Sparsity-Aware Ordered Successive Interference Cancellation for Massive Machine-Type Communications

Abstract

In massive machine-type communication systems, by exploiting sporadic device activities, compressed sensing-based multi-user detection (CS-MUD) is used to recover sparse multi-user vectors. In CS-MUD, multi-user vectors are detected based on a sparsity-aware maximum a posteriori probability (S-MAP) criterion. To reduce the computational complexity of S-MAP detection, a sparsity-aware successive interference cancellation (SA-SIC) technique can be used. However, SA-SIC does not perform well without proper layer sorting due to error propagation. In this letter, we propose a novel sparsity-aware ordered SIC scheme that finds the optimal detection order based on the activity probabilities and channel gains of devices. Simulation results verify that the proposed scheme greatly improves the performance of SA-SIC.