Performance analysis and soft demapping for coded MIMO-OFDM systems

Multiple input multiple output-Orthogonal frequency division multiplexing (MIMO-OFDM) is a viable solution for providing high data rate services in harsh channel environments. The optimum receivers for them are those based on the maximum likelihood criterion. However, they have a prohibitive complexity especially when channel dimensions are high and coding is employed. Zero Forcing (ZF) and Linear Minimum Mean Square Error (MMSE) receivers on the other hand provide practicable and low complexity solutions for detection, but require soft demappers to deduce the soft bits information contained in each of the received symbols. In this work, we present the ZF and MMSE receiver analysis of a bit interleaved and coded MIMO-OFDM system and propose a soft output demapper based on MMSE equalizer output to demap the information needed for viterbi decoding. A comparison of the proposed soft demapper with conventional soft demappers in literature show a significant performance improvement. We also noticed that it is more advantageous to apply the proposed demapper on a MIMO-OFDM system employing higher modulation schemes.