Adaptive stream mapping multi antenna systems with low complexity iterative detection

In this paper, we propose an adaptive linear precoding technique combined with low complexity iterative detection which can significantly improve the performance of spatial multiplexing multiple input, multiple output (MIMO) systems. By allowing different data stream mappings to the transmit antennas, linear precoding techniques can provide full diversity and increased capacity gain, assuming maximum likelihood (ML) detection is used. There has been previous work on selection criteria to select the appropriate precoding matrix to perform the stream mapping, based on ML or minimum mean square error (MMSE) detection. However these selection criteria were designed under uncoded conditions. In this paper we present a selection criterion based on the Extrinsic information transfer (EXIT) chart specifically designed for a reduced complexity iteratively detected coded system with interference cancellation detection. The results show that the detector maintains full diversity and has performance within 0.4 dB of the ML detector. Moreover precoding increases performance by 1.2 dB compared to the ML detector with fixed stream mapping. In addition it outperforms the SNR criterion from previous work.