A low-complexity implementation of sampling-based MIMO detection

Abstract

A low-complexity implementation of sequential Monte Carlo (SMC) sampling-based detector is developed for multiple-input multiple-output (MIMO) communication systems. Unlike previous reports about SMC sampling that widely sequentially draw samples and process each sample independently, we present a novel sampling method which collaboratively processes all samples and extracts information from a collection of samples to establish the sampling space to draw next samples. Simultaneously, the proposed method adopts a reselection step to save storage resource and decrease computation burden. Simulations indicate that the proposed solution decreases the necessary amount of samples and improves the system performance compared with the classical SMC detector. The revised detector is also compared with sphere decoding (SD) that has the comparable computation burden, and simulation result shows that it can obtain the same performance as SD with lower complexity.