FLOWS 1ST-2001-32125 Deliverable No:D14 Figure 4.22 Distribution of the MTs in the considered mobile radio cell,K=2...... .127 F9ee4aCiembstomt和netomProacRna)oicRof425句orCenano1L30 and enario2（- Figure 5.1 Model of multi-user MIMO downlink.. 137 Figure 5.2 Constitution of the K MT specific demodulator matrices D(based on the MT specific demodulator matrices D for the single symbol/single MT case.. 143 Figure 5.3 CDF of the average energye iciency of(5.32)for MTs with MIMO model 1 Figure 5.4 CDF of the average energy efficiency of(5.32)for MTs with MIMO model 1... …14 Figure 5.5 CDF of the average energy efficiency of (5.32)for MTs with MIMO model 1..... .147 Figure 5.6 CDF of the average energy efficiency of(5.32)for MTs with MIMO model 2.. …148 Figure 5.7 CDF of the average energy efficiency of(5.32)for MTs with MIMO model 2..... .…149 Figure 5.8 CDF of the average eneray efficiency of(5.32)for MTs with MIMO model 2...... ,.150 Figure 5.9 Magnitudes of the elements of the dominant eigenvector... .153 Figure 5.10 Arguments of the elements of the dominant eigenvector.. .154 Fiaure 5 11 Cross-correlation coefficient of (5 45)versus 154 Figure 5.12 CDF of the cross-correlation coefficients for two and four tap channel. 155 re 513 Cross-co relation coefficient of (5.64)versus 159 of the ratio of the second large Ithe largest eigenvalue for 3GPP cha 161 Figure 5.15 Flow chart of the cost efficient eigenvector selection algorithm.. .164 Figure 5.16 CDE of the average eneray efficiency of (5 32)for MTs. .165 Fiaure 5 17 CDE of the average eneray efficieney of (5 32)for MTs 166 Fiaure 5 18 cDE of the average eneray efficiency of (5 32)for Mts 167 re 5 19 CDE of the av age enet ergy efficiency of (5.32)for MTs. 46R 171 re 6-2:Simplified OFDM S 173 e2.c. rrier MIMO in OFDM. Figure 6-4. FDM t on system coder for genera spa equency cod Figure 6-7:Multichannel syst m equivalent to SVD based MIMO-OFDM 181 Figure 6-8:Block diagram of MIMO-OFDM transmitter with adaptive modulation......... .18 Figure 6-9:Antenna selection.. 183 num antennas. 185 tensl 186 Fia Figure 6-14:Space-time code with 2transmit antenna198 Figure 6-15:Spatial 31December 2003 Page 12
FLOWS IST-2001-32125 Deliverable No: D14 31st December 2003 Page 12 Figure 4.22 Distribution of the MTs in the considered mobile radio cell, K = 2 .........................127 Figure 4.23 Cumulative distribution function Prob(CR ) of CR of (4.255) for scenario 1 ( – ) and scenario 2 ( – – ) ..................................................................................................................130 Figure 4.24 Cumulative distribution function Prob(CR ) of CR of (4.255) for scenario 1 ( – ) and scenario 4 ( – – ) ..................................................................................................................133 Figure 5.1 Model of multi-user MIMO downlink ..............................................................................137 Figure 5.2 Constitution of the K MT specific demodulator matrices ( ) k D based on the MT specific demodulator matrices ( ) 0 k D for the single symbol/single MT case........................143 Figure 5.3 CDF of the average energy efficiency of (5.32) for MTs with MIMO model 1.............145 Figure 5.4 CDF of the average energy efficiency of (5.32) for MTs with MIMO model 1.............146 Figure 5.5 CDF of the average energy efficiency of (5.32) for MTs with MIMO model 1.............147 Figure 5.6 CDF of the average energy efficiency of (5.32) for MTs with MIMO model 2.............148 Figure 5.7 CDF of the average energy efficiency of (5.32) for MTs with MIMO model 2.............149 Figure 5.8 CDF of the average energy efficiency of (5.32) for MTs with MIMO model 2.............150 Figure 5.9 Magnitudes of the elements of the dominant eigenvector..........................................153 Figure 5.10 Arguments of the elements of the dominant eigenvector.........................................154 Figure 5.11 Cross-correlation coefficient of (5.45) versus ............................................................154 Figure 5.12 CDF of the cross-correlation coefficients for two and four tap channel .................155 Figure 5.13 Cross-correlation coefficient of (5.64) versus ............................................................159 Figure 5.14 CDF of the ratio of the second largest and the largest eigenvalue for 3GPP channel model case 3 ...............................................................................................................................161 Figure 5.15 Flow chart of the cost efficient eigenvector selection algorithm .............................164 Figure 5.16 CDF of the average energy efficiency of (5.32) for MTs.............................................165 Figure 5.17 CDF of the average energy efficiency of (5.32) for MTs.............................................166 Figure 5.18 CDF of the average energy efficiency of (5.32) for MTs.............................................167 Figure 5.19 CDF of the average energy efficiency of (5.32) for MTs.............................................168 Figure 6-1: Basic OFDM System.......................................................................................................171 Figure 6-2: Simplified OFDM System...............................................................................................173 Figure 6-3: Subcarrier-specific MIMO in OFDM ..............................................................................174 Figure 6-4: Space-Frequency Coding ..............................................................................................177 Figure 6-5: Coded MIMO-OFDM transmission system...................................................................178 Figure 6-6: MIMO-OFDM encoder for general space-time-frequency code .................................178 Figure 6-7: Multichannel system equivalent to SVD based MIMO-OFDM ....................................181 Figure 6-8: Block diagram of MIMO-OFDM transmitter with adaptive modulation .....................182 Figure 6-9: Antenna selection...........................................................................................................183 Figure 6-10: Mean SNR gain compared to a one-antenna system; System with MRC diversity at Rx, selection diversity at Tx and different numbers of Tx antennas ....................................184 Figure 6-11: Bit Error Rate (analytically) for different numbers of Rx antennas and Tx antennas. Red, green, yellow, blue curves have 1, 2, 3, and 4 Rx antennas, each with 1, 2, or 3 Tx antennas. .....................................................................................................................................185 Figure 6-12: SNR gain at BER 10-2 relative to single antenna transmission; System with MRC diversity at Rx, antenna selection at Tx, uncoded BPSK, and different numbers of Tx antennas ......................................................................................................................................186 Figure 6-13: SNR gain at BER 10-2 relative to one transmit antenna; System with MRC diversity at Rx, antenna selection at Tx, uncoded BPSK, and different numbers of Tx antennas....186 Figure 6-14: Space-time code with 2 transmit antenna .................................................................198 Figure 6-15: Spatial multiplexing......................................................................................................199
FLOWS IST-2001-32125 Deliverable No:D14 Figure 6-16:Comparison of SM decoding techniques 202 Figure 6-17:Differential space-time block code... ...203 Figure 6-18 MIMO-OFDM simulation result for different MIMO techniques with the same band efficiency Figure 7-18:RAKE receiver for WCDMA downlink 23 Figure 7-19:Chip equalisation for WCDMA downlink 231 Figure 7-20:Power delay profile for ITU Vehicular channel A. 235 power a Perte 236 ure 7-22:BER performance of SISO optimal chip-equalised WCDMA downlink over quasi- 238 Figu variable number of active users,K.N,=8.Perfect channel information 238 Figure 7-24:Alamouti two-transmitter space-time block code. 240 Figure 7-25:STBC with Rake receiver for WCDMA downlink.... ..243 Figure 7-26:STBC with chip equalisation for WCDMA downlink..... 244 Fiaure 7.27.Channel matrices for 2x2 MIMO transmission .245 Fig e 7-28:BER D qualiser 248 rmance of chip equalis ct cand SI eq atic Rayleigh channel with ITU Veh.A power delay profile. ,249 Fig re 7-30:BER performance of 2x2 MIMO chip equalise and 2x2 MIMO Rake receiver for ITU vand re 7-31:BER delay profile 251 Figure 7-32:BER perfc mpu ayleigh chann .252 WCDMA downlink for fully loaded ay pr 253 r WC MA T Veh.A power delay profile.. .253 Fig re 7-35:BER performance of 2x2 MIMO chip equaliser WCDMA downlink for fully loaded cell with chann el es ength,N,in chips.Quasi-static delay pr 267 Figure 7- -eq Figure 7- a UMTS Figure 7-38 rame structure for UMTS 360 Figu 39 rm chipe 2560 Rayleigh channel with ITU Veh.A power delay profile.. ,.261 31"December 2003 Page 13
FLOWS IST-2001-32125 Deliverable No: D14 31st December 2003 Page 13 Figure 6-16: Comparison of SM decoding techniques ..................................................................202 Figure 6-17: Differential space-time block code.............................................................................203 Figure 6-18 MIMO-OFDM simulation result for different MIMO techniques with the same band efficiency .....................................................................................................................................206 Figure 7-18: RAKE receiver for WCDMA downlink.........................................................................230 Figure 7-19: Chip equalisation for WCDMA downlink....................................................................231 Figure 7-20: Power delay profile for ITU Vehicular channel A ......................................................235 Figure 7-21: BER performance of Rake receiver WCDMA downlink over quasi-static Rayleigh fading channel with ITU Vehicular A power delay profile. Perfect channel information. Nsf=8. ............................................................................................................................................236 Figure 7-22: BER performance of SISO optimal chip-equalised WCDMA downlink over quasistatic Rayleigh fading channel with ITU Vehicular A power delay profile, with variable number of active users, K. Nsf=8. Perfect channel information............................................238 Figure 7-23: BER performance of SISO sub-optimal chip-equalised WCDMA downlink over quasi-static Rayleigh fading channel with ITU Vehicular A power delay profile, with variable number of active users, K. Nsf=8. Perfect channel information. ............................238 Figure 7-24: Alamouti two-transmitter space-time block code.....................................................240 Figure 7-25: STBC with Rake receiver for WCDMA downlink.......................................................243 Figure 7-26: STBC with chip equalisation for WCDMA downlink.................................................244 Figure 7-27: Channel matrices for 2×2 MIMO transmission ..........................................................245 Figure 7-28: BER performance of SISO (1×1), MISO (2×1) and MIMO (2×2) chip equaliser WCDMA downlink, single user, perfect channel knowledge. Frequency-flat quasi-static Rayleigh channel. .......................................................................................................................248 Figure 7-29: BER performance of 2×2 MIMO chip equaliser and SISO chip equaliser WCDMA downlink for single user and fully loaded cell, perfect channel knowledge. Quasi-static Rayleigh channel with ITU Veh. A power delay profile...........................................................249 Figure 7-30: BER performance of 2×2 MIMO chip equaliser and 2×2 MIMO Rake receiver WCDMA downlink for single user and fully loaded cell, perfect channel knowledge. Quasistatic Rayleigh channel with ITU Veh. A power delay profile. ...............................................249 Figure 7-31: BER performance of 2×2 MIMO Rake receiver WCDMA downlink for single user, versus channel impulse SNR (dB). Quasi-static Rayleigh channel with ITU Veh. A power delay profile.................................................................................................................................251 Figure 7-32: BER performance of 2×2 MIMO chip equaliser WCDMA downlink for single user, versus channel impulse SNR (dB). Quasi-static Rayleigh channel with ITU Veh. A power delay profile.................................................................................................................................252 Figure 7-33: BER performance of 2×2 MIMO Rake receiver WCDMA downlink for fully loaded cell, versus channel impulse SNR (dB). Quasi-static Rayleigh channel with ITU Veh. A power delay profile.....................................................................................................................253 Figure 7-34: BER performance of 2×2 MIMO chip equaliser WCDMA downlink for fully-loaded cell, versus channel impulse SNR (dB). Quasi-static Rayleigh channel with ITU Veh. A power delay profile.....................................................................................................................253 Figure 7-35: BER performance of 2×2 MIMO chip equaliser WCDMA downlink for fully loaded cell with channel estimation, versus pilot sequence length, Nu, in chips. Quasi-static Rayleigh channel with ITU Veh. A power delay profile...........................................................257 Figure 7-36: Chip-equalised UMTS downlink..................................................................................259 Figure 7-37: Transmitted WCDMA signal for UMTS .......................................................................259 Figure 7-38: Frame structure for UMTS downlink ..........................................................................260 Figure 7-39: BER performance of uncoded SISO chip equalised UMTS downlink with common pilot and ‘ideal’ channel estimation. Pilot sequence length=2560 chips. Quasi-static Rayleigh channel with ITU Veh. A power delay profile...........................................................261
FLOWS 1ST-2001-32125 Deliverable No:D14 Figure 7-40:BER per e of uncoded H(50Km/)(100m/h)370H(20Km/).common estimation. …262 d UMTS downlink with fast km/h),92.5Hz (50km/h),185Hz(100km/h),370Hz(200km/h).Common pilot-aided channel 263 Figua7BERerTance6awaastbaS9tgdyi0oRaa88tervMT7Rgwea 0) ppler frequencies OHz(0 km/h),92.5Hz(50km/h),185Hz(100km/h),370Hz ommon pilot-aid 263 Figu =15 N rate1/3 turb ed SISO chip eq JMTS downlink in ppler frequencies 0Hz (0 km/h),92.5Hz(50km/h),85H(100km/h).370H (200kmh. on pilot-aide 264 Figu ation in MIMO channels with Hadamard(orthogonal) 275 FigugigRnagnepiotlengths16 82 The CRLB fo el estimation in MImo char els with random binary pilot 275 FougagSRl8erGgetmaioanwochanaswhramcomcausanpi 276 9oth8nge2 ation in MIMO channels with Hadamard (orthogonal) 276 ation in MIMO channels nath is 256 stin with random binary pilot 277 Figure 8.6 The cRLB for channel estimation in mImo channels with random Gaussian pilot signals;the pilot length is 256 277 Figur MIMO cha nels with Nr receive antennas and Hadamard 280 Fig Fig re 8.9 CRLB for deterministic MIMO channels with SN_r$receive antennas and random Gaussian pilot sequences vsnumber of transmit antennas 283 or all real binary of lengt 4 symb 284 re 8.11 Lo nd un r bou Fig a upp unas I binary in de CR antennas:SNR=15 dB. .285 of length 6 symbols in deterministic MIS 286 g ngymbol tanrat n r bounds of the crlb of fre estimation for optimal binav Figu 15 MCRLB ACRLE Fig e 8.16 MCRLB.ACRLB.and c ntional CRLB for MISO fading channels with random binary pilot sequences vsnumber of transmit antennas.. 295 31December 2003 Page 14
FLOWS IST-2001-32125 Deliverable No: D14 31st December 2003 Page 14 Figure 7-40: BER performance of uncoded SISO Rake receiver UMTS downlink with fast fading Rayleigh channel, ITU Veh. A power delay profile. Doppler frequencies 0Hz (0 km/h), 92.5Hz (50km/h), 185Hz (100km/h), 370Hz (200km/h). Common pilot-aided channel estimation....................................................................................................................................262 Figure 7-41: BER performance of uncoded SISO chip equalised UMTS downlink with fast fading Rayleigh channel, ITU Veh. A power delay profile. Doppler frequencies 0Hz (0 km/h), 92.5Hz (50km/h), 185Hz (100km/h), 370Hz (200km/h). Common pilot-aided channel estimation....................................................................................................................................263 Figure 7-42: BER performance of rate 1/3 turbo coded SISO Rake receiver UMTS downlink in fully loaded cell (K=15, Nsf=16) with fast fading Rayleigh channel, ITU Veh. A power delay profile. Doppler frequencies 0Hz (0 km/h), 92.5Hz (50km/h), 185Hz (100km/h), 370Hz (200km/h). Common pilot-aided channel estimation. .............................................................263 Figure 7-43: BER performance of rate 1/3 turbo coded SISO chip equalised UMTS downlink in fully loaded cell (K=15, Nsf=16) with fast fading Rayleigh channel, ITU Veh. A power delay profile. Doppler frequencies 0Hz (0 km/h), 92.5Hz (50km/h), 185Hz (100km/h), 370Hz (200km/h). Common pilot-aided channel estimation. .............................................................264 Figure 8.1 The CRLB for channel estimation in MIMO channels with Hadamard (orthogonal) pilot signals; the pilot length is 16............................................................................................275 Figure 8.2 The CRLB for channel estimation in MIMO channels with random binary pilot signals; the pilot length is 16 ...................................................................................................275 Figure 8.3 The CRLB for channel estimation in MIMO channels with random Gaussian pilot signals; the pilot length is 16 ....................................................................................................276 Figure 8.4 The CRLB for channel estimation in MIMO channels with Hadamard (orthogonal) pilot signals; the pilot length is 256 .........................................................................................276 Figure 8.5 The CRLB for channel estimation in MIMO channels with random binary pilot signals; the pilot length is 256 ..................................................................................................277 Figure 8.6 The CRLB for channel estimation in MIMO channels with random Gaussian pilot signals; the pilot length is 256 .................................................................................................277 Figure 8.7 CRLB for deterministic MIMO channels with Nr receive antennas and Hadamard pilot sequences vsnumber of transmit antennas....................................................................280 Figure 8.8 CRLB for deterministic MIMO channels with Nr receive antennas and random binary pilot sequences vsnumber of transmit antennas ...................................................................282 Figure 8.9 CRLB for deterministic MIMO channels with $N_r$ receive antennas and random Gaussian pilot sequences vsnumber of transmit antennas .................................................283 Figure 8.10 Lower and upper bounds of the CRLB of frequency estimation for all real binary pilot signals of length 4 symbols in deterministic MISO channels with 2 transmit antennas; SNR=15 dB .................................................................................................................................284 Figure 8.11 Lower and upper bounds of the CRLB of frequency estimation for optimal binary pilot signals of length $N_0=4$ symbols in deterministic MISO channels with 2 transmit antennas; SNR=15 dB ................................................................................................................285 Figure 8.12 Lower and upper bounds of the CRLB of frequency estimation for optimal binary pilot signals of length 5 symbols in deterministic MIMO channels with 2 transmit antennas; SNR=15 dB ...............................................................................................................285 Figure 8.13 Lower and upper bounds of the CRLB of frequency estimation for optimal binary pilot signals of length 6 symbols in deterministic MISO channels with 2 transmit antennas; SNR=15 dB .................................................................................................................................286 Figure 8.14 Lower and upper bounds of the CRLB of frequency estimation for optimal binay pilot signals of length 6 symbols in deterministic MISO channels with 3 transmit antennas; SNR=15 dB .................................................................................................................................287 Figure 8.15 MCRLB, ACRLB, and conventional CRLB for MISO fading channels with Hadamard pilot sequences vsnumber of transmit antennas....................................................................294 Figure 8.16 MCRLB, ACRLB, and conventional CRLB for MISO fading channels with random binary pilot sequences vsnumber of transmit antennas .......................................................295
FLOWS IST-2001-32125 Deliverable No:D14 Figure 8.17 MCRLB,ACRLB,and conventional CRLB for MISO fading channels with random of transm ante 296 Figu of non-diagona th random bina Fig e 8.19 Maxir n of non-diagonal elements of the correlation matrix of the random Gaussian pilot signals... .298 Figure 8.20 MCRLB,ACRLB,and conventional CRLB for MISO fading channels with Hadamard 299 Fig hi ACRLB,a RonventonalcRLBorMS0fadingchannelswithFando .300 22 MCRI B ACRI B a ntional CRLB for MISO fading channels with andon 301 Figure 8.23 ML fre ncy estimator for MISO Rayleigh fading char nels with AWGN when the fading covariance matrix,correlation matrix of the pilot signals,and noise variance are 304 8 24 Ml fre tn MISO BA fading co ariance matri ix,correlation matrix of the riance ar a I are diagonal, eindepen t fading of channel coeffic clents an 305 825fr ator for Mlso ravleigh fading cha els with AWGN when the fading covariance matrx is unknown,but the correlatior matrix of th e pilot signals and are kn and are diagonal,L.eindependent fading of cha nts an 20 imates with the crLB in MIso fading channels;Hadamard pilot sequences:2 transmit antennas:a frequency offset of 0.01 Fig e.27 Comp with the CRL 0.0139 309 0.01ng chan 310 Fig re 8 29 ML fr it ante in MISO fading channels:Hadamard pilot sequences of length 64:a frequency offset of 0.013 310 Figure 8.30 ML fre ad in a Misc 1s4 314 Fig 8.32 ML fre is ng el:Hacy est matio on as a functio o of th .312 Figure 8.33 ML fre uency estimation as a function of the frequency to be estimated in a MISO :Hadamard umber of transmit antenna Fig 8.34 ML fre ed in a Mlso inary random pilotncof tong32 tho numbor of transm fading char 313 Figure 8.35 ML frequ nun thof trans nit ante s in MISO fading c gth 128;a frequ encyosatof02c 314 31"December 2003 Page15
FLOWS IST-2001-32125 Deliverable No: D14 31st December 2003 Page 15 Figure 8.17 MCRLB, ACRLB, and conventional CRLB for MISO fading channels with random Gaussian pilot sequences vsnumber of transmit antennas .................................................296 Figure 8.18 Maximum of non-diagonal elements of the correlation matrix of the random binary pilot signals ................................................................................................................................297 Figure 8.19 Maximum of non-diagonal elements of the correlation matrix of the random Gaussian pilot signals ...............................................................................................................298 Figure 8.20 MCRLB, ACRLB, and conventional CRLB for MISO fading channels with Hadamard pilot sequences vsSNR ..............................................................................................................299 Figure 8.21 MCRLB, ACRLB, and conventional CRLB for MISO fading channels with random binary pilot sequences vsSNR .................................................................................................300 Figure 8.22 MCRLB, ACRLB, and conventional CRLB for MISO fading channels with random Gaussian pilot sequences vsSNR ............................................................................................301 Figure 8.23 ML frequency estimator for MISO Rayleigh fading channels with AWGN when the fading covariance matrix, correlation matrix of the pilot signals, and noise variance are known...........................................................................................................................................304 Figure 8.24 ML frequency estimator for MISO Rayleigh fading channels with AWGN when the fading covariance matrix, correlation matrix of the pilot signals, and noise variance are knownThe matrices and are diagonal, i.eindependent fading of channel coefficients and orthogonal pilot signals.............................................................................................................305 Figure 8.25 ML frequency estimator for MISO Rayleigh fading channels with AWGN when the fading covariance matrix is unknown, but the correlation matrix of the pilot signals and noise variance are knownThe matrices and are diagonal, i.eindependent fading of channel coefficients and orthogonal pilot signals ...............................................................................307 Figure 8.26 Comparison of frequency errors of ML frequency estimates with the CRLB in MISO fading channels; Hadamard pilot sequences; 2 transmit antennas; a frequency offset of 0.013 ............................................................................................................................................309 Figure 8.27 Comparison of frequency errors of ML frequency estimates with the CRLB in MISO fading channels; Hadamard pilot sequences; 4 transmit antennas; a frequency offset of 0.013 ............................................................................................................................................309 Figure 8.28 Comparison of frequency errors of ML frequency estimates with the CRLB in MISO fading channels; Hadamard pilot sequences; 8 transmit antennas; a frequency offset of 0.013 ............................................................................................................................................310 Figure 8.29 ML frequency estimation vsthe number of transmit antennas in MISO fading channels; Hadamard pilot sequences of length 64; a frequency offset of 0.013 ................310 Figure 8.30 ML frequency estimation as a function of the frequency to be estimated in a MISO fading channel; Hadamard pilot sequences of length 8; the number of transmit antennas is 4 ....................................................................................................................................................311 Figure 8.31 ML frequency estimation as a function of the frequency to be estimated in a MISO fading channel; Hadamard pilot sequences of length 16; the number of transmit antennas is 4 ................................................................................................................................................312 Figure 8.32 ML frequency estimation as a function of the frequency to be estimated in a MISO fading channel; Hadamard pilot sequences of length 16; the number of transmit antennas is 8 ................................................................................................................................................312 Figure 8.33 ML frequency estimation as a function of the frequency to be estimated in a MISO fading channel; Hadamard pilot sequences of length 32; the number of transmit antennas is 8 ................................................................................................................................................313 Figure 8.34 ML frequency estimation as a function of the frequency to be estimated in a MISO fading channel; binary random pilot sequences of length 32; the number of transmit antennas is 8 ...............................................................................................................................313 Figure 8.35 ML frequency estimation vsnumber of transmit antennas in MISO fading channels; binary random pilot sequences of length 64; a frequency offset of 0.2 ...............................314 Figure 8.36 ML frequency estimation vsnumber of transmit antennas in MISO fading channels; binary random pilot sequences of length 128; a frequency offset of 0.2 .............................314
FLOWS 1ST-2001-32125 Deliverable No:D14 Figure 9-1 The demultiplexer for Space-Time Turbo Codes32 Figure 9-2 Simulation results for MIMO detectors.... 322 Figure 9-3 Receiver block diagram.. ..323 Figure 9-4 Comparison of BER performance of different estimation schemes....325 Figure 9-5 MSE versus Eb/No as the number of iteration increases... ..326 Figure 9-6 BER versus Eb/No as the number of iterations increases.. .326 Figure 10.1 Concept of the finite scatterers channel model,illustrating(a)typical single scatterered path p and (b)reflected path.... 328 31December 2003 Page 16
FLOWS IST-2001-32125 Deliverable No: D14 31st December 2003 Page 16 Figure 9-1 The demultiplexer for Space-Time Turbo Codes .........................................................320 Figure 9-2 Simulation results for MIMO detectors..........................................................................322 Figure 9-3 Receiver block diagram ..................................................................................................323 Figure 9-4 Comparison of BER performance of different estimation schemes ..........................325 Figure 9-5 MSE versus Eb/No as the number of iteration increases............................................326 Figure 9-6 BER versus Eb/No as the number of iterations increases..........................................326 Figure 10.1 Concept of the finite scatterers channel model, illustrating (a) typical single scatterered path p and (b) reflected path.................................................................................328