Optimal Distortion-Aware Multi-User Power Allocation for Massive MIMO Networks

Real-world wireless transmitter front-ends exhibit certain nonlinear behavior, e.g., signal clipping by a Power Amplifier (PA). Although many resource allocation solutions do not consider this for simplicity, it leads to inaccurate results or a reduced number of degrees of freedom, not achieving the global performance. In this work, we propose an optimal PA distortion-aware power allocation strategy in a downlink orthogonal frequency division multiplex (OFDM) based massive multiple-input multiple-output (M-MIMO) system. Assuming a soft-limiter PA model, where the transmission occurs under small-scale independent and identically distributed (i.i.d) Rayleigh fading channel, we derive the wideband signal-to-noise-and-distortion ratio (SNDR) and formulate the power allocation problem. Most interestingly, the distortion introduced by the PA leads to an SNDR-efficient operating point without explicit transmit power constraints. While the optimization problem is non-convex, we decouple it into a non-convex total power allocation problem and a convex power distribution problem among the users (UEs). We propose an alternating optimization algorithm to find the optimum solution. Our simulation results show significant sum-rate gains over existing distortion-neglecting solutions, e.g., a median 4 times increase and a median 50% increase for a 64-antenna and 512-antenna base station serving 60 users, respectively.