Youssef Nasser - Innovative 3D MIMO schemes in DVB systems

Transcript

1. Innovative 3D MIMO schemes in Digital Video Broadcasting (DVB) systems 1 UMR 6164 Youssef Nasser Research Engineer with INSA - Rennes
2. Summary Brief overview of DVB – Some Spec. of DVB-T2 Reminder on MIMO systems Proposition of 3D code for inter-cell and intra-cell coding SFN architecture Open Area Environment Gap Area Environment Simulation results Youssef NASSER 2 MIMO-OFDM Study Simulation results Proposition of 3D code for hybrid satellite-terrestrial transmission LMS Channel Simulations results Adaptation of the 3D code to a complete network Characterisation Simulations results Conclusions
3. Summary Brief overview of DVB – Some Spec. of DVB-T2 Reminder on MIMO systems Proposition of 3D code for inter-cell and intra-cell coding SFN architecture Open Area Environment Gap Area Environment Simulation results Youssef NASSER 3 MIMO-OFDM Study Simulation results Proposition of 3D code for hybrid satellite-terrestrial transmission LMS Channel Simulations results Adaptation of the 3D code to a complete network Characterisation Simulations results Conclusions
4. The second generation of Digital TV 1992 1992 1993 1993 1994 1994 1995 1995 1996 1996 1997 1997 1998 1998 1999 1999 2000 2000 2001 2001 2002 2002 2003 2003 2004 2004 2005 2005 2006 2006 2007 2007 2008 2008 2009 2009 2010 2010 2011 2011 2012 2012 From From Analog Analog TV to Digital TV TV to Digital TV ASO : ASO : Analog Analog TV Switch TV Switch- -Off Off From SDTV to HDTV From SDTV to HDTV Youssef NASSER 4 MIMO-OFDM Study From Digital TV to Mobile TV From Digital TV to Mobile TV
5. Objective of the 2nd Generation HDTV HDTV over over TERRESTRIAL TERRESTRIAL HDTV HDTV over over TERRESTRIAL TERRESTRIAL Youssef NASSER 5 MIMO-OFDM Study
6. what is new… BICM BICM BICM BICM BICM BICM BICM BICM BICM BICM ... BB BB BB BB BB BB BB BB BB BB Sync Sig PLP PLP PLP PLP PLP PLP PLP PLP PLP PLP Sync RF RF OFDM OFDM Modulator Modulator RF RF OFDM OFDM Modulator Modulator FRAME FRAME BUILDER BUILDER Bit Bit Interleaved Interleaved Coded Coded Modulation Modulation & & Rotated Rotated INPUT INPUT PROCESS PROCESS Sound Data Sig ... ... Youssef NASSER 6 MIMO-OFDM Study BICM BICM BICM BICM BB BB BB BB Sound Data PLP PLP PLP PLP RF RF Modulator Modulator (MISO option) (MISO option) Rotated Rotated Constellation Constellation
7. Some specifications Multiple services, flexibility, no “magic” number… FFT size: 1k, 2k, 4k, 8k, 16k, 16kE, 32k, 32kE Bandwidth: 1.7, 5, 6, 7, 8, 10 MHz Encoding: BCH outer code and LDPC inner code PAPR: ACE and TR First Broadcasting Standard implementing an Alamouti Alamouti code Block 1 Youssef NASSER 7 MIMO-OFDM Study First Broadcasting Standard implementing an Alamouti code using a MISO scheme Alamouti code Block 2
8. Summary Brief overview of DVB – Some Spec. of DVB-T2 Reminder on MIMO systems Proposition of 3D code for inter-cell and intra-cell coding SFN architecture Open Area Environment Gap Area Environment Simulation results Youssef NASSER 8 MIMO-OFDM Study Simulation results Proposition of 3D code for hybrid satellite-terrestrial transmission LMS Channel Simulations results Adaptation of the 3D code to a complete network Characterisation Simulations results Conclusions
9. Reminder on MIMO Basis: 2D encoding such as Space-Time (ST), Space- Frequency (SF)… (s1 ,s2 ) ST encoder x1 =f1 (s1 ,s2 ) x2 =f2 (s1 ,s2 ) ST detector Y1 =g1 (x1 ,x2 ,h) Y2 =g2 (x1 ,x2 ,h) h11 h22 h21 h12 Youssef NASSER 9 MIMO-OFDM Study ( ) ( ) ( ) ( ) 2 2 2 22 12 * 21 1 21 * 22 11 * 21 2 1 2 22 * 12 12 * 11 1 21 * 12 2 11 1 ˆ ˆ w s h h h s h h h h s w s h h h h s h h h s + + + + = + + + + = Received signal Useful signal Interfering signal
10. MIMO receiver Orthogonal ST schemes (e.g. Alamouti): no interfering signal Very simple ML decoder. Non-orthogonal ST schemes (e.g. Golden): interfering signal Complex ML decoder, Iterative receiver. Youssef NASSER 10 MIMO-OFDM Study RX_OFDM Sink_Bits SISOLogMap Decoder Depuncturing Bit_Deinterleaver Demapping Iterative Equalizer FFT FFT Symb. Deintrlver CP Remove CP Remove RecAnt=1 RecAnt=MR Symb. Deintrlver Bit_Interleaver SoftGrayMapper Iterative RX_OFDM Interference canceller RX_OFDM Sink_Bits SISOLogMap Decoder Depuncturing Bit_Deinterleaver Demapping Iterative Equalizer FFT FFT Symb. Deintrlver CP Remove CP Remove RecAnt=1 RecAnt=MR Symb. Deintrlver Bit_Interleaver SoftGrayMapper Iterative RX_OFDM Interference canceller
11. ST schemes         = = = ) ,..., ( ... ) ,..., ( ) ,..., ( ... ) ,..., ( 1 21 1 12 2 1 21 1 11 1 Q Q Q Q s s f s s f x s s f s s f x X (s1 ,…,sQ ) ST encoder Q T ST rate L=Q/T Space-Time Block Codes (STBC). Youssef NASSER 11 MIMO-OFDM Study Orthogonal (X.XH=I) Non-Orthogonal       − = * 1 * 2 2 1 s s s s X Alamouti (Q=T=2) Space-Time Block Codes (STBC).       + + + + = ) ( ) ( ) ( ) ( 5 1 2 1 4 3 4 3 2 1 s s s s s s s s X θ β θ β µ θ β θ β 2 5 1+ = θ θ θ − =1 ) 1 ( 1 θ β − + = j ) 1 ( 1 θ β − + = j j = µ 1 − = j Golden Code (Q=4, T=2)
12. Benefits: Capacity SISO: C = E{log2 ( 1 + |h|2 SNR)} [b/s/Hz] (with E{|h|2}=1) Low increase with SNR Youssef NASSER 12 MIMO-OFDM Study MIMO: C = E{ log2 det[ I+MR /MT SNR HHH] } [b/s/Hz] with E{HHH}=I High SNR: C ≈ min(MR ,MT ) log2 (SNR) Linear increase with min(MR ,MT ) Low SNR: C ≈ MR . SNR log2 (e) Receiving diversity
13. Benefits: Diversity SISO: n hs y + = n h s h s * 2 ˆ + = MRC: Diversity d=1 since law with one degree of freedom 2 χ MIMO: Youssef NASSER 13 MIMO-OFDM Study MIMO:         +                 =         2 1 2 1 22 21 12 11 2 1 n n s s h h h h y y ( ) ( ) ( ) ( ) 2 2 2 22 12 * 21 1 21 * 22 11 * 21 2 1 2 22 * 12 12 * 11 1 21 * 12 2 11 1 ˆ ˆ w s h h h s h h h h s w s h h h h s h h h s + + + + = + + + + = MRC: Diversity d=2 since law with 4 degrees of freedom 2 χ
14. Summary Brief overview of DVB – Some Spec. of DVB-T2 Reminder on MIMO systems Proposition of 3D code for inter-cell and intra-cell coding SFN architecture Open Area Environment Gap Area Environment Simulation results Youssef NASSER 14 MIMO-OFDM Study Simulation results Proposition of 3D code for hybrid satellite-terrestrial transmission LMS Channel Simulations results Adaptation of the 3D code to a complete network Characterisation Simulations results Conclusions
15. SISO in SFN Architecture SFN : Several transmitters transmit at the same moment the same signal on the same frequency. D d d1 d D d d1 d Open area Youssef NASSER 15 MIMO-OFDM Study P1 d1 d2 P2 P1 d1 d2 P2 Gap area τ h(τ) ∆τ =(d1 -d1 )/c β = 10.log10 (P2 /P1 ) τ h(τ) β = ?? ∆τ =??
16. MIMO In SFN Architecture D d d MT antennas MT antennas d1 d2 Gap area Youssef NASSER 16 MIMO-OFDM Study P1 P2 d2 Open area
17.       − = * * 2 1 s s s s X • Dispersion Matrix First construction : Alamouti code MIMO In SFN Architecture Youssef NASSER 17 MIMO-OFDM Study P1 d1 d2 P2 Open area Gap area 1 s * 1 s * 2 s − * 2 s − 2 s 2 s    − 1 2 s s
18. • Dispersion Matrix First construction : Golden code MIMO In SFN Architecture       + + + + = ) ( ) ( ) ( ) ( 5 1 2 1 4 3 4 3 2 1 s s s s s s s s X θ β θ β µ θ β θ β 2 5 1+ = θ θ θ − =1 ) 1 ( 1 θ β − + = j Youssef NASSER 18 MIMO-OFDM Study   + + ) ( ) ( 5 2 1 4 3 s s s s θ β θ β µ ) 1 ( 1 θ β − + = j ) 1 ( 1 θ β − + = j j = µ 1 − = j P1 d1 d2 P2 Open area Gap area 1 2 ( ) s s β θ + 1 2 ( ) s s β θ + 3 4 ( ) s s µβ θ + 1 2 ( ) s s β θ + 1 2 ( ) s s β θ +
19. [ ] 1 2 tr s s = X MIMO In SFN Architecture • Dispersion Matrix First construction : spatial multiplexing Youssef NASSER 19 MIMO-OFDM Study P1 d1 d2 P2 Open area Gap area 1 s 2 s P1 d1 d2 P2 Open area Gap area 1 s 1 s 2 s 2 s
20. Simulations Parameters FFT size 8K Sampling frequency (f s =1/T s ) 9.14 MHz Guard interval (GI) duration 1024×T s =112 µs Rate R c of convolutional code 1/2, 2/3, 3/4 Polynomial code generator (133,171) o Youssef NASSER 20 MIMO-OFDM Study Polynomial code generator (133,171) o Channel estimation perfect Constellation 16-QAM, 64-QAM, 256-QAM Spectral Efficiencies η= 4 and 6 [b/s/Hz] Channels Rayleigh i.i.d. and COST 207 TU-6 Adaptation of the Interleaver for the 256-QAM constellation
21. Spectral Efficiency ST scheme ST rate (L) Constellation Rc η=4 [b/sec/Hz] Alamouti 1 64-QAM 2/3 SM 2 16-QAM 1/2 Golden 2 16-QAM 1/2 Simulation Parameters Youssef NASSER 21 MIMO-OFDM Study Golden 2 16-QAM 1/2 η=6 [b/sec/Hz] Alamouti 1 256-QAM 3/4 SM 2 64-QAM 1/2 Golden 2 64-QAM 1/2
22. MIMO Results in SFN 10 12 14 16 18 E b /N 0 [dB] Alamouti SM Golden 3 dB 14 16 18 20 22 24 E b /N 0 [dB] Alamouti SM Golden 3 dB Frequency Rayleigh i.i.d channel Youssef NASSER 22 MIMO-OFDM Study -12 -10 -8 -6 -4 -2 0 6 8 β [dB] β lim -12 -10 -8 -6 -4 -2 0 10 12 β [dB] β lim
23. 3D in SFN Architecture D d d P1 P2 MT antennas MT antennas Gap area Youssef NASSER 23 MIMO-OFDM Study (2) (2) (1) 11 1 (2) (2) 21 2 U U   =     X X X X X L L First Layer ,11 1 ,1 1 (2) , 1 1 , 1 ( ,... ) ( ,... ) ( ,... ) ( ,... ) T T pq M pq V M pq pq M M pq M V M f s s f s s f s s f s s     =       X K M O M L Second Layer Open area Inter-cell ST code Intra-cell ST code
24. MIMO Results in SFN 16 18 20 22 24 E b /N 0 [dB] 3D code Alamouti Golden η= 6[b/s/Hz] 3.1 dB 19 20 21 22 23 24 b /N 0 [dB] 3D code 10 Km/h Alamouti 10 Km/h Golden 10 Km/h 3D code 60 Km/h Alamouti 60 Km/h Golden 60 Km/h Youssef NASSER 24 MIMO-OFDM Study -12 -10 -8 -6 -4 -2 0 10 12 14 16 β [dB] E η= 4[b/s/Hz] 1.5 dB -12 -10 -8 -6 -4 -2 0 14 15 16 17 18 β [dB] E b h(τ) β = f(∆τ) ∆τ
25. MIMO Results in SFN 14 16 18 20 22 E b /N 0 [dB] 3D code Alamouti Golden η η η η = 6 [b/s/Hz] Youssef NASSER 25 MIMO-OFDM Study 0 50 100 150 200 250 300 350 400 450 8 10 12 14 ∆ τ [samples] E η η η η = 4 [b/s/Hz] τ h(τ) ∆τ
26. Summary Brief overview of DVB – Some Spec. of DVB-T2 Reminder on MIMO systems Proposition of 3D code for inter-cell and intra-cell coding SFN architecture Open Area Environment Gap Area Environment Simulation results Youssef NASSER 26 MIMO-OFDM Study Simulation results Proposition of 3D code for hybrid satellite-terrestrial transmission LMS Channel Simulations results Adaptation of the 3D code to a complete network Characterisation Simulations results Conclusions
27. MIMO scheme for DVB-SH / DVB-NGH DVB-SH broadcast head-end Satellite links data content Youssef NASSER 27 MIMO-OFDM Study DVB-SH broadcast head-end Satellite links Terrestrial links
28. Satellite link and channel problem Youssef NASSER 28 MIMO-OFDM Study S1 S2 S3
29. Land Mobile Satellite Channel Fontan model is widely used (narrow-band channel model). Based on Loo’s distribution and Markov chain model. The Loo’s distribution assumption: and Youssef NASSER 29 MIMO-OFDM Study The received signal (according to Loo’s distribution) is a Rice distributed signal where its mean is log-normally distributed. The received signal could be generated according to:
30. Land Mobile Satellite Channel The Markov chain states are defined by: S1: LOS conditions S2: moderate shadowing conditions S3: deep shadowing conditions The different states (and their probabilities) depend on the elevation angle (θ°) p22 Youssef NASSER 30 MIMO-OFDM Study S1 S2 S3 p12 p21 p13 p23 p32 p31 p11 p22 p33
31. Land Mobile Satellite Channel 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 -40 -35 -30 -25 -20 -15 -10 -5 0 5 Traveled distance (m) Signal level (dB/LOS) -40 -35 -30 -25 -20 -15 -10 -5 0 5 0 1000 2000 3000 4000 5000 6000 Received Signal Level relative to LOS (dB) N° samples Youssef NASSER 31 MIMO-OFDM Study Traveled distance (m) Received Signal Level relative to LOS (dB) 10 20 30 40 50 60 70 80 90 100 -14 -12 -10 -8 -6 -4 -2 0 2 4 Traveled distance (m) Signal level (dB/LOS)
32. Elevation S1: LOS S2: Interm. Shad S3: Deep Shad. µ Σ Σ Σ Σ MP µ Σ Σ Σ Σ MP µ Σ Σ Σ Σ MP 10° -0.1 0.5 -19 -8.7 3 -12 -12.1 6 -25 Average Loo model parameters in dB for various angles and suburban area Simulation Parameters Youssef NASSER 32 MIMO-OFDM Study 30° -0.5 1 -15 -4.7 1.5 -19 -7 3 -20 50° -0.5 1 -17 -6.5 2.5 -17 -14 2.5 -20 70° -0.2 0.5 -15 -6.0 2.1 -17 -11.5 2 -20
33. Markov chain matrices P and W for various angles and suburban area Matrix P (10°) 0.9174 0.0512 0.0314 Matrix W (10°) 0.4389 0.1064 0.7802 0.1134 0.2599 0.0285 0.1151 0.8564 0.3012 Matrix P (30°) 0.9531 0.0350 0.0119 Matrix W (30°) 0.7467 0.1891 0.6198 0.1911 0.1511 0.0631 0.3065 0.6304 0.1022 Matrix P 0.7498 0.2462 0.004 Matrix W 0.1626 Simulation Parameters Youssef NASSER 33 MIMO-OFDM Study Matrix P (50°) 0.7498 0.2462 0.004 Matrix W (50°) 0.1626 0.0479 0.9160 0.0361 0.7642 0.0554 0.3296 0.615 0.0732 Matrix P (70°) 0.8969 0.0898 0.0406 Matrix W (70°) 0.0634 0.0106 0.9336 0.0558 0.5337 0.0064 0.0738 0.9198 0.4029
34. 12 14 16 18 [dB] to obtain a BER=10-4 Required Eb/N0 to obtain a BER=10-4, single layer case One satellite antenna and one terrestrial antenna Simulation Results Youssef NASSER 34 MIMO-OFDM Study 2 2.5 3 3.5 4 4.5 5 5.5 6 4 6 8 10 12 η b/s/Hz Required E b /N 0 [dB] to obtain a BER=10 Alamouti θ = 30° Golden θ = 30° Alamouti θ = 50° Golden θ = 50°
35. Required Eb/N0 to obtain a BER=10-4, double layer case, η=2, 6 b/s/Hz 12 14 16 18 20 22 [dB] to obtain a BER=10-4 Alamouti Golden LSTBC η η η η = 6 b/s/Hz Simulation Results Youssef NASSER 35 MIMO-OFDM Study 10 20 30 40 50 60 70 4 6 8 10 φ° Required E b /N 0 [dB] to obtain a BER=10 η η η η = 2 b/s/Hz
36. Required Eb/N0 to obtain a BER=10-4, double layer case, η=4 b/s/Hz 12 13 14 15 16 17 18 [dB] to obtain a BER=10-4 Alamouti Golden LSTBC Simulation Results Youssef NASSER 36 MIMO-OFDM Study 10 20 30 40 50 60 70 8 9 10 11 12 φ° Required E b /N 0 [dB] to obtain a BER=10
37. Summary Brief overview of DVB – Some Spec. of DVB-T2 Reminder on MIMO systems Proposition of 3D code for inter-cell and intra-cell coding SFN architecture Open Area Environment Gap Area Environment Simulation results Youssef NASSER 37 MIMO-OFDM Study Simulation results Proposition of 3D code for hybrid satellite-terrestrial transmission LMS Channel Simulations results Adaptation of the 3D code to a complete network Characterisation Simulations results Conclusions