Reinforcement Learning-Based Beam Selection in mmWave Networks Using Madrid Grid

In this paper, we propose a beam selection method using reinforcement learning for mmWave networks. The proposed method is evaluated within a realistic simulation environment based on Madrid Grid model. By leveraging reinforcement learning, base stations can autonomously select optimal beams to maximize data rates, overcoming challenges associated with beamforming in mmWave frequencies. Simulation results demonstrate that the AI-based approach effectively improves network performance, increasing both coverage and capacity in dynamic environments.