Scholarlink

<p>The millimeter-wave integrated access and backhaul networks (mmWave IABN) is viewed as a potential technology for B5G systems and has been received widespread attention in the industry. Due to the challenge posed by the shared wireless resources between the access and backhaul parts in the mmWave IABN, this paper investigates the power allocation problem of the concurrent transmission links in the considered mmWave IABN to maximize the energy efficiency (EE) of the systems. Firstly, we formulate the power allocation problem of concurrent links including access and backhaul links in mmWave IABN as a non-convex optimization problem under the full consideration of the user’s transmission rate requirements. Secondly, we further transform the original problem into a Markov decision process through the Markov characteristic of the constraints in the formulated problem. Thirdly, a deep reinforcement learning (DRL) based transmission power controlling framework of the concurrent links is proposed, where the optimal transmission power of each concurrent link is predicted through the deep Q-network (DQN) for higher EE. Finally, it is demonstrated that the proposed power controlling framework is superior to the other reference algorithms in terms of EE and convergence when it is compared with these baselines in diversified scenarios through affluent simulations. </p>

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

Energy-Efficient Power Controlling for mmWave Integrated Access and Backhaul Network: Deep Reinforcement Learning Based Designment

  • 期刊论文

  • 发表日期：

    2025年7月10日

查看全文
  • Zhongyu Ma,
  • Yajuan Gao,
  • Jing Li,
  • Qun Guo,
  • Yingting Liu

摘要

The millimeter-wave integrated access and backhaul networks (mmWave IABN) is viewed as a potential technology for B5G systems and has been received widespread attention in the industry. Due to the challenge posed by the shared wireless resources between the access and backhaul parts in the mmWave IABN, this paper investigates the power allocation problem of the concurrent transmission links in the considered mmWave IABN to maximize the energy efficiency (EE) of the systems. Firstly, we formulate the power allocation problem of concurrent links including access and backhaul links in mmWave IABN as a non-convex optimization problem under the full consideration of the user’s transmission rate requirements. Secondly, we further transform the original problem into a Markov decision process through the Markov characteristic of the constraints in the formulated problem. Thirdly, a deep reinforcement learning (DRL) based transmission power controlling framework of the concurrent links is proposed, where the optimal transmission power of each concurrent link is predicted through the deep Q-network (DQN) for higher EE. Finally, it is demonstrated that the proposed power controlling framework is superior to the other reference algorithms in terms of EE and convergence when it is compared with these baselines in diversified scenarios through affluent simulations.