Deep Reinforcement Learning-Based High Data Rate Transmission Techniques for Metaverse

With the growing demand for ultra-high bandwidth in metaverse applications, the IEEE 802.11ay wireless local area network (WLAN) standard, which supports data rates of several hundred Gbps, has recently been developed. Downlink multi-user multiple-input-multiple-output (MU-MIMO) transmission is one of the key advancements introduced in the IEEE 802.11ay standard. However, to enable downlink MU-MIMO transmission, the beamforming training (BFT) process must be performed frequently, which incurs considerable latency and additional signaling overhead. In this paper, we propose a deep reinforcement learning (DRL) based BFT scheme aimed at reducing latency and failure probability of the BFT process.