Low-Complexity CS-Aided MPA Detector for SCMA Systems

In this letter, we present a novel low-complexity multiuser detection for sparse code multiple access systems, combined message passing algorithm (MPA), and the compressed sensing (CS) technique, referred to as a CS-aided MPA detector. Specifically, the detector is divided into a two-step procedure, i.e., initial detection and sparse error correction. In the first step, MPA with a few iterations is employed as initial detection. In the second step, based on the observation that the symbol error produced by initial detection tends to be sparse, we propose a CS-based greedy pursuit algorithm, called threshold-aided adaptive subspace pursuit, to estimate the error accurately. Then, the final estimated vector is obtained by adding the initial estimated vector produced by MPA to the estimated error vector. Simulation results demonstrate that the proposed CS-aided MPA detector provides very similar symbol error rate performance compared with the original one with substantially reduced complexity.