Scholarlink

The rapid evolution of 5G networks requires new solutions to manage user mobility, changing environments, and high data needs. This paper introduces a reinforcement learning (RL) beam selection framework for 5G millimeter-wave (mmWave) networks to improve signal quality and network performance. Using the Q-learning algorithm, the system selects optimal beam directions and handles handovers in real time based on Signal-to-Interference-plus-Noise Ratio (SINR), user mobility, and positional feedback. The framework was trained and tested in a MATLAB-based simulated 5G mmWave environment that included realistic conditions like blockage and movement. Simulation results indicate an 18% improvement in SINR and a 25% reduction in handover latency when compared to static beam selection methods. These results show the effectiveness and scalability of RL-based beam management for future wireless networks.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Reinforcement Learning Based Adaptive Beam Selection for 5G mmWave Networks

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Nikita P. Girimath,
  • Ananya G. Patil,
  • Prajwal P. Nayak,
  • Megha T. Khoday,
  • M. R. Kiran,
  • Suneeta V. Budihal

摘要

The rapid evolution of 5G networks requires new solutions to manage user mobility, changing environments, and high data needs. This paper introduces a reinforcement learning (RL) beam selection framework for 5G millimeter-wave (mmWave) networks to improve signal quality and network performance. Using the Q-learning algorithm, the system selects optimal beam directions and handles handovers in real time based on Signal-to-Interference-plus-Noise Ratio (SINR), user mobility, and positional feedback. The framework was trained and tested in a MATLAB-based simulated 5G mmWave environment that included realistic conditions like blockage and movement. Simulation results indicate an 18% improvement in SINR and a 25% reduction in handover latency when compared to static beam selection methods. These results show the effectiveness and scalability of RL-based beam management for future wireless networks.