Game Theoretic Based Intelligent Multi-User Millimeter-Wave MIMO Systems under Uncertain Environment and Unknown Interference

This paper investigates the intelligent development problem for of multi-user millimeter-wave (mmWave) MIMO systems under uncertain environment and unknown interference. To effectively increase the real-time performance of millimeter-wave (mmWave) MIMO system, a series of advanced communication techniques have been proposed recently, such as Hybrid precoding and MIMO relay, which integrated analog and digital schemes to reduce the hardware complexity of the conventional fully-digital beamforming and further maximizing the capacity of the MIMO channel through introducing relays and formulating multi-hop system. However, in practice harsh conditions such as uncertainty environment like rain, wind, snow, users’ movement, and the unknown interference, would seriously affect the effectiveness and practicality of those emerging communication techniques. This paper proposes a Game Theoretic Based Intelligent Multi-User Millimeter-Wave MIMO (GT-MU-MIMO) system that includes a novel dynamic code book based beam training protocol and an online reinforcement learning algorithm supervising the mobility of multi-robot-relay as well as handling the serious effects from the uncertain environment and unknown interference. Firstly, a novel dynamic codebook development has been introduced to lower the complexity during multi-user beam training. Then, a decentralized game theoretic deep reinforcement learning based intelligent algorithm has been developed. It will not only optimize GT-MU-MIMO beam training efficiency and managing mobility of multi-robot-relay online, and also effectively handle uncertainty while user moving and avoiding the signal interference from unknown radio jamming attack, etc. The effectiveness of proposed design has been demonstrated through computer aided simulation.