A WMMSE Approach to Distortion-Aware Beamforming Design for Millimeter-Wave Massive MIMO Downlink Communication

Hardware impairments, such as power amplifier (PA) nonlinear distortion, present as the key source to system performance degradation in millimeter-wave (mmWave) communications. In this paper, we consider a mmWave massive MIMO downlink system with nonlinear PA at each transmit antenna and investigate the design of distortion-aware beamforming for efficient data transmission. In particular, we formulate a sumrate optimization problem that accounts for the characteristics of PA distortion. Rather than directly solving the original highly non-convex sum-rate optimization problem, we embrace the classic weighted-minimum-mean-square-error (WMMSE) framework and propose an algorithm to solve its equivalent WMMSE optimization problem for beamforming design. Numerical results confirm the effectiveness of our proposed WMMSE-based algorithm, which outperforms one existing distortion-aware beamforming algorithm and provides significant performance gain compared to conventional beamforming baselines that do not account for the knowledge of the PA distortion.