WIPT-Enabled Massive MIMO URLLC Relaying: Hardware Distortion Aware SE Analysis and Low-Complexity Optimization

We consider a multi-pair two-way massive multi-input-multi-output relaying system, where the relay is equipped with low-quality radio frequency (RF) chains and low-resolution analog-to-digital/digital-to-analog converters (ADC/DACs). The user equipments (UEs) have simultaneous wireless information and power transfer capability, and perform ultra-reliable and low latency communication (URLLC). We design novel distortion-aware minimum mean square error and regularized zero forcing precoders, which mitigate the impact of hardware impairments due to low-cost RF chains and low-resolution ADC/DACs. We employ random matrix theory to derive the deterministic equivalent expression for the spectral efficiency (SE), which is tight for a system with finite number of relay antennas and UEs. We propose a novel low-complexity optimization framework to maximize the non-convex SE-metric. Our framework provides closed-form fixed point solutions for the optimal relay and UE transmit powers, and charging time fraction. We numerically show that the i) energy harvested by the UEs increases with increasing spatial correlation; ii) proposed distortion-aware precoders are able to compensate for the SE loss due to URLLC implementation; and iii) proposed optimization framework not only has a lower complexity, but also provides close-to-optimal solution by comparing it with exhaustive grid-based search.