Optimal Combining of STBC and Spatial Multiplexing for MIMO-OFDM

1
Optimal Combining of STBC and Spatial
Multiplexing for MIMO-OFDM

• Taehyun Jeon, Heejung Yu, and Sok-kyu Lee
• Wireless LAN Modem Research Team, ETRI
• Jihoon Choi and Yong H. Lee
• KAIST

2
Contents

• Spatial Multiplexing for MIMO Systems
• Detection Methods
• Transmit Diversity for MIMO Systems
• Transmit Diversity using STBC
• Combining of STBC and Spatial Multiplexing
• Simulation results
• Conclusion

3
Spatial Multiplexing for MIMO Systems

• Transmit independent parallel data streams
  through multiple antennas
• Increase the data rate T times faster than SISO

4
Detection of Spatial Multiplexed Signal

• Maximum Likelihood (ML) Detection
• Complexity LT operation required (L
  constellation size)
• Linear Detection
• V-BLAST Detection
• Nulling Canceling Successive detection of
  data streams (layer by layer)
• Complexity O(29T3/3)

5
Spatial Multiplexing for MIMO-OFDM

• OFDM sub-divides the wideband channel into
  multiple flat fading subcarriers and can be
  implemented simpler and more efficient
  demodulation processing (IFFT/FFT and FEQ) over
  single carrier systems
• MIMO-OFDM with TR2 Complexity O(29NT3/3)
  when V-BLAST detection used

6
Transmit Diversity for MIMO Systems

• Techniques compatible to IEEE 802.11a
• Delay Diversity and Random Phase Diversity
• Simple to implement but not significant
  performance gain with number of antenna increase
• Space-Time Block Coding (STBC)
• Simpler implementation than Trellis Coding
• Alamouti code provides diversity order 2 with T2
  and R1

7
STBC for MIMO-OFDM
8
N x N MIMO Channels

• In terms of Spatial Multiplexing
• Data rate increases as number of antenna
  increases
• No significant diversity advantages

• In terms of Diversity
• Maximum diversity gain can be achieved
• Higher order modulation needed to increase the
  data rate ? SNR loss

• For best performance for a target data rate,
  optimal combining of above should be considered

9
Combining of 2-layer Spatial Multiplexing and
Alamouti Code
Tx 1 Tx 2 Tx 3 Tx 4
t2n a2n a2n1 b2n b2n1
t2n1 -a2n1 -a2n -b2n1 b2n
10
Simulation Parameters

• IEEE 802.11a PHY Based Frame
• Number of Subcarriers (data subcarriers) 64 (48)
• Number of Cyclic Prefix 16
• Sampling Rate 20MHz
• Modulation QPSK, 16QAM, 256QAM
• Number of Tx and Rx Antenna 2, 4
• Channel Coding None
• Channel Model Independent MIMO Channel
• ETSI/BRAN Channel Model B (RMS Delay Spread
  100ns)
• Quasi Static Channel (no change within one frame)

11
Simulation Results (T2,R2)

• Mode 2 (16QAM and Alamouti) performs better than
  Mode 1 (QPSK and 2-layered)
• Diversity gain of order 4 in Mode 2 overcomes the
  4dB degradation of higher order modulation

12
Simulation Results (T4,R2)

• Mode 2 (16QAM and 4x4 STBC) performs better than
  Mode 1 (QPSK and 2-layered Alamouti) beyond
  Eb/No8dB
• Diversity order 4 over 3

13
Simulation Results (T4,R4)

• Three different combinations are tested
• Mode 2 (16QAM and 2-layered Alamouti) performs
  best beyond 5dB
• Mode 1 (QPSK and 4-layered) Average diversity
  gain 2.5
• Mode 2 and Mode 3 diversity gain about the same
  but 256QAM performs 8dB worse than 16QAM

14
Conclusions

• Optimal MIMO-OFDM system should be selected by
  trading-off the spatial multiplexing and
  diversity gain for a given antenna arrangement
• Candidate combined systems of STBC and spatial
  multiplexing proposed and their performances
  compared for extended IEEE 802.11a systems
• Performance evaluations with realistic MIMO
  channel model for further work items