Network level techniques to increase spectral• Network-level techniques to increase spectral efficiency

• BS (Base Station) cooperation• BS (Base Station) cooperation31

BS CooperationBS Cooperation• Same frequency for MTsq y

in adjacent cells• Inherent macro-diversity

nature• Muliuser scenario • IB-DFE-based receiver

as in MIMO systems

32

BER PerformanceBER Performance

Macro-Diversity Interference scenario without BS cooperation (P=2, R=1) 33

BER Performance with P=R=2BER Performance with P R 2

Equal-power users Low and high power users 34

Interference AlignementInterference AlignementAnt. 1 Ant. 1

Ant. MBS 1...1,lW

.

.

.

Ant. MUT 1

1,lG

1,1,lH

2,1,lH

1K lHAnt. 1

Ant. MBS 2..

Ant. 1

.

.Ant. M

UT 2

G

,1,K l

.2,lW

Ant 1

.

.

. 2,lG

Ant 1

.

.2 K lH

1, ,K lH

Ant. 1

Ant. MBS K...

,K lW

. Ant. 1

.

.Ant. M

UT K

,K lG

.

, ,K K lH

2, ,K lH

K-user MIMO interference channel 35

, .

MC-CDMA with Interference AlignementAlignement

• Transmitter Spreading. IA

PrecoIFFT

+Transmitter structure for a generic user with

s/p

.

.

.

Spreading(L)

.

.

.

Precoding

.

.

.

+CP

.

.

.

Antenna 1

IA s/p Spreading(L)

.

.

.IA

Precoding

CSI from

IFFT+

CP Antenna M

• Receiver structure

Other Tx

Antenna 1 FFT-

CP

.

.

.

Antenna M FFT

.

.

.

Equalizer+

Noise Whitening

.

.

.

.

.

.IB-DFE

36

Antenna M-

CP

.

.

CSI Eq. Channels

Receiver Structure

• IB-DFE (PIC approach) applied to the receivedIB DFE (PIC approach) applied to the received signal after noise whitening process

Despreading

x+

-

Despreading(L) Demod.

S/P...

Despreading Demod

M dSpreading

(L) Demod.

xMod.p g

(L)

Mod.Spreading(L)

P/S...

37

(L)

BER Performance for K=3, M=4BER Performance for K 3, M 4

10-1 10-1

4x4 IA ZF4x4 IA MMSE

-3

10-2

ER

4 4 IA ZF 3

10-2

ER

4x4 IA MMSE+IB-DFE 1 iter4x4 IA MMSE+IB-DFE 2 iter4x4 IA MMSE+IB-DFE 3 iter4x4 IA MMSE+IB-DFE 4 iterMF L=128

10-4

10 3

BE 4x4 IA ZF

4x4 IA MMSE4x4 IA ZF+IB-DFE 1 iter4x4 IA ZF+IB-DFE 2 iter4x4 IA ZF+IB-DFE 3 iter

10-4

10-3

BE

-2 0 2 4 6 8 10 12 1410

-5

10

4x4 IA ZF+IB-DFE 4 iterMF L=128

2 0 2 4 6 8 10 12 1410-5

10

ZF IA MMSE IA 38

2 0 2 4 6 8 10 12 14Eb/No (dB)

-2 0 2 4 6 8 10 12 14Eb/No (dB)

Conclusions

• IB-DFE receivers for SC-FDEIB-DFE receivers for SC-FDE• Iterative receivers implemented in the

frequency domainfrequency domain • Extension for network diversity (DC and/or

NDMA) BS iNDMA), BS cooperation, etc. • BER performance close to the MFB• Suitable for broadband wireless systems

39

References

• General IB-DFE issues– A.Gusmão, P.Torres, R.Dinis and N.Esteves, “A Turbo FDE Technique for Reduced-CP SC-Based Block

Transmission Systems”, IEEE Trans. on Comm., Vol. 55, No. 1, pp. 16-20, Jan. 2007.– A.Gusmão, P.Torres, R.Dinis and N.Esteves, “A Reduced-CP Approach to SC/FDE Block Transmission for

Broadband Wireless Communications”, IEEE Trans. on Comm., Vol. 55, No. 4, pp. 801-809, April 2007.– R.Dinis, P.Silva and A.Gusmão, “IB-DFE Receivers with Space Diversity for CP-Assisted DS-CDMA and

MC-CDMA Systems”, European Trans. on Telecomm., Vol. 18, No. 7, pp. 791-802, Nov. 2007.– N.Benvenuto, R.Dinis, D.Falconer and S.Tomasin, "Single Carrier Modulation with Non Linear Frequency

Domain Equalization: An Idea Whose Time Has Come – Again", IEEE Proceedings, Vol. 98, No. 1, page 69-96 Jan 201096, Jan. 2010.

– F. Silva, R. Dinis and P. Montezuma, “Estimation of the Feedback Reliability for IB-DFE Receivers”, ISRNCommunications and Networking Journal, 2011.

– F. Amaral, R. Dinis, P. Montezuma, N. Souto, “Approaching the MFB with Block Transmission Techniques”, European Trans. on Telecommunications, Vol. 23, No. 1, pp. 76 - 86, February, 2012.European Trans. on Telecommunications, Vol. 23, No. 1, pp. 76 86, February, 2012.

– J. Silva, R. Dinis, N. Souto, P. Montezuma, “Single-carrier frequency domain equalisation with hierarchical constellations: an efficient transmission technique for broadcast and multicast systems”, IET Communications, Vol. 6, No. 13, pp. 2065-2073, Sep. 2012

40

References

• MIMO systems– R.Dinis, R.Kalbasi, D.Falconer and A.Banihashemi, "Iterative Layered Space-Time Receivers

for Single-Carrier Transmission over Severe Time-Dispersive Channels", IEEE Comm. Letters, Set. 2004.

– P Silva and R Dinis “Multiuser Detection for the Uplink of Prefix-Assisted DS-CDMA SystemsP.Silva and R.Dinis, Multiuser Detection for the Uplink of Prefix-Assisted DS-CDMA Systems Employing Multiple transmit and Receive Antennas”, IEEE VTC’06(Fall), Montreal, Canada, Sep. 2006.

– P.Silva and R.Dinis, “Frequency-Domain Multiuser Detection for CP-Assisted DS-CDMA Signals” IEEE VTC’06 (Spring) Melbourne Australia May 2006Signals , IEEE VTC 06 (Spring), Melbourne, Australia, May 2006.

– R.Kalbasi, D.Falconer, A.Banihashemi and R.Dinis, "A Comparison of Frequency Domain Block MIMO Transmission Systems", IEEE Trans. on Vehicular Technology, Vol. 58, No. 1, pp. 165-175, Jan. 2009.

41

References

• Network Diversity– R.Dinis, P.Carvalho, L.Bernardo, R.Oliveira, M.Pereira and P.Pinto, “Frequency-Domain Multipacket Detection: A High

Throughput Technique for SC-FDE Systems”, IEEE Trans. on Wireless Communications, Vol. 8, No. 7, pp. 3798-3807, Jul. 2009.

– F. Ganhão, M. Pereira, L. Bernardo, R. Dinis, R. Oliveira, P. Pinto, “Performance of Hybrid ARQ for NDMA Access Schemes with Uniform Average Power Control”, Journal of Communications, Vol. 6, No. 9, pp. 691-699, Academy Publishers, Dec. 20112011.

– M. Pereira, L. Bernardo, R. Dinis, R. Oliveira, P. Montezuma, P. Pinto, “Performance of diversity combining ARQ error controlin a TDMA SC-FDE system”, IEEE Trans. on Communications, Vol. 60, No. 3, pp. 735-746, Mar. 2012.

– F. Ganhão, R. Dinis, L. Bernardo and R. Oliveira , “Analytical BER and PER Performance of Frequency-Domain Diversity Combining, Multipacket Detection and Hybrid Schemes”, IEEE Transactions on Communications, Vol. 60, No. 8, pp. 2353-2362 A 20122362, Aug. 2012.

– M. Silva, R. Dinis, P. Montezuma, “Iterative Frequency-Domain Packet Combining Techniques for UWB Systems with Strong Interference Levels”, Wireless Personal Communications, Vol. 9, No. 9, July, 2012.

– M. Pereira, L. Bernardo, R. Dinis, R. Oliveira, P. Pinto, “Frequency-Domain Cross-Layer Diversity Techniques: An Efficient Way of Coping with Lost Packets in Broadband Wireless Systems”, IEEE Wireless Communications Magazine, 2013.

– M. Pereira, L. Bernardo, R. Dinis, R. Oliveira, P. Pinto, “On the Use of Frequency-Domain Cross-Layer Diversity Techniques to Cope with Lost Packets”, Physical Communication”, Elsevier Physical Communication.

– F. Ganhão, M. Pereira, L. Bernardo, R. Dinis, R. Oliveira, P. Pinto, “Performance Analysis of an Hybrid ARQ Adaptation of NDMA Schemes”, IEEE Transactions on Communications, 2013.

42

References

• Joint detection/estimation– C.Lam, D.Falconer, F.Danilo-Lemoine and R.Dinis, “Channel Estimation for SC-FDE Systems Using

Frequency Domain Multiplexed Pilots”, IEEE VTC’06(Fall), Montreal, Canada, Sep. 2006.– R.Dinis, C.Lam and D.Falconer, "Joint Frequency-Domain Equalization and Channel Estimation using

Superimposed Pilots", IEEE WCNC'08.– R.Dinis, T.Araújo, P.Pedrosa and F.Nunes, "Joint Turbo Equalization and Carrier Synchronization for SC-

FDE Schemes", European Trans. on Telecomm., Vol. 21, No. 2, pp.131-141, Mar. 2010.– P. Pedrosa, R. Dinis and F. Nunes, “Iterative Frequency Domain Equalization and Carrier Synchronization for

Multi-resolution Constellations”, IEEE Trans. on Broadcasting, Vol.56, No.4, pp. 551-557, Dec. 2010.P P d R Di i d F N “A l ti l P f E l ti f Cl f R i ith J i t– P. Pedrosa, R. Dinis and F. Nunes, “Analytical Performance Evaluation of a Class of Receivers with Joint Equalization and Residual CFO Estimation”, accepted for publication at Transactions on Emerging Telecommunications Technologies.

– P. Pedrosa, R. Dinis, F. Nunes, J. Bioucas-Dias, “Phase Drift Estimation and Symbol Detection in Digital Communications: A Stochastic Recursive Filtering Approach”, IEEE Communications Letters, Vol. 16, No. 6,Communications: A Stochastic Recursive Filtering Approach , IEEE Communications Letters, Vol. 16, No. 6, pp. 854-857, June 2012.

– F. Silva, R. Dinis, P. Montezuma, “Channel Estimation and Equalization for Asynchronous Single Frequency Networks”, IEEE Transactions on Broadcasting, 2013.

43

References

• Offset modulations– J. Luzio, R. Dinis, P. Montezuma, “SC-FDE for Offset Modulations: An Efficient Transmission

Technique for Broadband Wireless Systems”, IEEE Trans. on Communications, Vol. 60, No. 7, pp. 1851-1861, July 2012.

– M Luzio R Dinis P Montezuma “Pragmatic Frequency Domain Equalization for SingleM. Luzio, R. Dinis, P. Montezuma, Pragmatic Frequency Domain Equalization for Single Carrier with Offset Modulations”, IEEE Transactions on Wireless Communications, 2013.

• NL channels– P.Silva and R.Dinis, "A Turbo SDMA Receiver for Strongly Nonlinearly Distorted MC-CDMA g y y

Signals", Canadian Journal of Electrical and Computer Engineering, 2008.– P.Silva and R.Dinis, “Joint Turbo Equalization and Multiuser Detection of MC-CDMA Signals

With Strongly Nonlinear Transmitters”, IEEE Trans. on Vehicular Technology, Vol. 58, No. 5, pp. 2288-2298, Jun. 2009.pp. 2288 2298, Jun. 2009.

– R.Dinis and P.Silva, "Iterative Detection of Multicode DS-CDMA Signals with Strong Nonlinear Distortion Effects", IEEE Tran. on Vehicular Technology, Vol. 58, No. 8, pp. 4169-4181, Oct. 2009.R Di i P Sil d T A új “T b E li ti ith C l ti f N li Di t ti f

44

– R.Dinis, P.Silva and T.Araújo, “Turbo Equalization with Cancelation of Nonlinear Distortion for CP-Assisted and Zero-Padded MC-CDM Schemes”, IEEE Trans. on Communications, Vol. 57, No. 8, pp. 2185-2189, Aug. 2009.