Enhanced Channel Hardening-Exploiting Message Passing Algorithm For Large-scale MIMO Detection

In this paper, the enhanced channel hardening-exploiting message passing (ECHEMP) algorithm is proposed for large-scale MIMO detection to improve the decoding performance with less complexity cost, so as to a better decoding tradeoff. Specifically, the proposed algorithm consists of two stages. In the first stage, to enhance the convergence performance of the iterative detection symbols estimation, we propose to use the discrete Gaussian distribution rather than the uniform one as the default initial probability distribution, and give the initial estimation as the mean of the discrete Gaussian distribution under channel hardening phenomenon. In the second stage, compared to iterative calculations of all detection symbols, the number of calculations of detection symbols in the iterative process is reduced through a probability threshold, which efficiently reduces computational complexity. Finally, simulation results about MIMO detection are presented to verify the performance improvement and the complexity reduction.