Comparison of Optimization Techniques and Machine Learning Methods for Optimized Beamforming in Wireless Networks

In today’s wireless networks, particularly in the context of 5G and upcoming 6G networks, parameter optimization using machine learning are crucial. Classical optimization techniques aim to enhance efficiency, while machine learning contributes to predictive capabilities, both critical for the success of next-generation networks. This article explores the utilization of ray tracing to optimize beamforming in wireless communication systems. Specifically, it employs two optimization methods and two classical reinforcement learning approaches to determine the optimal beamforming angles for transmitter and receiver antennas. The goal is to maximize total channel capacity while minimizing interference. By comparing these four methods in terms of accuracy, efficiency, and execution time, the study aims to identify the most effective approach. This analysis provides information on the strengths and weaknesses of each method, supporting future developments in wireless network optimization.