Improving Beamforming in 5G Networks Using Reinforcement Learning

This paper presents a study on training of Reinforcement Learning (RL) for balancing of beamforming in 5G network. Major challenges in giving directions in traditional beamforming collapse in dynamic surroundings: Beamforming which plays an essential role in increasing the strength and quality of a signal. RL provides a solution by allowing system to discover that actions that lead to a favorable outcome in the environment. The research process will involve first having a 5G network emulation, designing a DQN for beamforming, and finally having a comparison with traditional methods or other established algorithms. Effects that are observed on signal-to-noise ratio and throughput, and latencies further showed that RL based beamforming yield enhanced results. This paper also captures the possibilities of RL in handling aspects of 5G network and the development of wireless communication.