An Improved Soft Interference Cancellation Based Combined Probability Data Association and Sphere Decoding algorithm for MIMO detection

This paper presents a multi-input multi-output (MIMO) detector that combines probability data association (PDA) and sphere decoding (SD) algorithms. Because of the PDA's natural soft output, soft interference cancellation (SIC) idea is taken to improve the performance. In the presented combining scheme, two different iterative stopping criterions for PDA's module are discussed. Simulation results show that the combination can lead to a tradeoff between performance and complexity of the independent PDA and SD algorithms. In low SNR region, the proposed scheme show s a significant complexity reduction compared to SD, which compensate the deficiency of SD in that region. The SIC based scheme can perform better than the one without SIC, especially at high modulation level. It shows that about 1.8 dB SNR reduction can be obtained by SIC based scheme in the 64QAM situation at the BER =10
^-3
compared to HIC based one.