Greedy Data-Aided Active User Detection for Massive Machine Type Communications

Abstract

A grant-free access (GFA) is considered as a promising solution for realizing massive connectivity with reduced signal overhead and latency in massive machine type communications (MMTC). In GFA, users autonomously send the data packet without prior resource assignment and the system identifies the active users that have sent the data based on the received data. Compressed sensing (CS) techniques have been employed to perform active user detection (AUD) leveraging the sparse nature of traffic in MMTC. In this letter, we propose an enhanced CS-based AUD technique, which bases the detection of active users on both pilot and data symbols received. The proposed AUD algorithm successively detects the active users in a greedy manner. For the user candidates detected up to the current iteration, our method finds the joint estimate of the channel and the data symbols, which are in turn utilized in detecting new users in the next iterations. Our numerical evaluation shows that the proposed method achieves better AUD, channel estimation, and data detection performance than the existing methods.