An ALOHA Massive Machine Type Communication System Based on Blind Source Separation

Abstract

This paper demonstrates that Blind Source Separation (BSS) can effectively replace orthogonal pilots (OP) in Coded Random Access (CRA) diversity ALOHA for massive machine-type communications. A novel scheme for integrating BSS within the ALOHA framework is proposed. The scheme comprises various components working in synergy to achieve optimal results. The problem’s complexity is first mitigated using a singular value decomposition (SVD) technique and by estimating the number of active users in each slot. Quadrature Phase-Shift Keying (QPSK) modulation is adopted, and the Constant Modulus Algorithm (CMA) is selected as the BSS method. CMA is further enhanced through a phase correction algorithm and a double gradient descent approach. Both intra-slot and intra-frame successive interference cancellation are considered, with two potential methodologies analyzed. Extensive simulation results, including a density evolution analysis, highlight the parameters influencing detection and demonstrate the performance achievable compared to OP CRA techniques.