Differential Space-Time Coding for Future Wireless Systems

1. Differential Space-Time Coding for Future Wireless Systems Lingyang Song

2. Contents • Differential Preliminaries • Differential Schematic • Fields of Application • Future Work

3. Differential Preliminaries Lingyang Song Communications Research Group March 6, 2006

4. Differential Preliminaries • Multiple Input Multiple Output Systems (MIMO) • Transmit diversity • Channel model • Maximum likelihood decoding • Space-Time Block Codes • Alamouti scheme • Differential Space-Time Block Codes • Advantage of differential application • Differential encoding and decoding • Quasi-Orthogonal Space-Time Block Codes • Mainly for four transmit antennas, why?

5. Multiple Input Multiple Output (MIMO) • MIMO • Concept • Usually multiple transmit antennas and receive antennas • Expression: • Transmit Diversity • Down-link of a mobile system provides diverse antennas at the transmit end, while requiring only a single antenna on the terminal. • Channel • Slow and flat fading • ML Decoder

6. Space-Time Block Codes: Alamouti Scheme • Achieve full transmit diversity, assuming a system with two transmit antennas, so the diversity is two. • Orthogonal properties • Channel knowledge available at the receiver, one receive antenna

7. Space-Time Block Codes: Alamouti Scheme • Low computational complexity • ML decoder can be simplified into linear processing

8. Differential Space-Time Block Codes • Why Differential? • Perfect channel is not available at the transmitter or receiver • It may be difficult or costly to estimate the channel accurately • Too many training symbols are required, such as MIMO • High mobility situations • Reduce the complexity of the handset • Channel estimation takes a large proportion!

9. Differential Space-Time Block Codes • Differential Encoding • Based on previous signals, the DE computes the next transmitted symbols • Encoded symbols are sent from STBC Encoder

10. Differential Space-Time Block Codes • Differential Decoding The received signals for time 2t+2 can be written as:

11. Differential Space-Time Block Codes • We can further get:

12. Quasi-Orthogonal Space-Time Block Codes • Why QO-STBC is required? • High data rate requirement • Full rate, full diversity orthogonal space-time block codes do not exist for more than two transmit antennas with linear processing at the receiver • low computational complexity • In practical SNR regime, it can provider very promising performance

13. Quasi-Orthogonal Space-Time Block Codes Quasi-Orthogonal Space-Time Coding Block where • Note that has a form similar to the Alamouti code • This code has rate one, but diversity order two, since each symbol passes through only two of the four transmitter antennas. • In comparison to O-STBC for four transmit antennas, this codes can provide better performance in practical SNR region. • Hence, we expect for differential scheme, D-QO-STBC could also obtain some advantages in the similar SNR regime.

14. Differential QO-STBC Lingyang Song

15. Differential QO-STBC • Differential Encoding Process • Differential Decoding Process • Relative Differential Scheme • Simulation Results • Extensions

16. Differential Encoding Schematic • Differential Encoding in each Alamouti partition

17. Differential Decoding Schematic The received signals for time 4t+4 can be written as: where

18. Differential Decoding Schematic • By further transformation, we obtain: where • Clearly, D-QO-STBC with four antennas now is simplified to differential Alomouti scheme, similar decoding methods can be used to recover the original data.

19. Simulation Results of D-QO-STBC • 4 transmit antennas • 1 receive antenna • QPSK, 2bps/Hz • slow and flat fading channels

20. Simulation Results Analysis • In low SNR region, our scheme can give better performance. • The complexity is very low, in the number of transmit antennas and rate • In high SNR, other two curves begin to perform better since it is the diversity that decides the slope. • If the codes have full diversity (only half in our scheme), it will provide better performance in the whole SNR regime!!

21. Extensions • There are other possible structures that can provide behaviours. A couple of examples are given below: • Also, similar ideas can be used to build up a rate ¾ transmission matrix based on the rate orthogonal space-time block code. Examples are given below: where

22. Extensions • Full-diversity differential quasi-orthogonal space-time block codes:

23. Fields of Application • MIMO techniques in 3GPP, ,available through https://www.3gpp.org/ • WINNER,available through https://www.ist-winner.org/

24. Future Work • Differential scheme based on the combination of space-time block codes and BLAST, aiming to maximize spatial multiplexing gain and transmit diversity gain simultaneously • Differential Space-Time Multiplexing • Differential Turbo Space-Time Multiplexing • Space-Time Multiplexing from Generalized Design

25. Differential Quasi-Orthogonal Space-Time Block Codes with Full Transmit Diversity Lingyang Song

26. Contents • Full Diversity Quasi-Orthogonal Space-Time Blocks • Differential Transmission Schematic • Differential Reception Schematic • Simulation Results and Performance Analysis

27. Quasi-Orthogonal STBC Quasi-Orthogonal Space-Time Coding Block where , S are complex signals which are picked up from two constellations

28. Differential Encoding Schematic Fig. 1 , Special Signal Mapping Scheme

29. Differential Encoding Schematic • Figure 2. Differential Encoding Scheme

30. Differential Decoding Schematic • Differential Encoding in each Sub-Block where • Entire QOSTBC code block can be then formed

31. Differential Decoding Schematic • Figure 3. Differential Decoding Scheme

32. Differential Decoding Schematic • Received signals can be written as: where , and

33. Differential Decoding Schematic • Recalling encoding process, we have • We can also write • where

34. Differential Decoding Schematic • Differential Encoding Function • where , ; and are the first row of each coding block

35. Differential Decoding Schematic • The received signals can be written in a matrix form as: • The estimated power can be written as:

36. Differential Decoding Schematic • The estimated interference can be written as:

37. Differential Decoding Schematic • The pair-wise signal detector can be written as:

38. Simulation Results • Figure 4. Differentia Performance by QPSK

39. Simulation Results • Figure 5. Constellations for 8QAM

40. Simulation Results • Figure 6. Differentia Performance by 8QAM

41. Interference Cancellation for Space-Time-Frequency Codes Lingyang Song

42. Contents • INTRODUCTION TO STBC • SFBC TRANSMISSION SCHEME • CONVENTIONAL STBC DETECTOR • ITERATIVE DECODING APPROACH • PERFORMANCE ANALYSIS • CONCLUSIONS

43. SPACE-TIME BLOCK CODES • Reasons For the STBC Existence • Offer Full Diversity Rate • Provide Better System Performance • STBC Structure For 2 Transmit Antennas

44. SFBC TRANSMISSION SCHEME Fig. 1. SFBC Transmission Scheme

45. CONVENTIONAL DETECTOR • Maximum-Likelihood (ML) Decoder • Conventional ML Pair-Wise Signal Detector

46. ITERATIVE DECODING • Step 1 [Initialization]: Set the iteration number k=0, and obtain , i=1,2, from the STBC decoder • Step 2 [Iteration]:Iteration number k=1,2,…,I, • Then, symbol can be obtained via a LS decoder:

47. PERMORMANCE ANALYSIS Fig. 2. New IC based Signal Detector Performance

48. CONCLUSIONS • Conventional signal detector for SFBC can suffer from an irreducible error floor over MIMO-OFDM systems; • The new IC based signal detector can effectively subtract the impact of the fast fading channels; • Performance is improved with the number of iteration increasing;

49. Differential Turbo Bell-Lab Layered Space-Time Architectures L.-Y. Song

50. Contents • Introduction System Model; BLAST, Motivation • Differential BLAST Differential Encoding; Differential Decoding; • Differential Turbo BLAST Serial Concatenated Turbo Codes; Iterative Decoder • Simulation Results • Conclusions