Dynamic Pilot Design and Channel Estimation Based on Structured Compressive Sensing for Uplink SCMA System

Sparse Code Multiple Access (SCMA) is expected to accommodate massive machine-type communications (mMTC) in 5G wireless networks. Since the overloading system creates enormous signaling overheads, massive connections with grant-free transmission methodology have received significant attention. In this paper, we study active user detection (AUD) and channel estimation (CE) based on compressed sensing technology in the uplink of a grant-free system. We firstly propose a pilot design scheme considering the optimization of sensing matrix, and then a dynamic sensing matrix-based Group Orthogonal Matching Pursuit (DSM-based GOMP) algorithm is proposed for block sparse channel estimation, and hence pilot overhead in the cellular network can realize self-adaptation with the number of potential users or communication channel states. In low SNR scenarios, the sensing matrix composed of Zadoff-Chu (ZC) sequence is considered. When the SNR exceeds the threshold, the sensing matrix is constructed by optimizing Gram matrix to reduce inter-cell interference. Simulation results prove that the proposed algorithm is capable of achieving multiple access with low detection error, and adjust pilot resource overhead adaptively.