Hybrid Beamforming for mmWave MU-MISO Systems Exploiting Multi-Agent Deep Reinforcement Learning

Abstract

In this letter, we investigate the hybrid beamforming based on deep\nreinforcement learning (DRL) for millimeter Wave (mmWave) multi-user (MU)\nmultiple-input-single-output (MISO) system. A multi-agent DRL method is\nproposed to solve the exploration efficiency problem in DRL. In the proposed\nmethod, prioritized replay buffer and more informative reward are applied to\naccelerate the convergence. Simulation results show that the proposed\narchitecture achieves higher spectral efficiency and less time consumption than\nthe benchmarks, thus is more suitable for practical applications.\n