Spectrum Efficiency Optimization for SWIPT-Enabled Hybrid Precoding in mm-Wave Massive MIMO-NOMA Systems

Abstract

Enhancing spectrum efficiency (SE) remains central to the evolution of 5G millimeter-wave (mm-wave) massive MIMO networks. This paper presents an integrated approach combining hybrid precoding and non-orthogonal multiple access (NOMA) for simultaneous information and power transfer. A three-stage framework is proposed: (1) a Dynamic Cluster-Head Selection algorithm that forms low-correlation user groups to reduce inter-beam interference; (2) a Compressed Sensing–Driven Committee Machine Design (CS-COMADE) that strengthens array gain and suppresses inter-user interference; and (3) an iterative joint power-allocation procedure that maximizes SE under energy and quality-of-service constraints. Simulation results show that the proposed method attains near-digital precoding performance—within 7% of fully digital zero-forcing benchmarks—while outperforming conventional MIMO-OMA and existing hybrid NOMA designs by up to 35% in SE. The proposed framework demonstrates strong scalability, offering a practical pathway for high-capacity, energy-aware mm-wave implementations in future 5G and beyond-5G systems.