Coding-assisted MIMO joint detection and decoding in turbo-coded OFDM

A new coding-assisted MIMO (CA-MIMO) receive algorithm is proposed that can approach the performance of an optimal MIMO receiver, but with a significantly reduced complexity. The algorithm starts with linear MIMO processing to obtain the log-likelihood ratios (LLRs) of the channel bits. Then, at the end of an intermediate turbo decoding iteration, it refines a few channel bit LLRs based on the computed distance to only a select number of symbol candidates, thus achieving the savings. The selection is based on the LLRs of the coded bits at that iteration. In a simulated two-transmit two-receive MIMO-OFDM 16-QAM system with a rate-1/2 turbo code, the proposed algorithm can improve the FER performance by more than 3 dB in a COST-259-style bad urban environment, compared with that of the MMSE MIMO receiver, and is only 0.2 dB away from that of the optimal receiver with exponential complexity.