Adaptive linear precoding for iterative maximum likelihood detection in multi-antenna systems

In this paper, we propose a novel precoding selection criterion for coded iterative maximum likelihood (ML) detection in spatial multiplexing multiple input, multiple output (MIMO) systems. With the help of this criterion, the adaptive linear precoding technique can significant improve the system performance and provide full diversity by allowing different data stream mappings to the transmit antennas. Previous work had been done on selection criterion to select the appropriate precoding matrix to perform the stream mapping, based on ML or minimum mean square error (MMSE) detection. However these selection criterion were designed under the uncoded conditions. The proposed criterion in this paper can predict the system performance for a particular channel and precoder combination through estimating the mutual information of the convergence point on the Extrinsic information transfer (EXIT) chart, hence can choose the optimum precoder from the codebook to achieve the maximum performance gain. The results show that the system performance is hugely increased by applying the precoding and the particular selection criterion.