A Low-Complexity Iterative MIMO Detection Scheme Using the Soft-Output M-Algorithm

This paper presents a soft-output version of the M-algorithm (SOMA) for complexity reduction in an iterative MIMO receiver. Unlike existing algorithms, SOMA utilizes paths that are traversed but discarded in a pruned tree to reduce complexity. It is shown that discarded paths in a heavily pruned tree carry a significant amount of soft-information which should not be ignored. We show a simple and effective method of using these paths without having to store them. BER performance is presented for both convolutionaland turbo-coded MIMO systems that use bit-interleaved coded-modulation. Flat fading with known channel state information is assumed. Improvement is observed in using soft-information from discarded paths for a convolutional outer code. For a powerful turbo outer code, assigning a constant to unknown likelihoods suffices.