Probabilistic Searching for MIMO Detection Based on Lattice Gaussian Distribution

Abstract

In this paper, a deterministic sampling decoding strategy for multiple-input multiple output (MIMO) systems is studied, which performs probabilistic searching according to a probability threshold in the lattice Gaussian distribution. Motivated by model probabilistic twin (MPT), the randomness in obtaining the target decoding solution is overcome by the proposed probabilistic searching decoding (PSD) algorithm, which brings considerable decoding gains in both performance and complexity. Specifically, the decoding radius of PSD is derived while the decoding complexity in terms of the number of visited nodes during the searching is also upper bounded, leading to an explicit decoding trade-off. Meanwhile, we generalize PSD by the mechanism of candidate protection so that it enjoys a flexible performance between the suboptimal successive interference cancelation (SIC) decoding and the optimal maximum likelihood (ML) decoding by adjusting the initial search size <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">K</i> . Methods for further optimization and complexity reduction of the proposed PSD algorithm are also given. Finally, simulation results based on MIMO detection are presented to confirm the tractable and flexible decoding trade-off of the proposed PSD algorithm.