Scholarlink

As an essential component of the integrated air-ground-space networks, the unmanned aerial vehicle (UAV) is indispensable for the mobile edge computing (MEC) processing. In this paper, considering the highly dynamic nature of information exchange, we apply the multi-agent reinforcement learning (MARL) for the UAV-based communication networks and introduce a low-complexity K-means algorithm for the pre-cluster/partition users to improve the system performance. We devise a general reward and punishment strategy within the multi-agent reinforcement learning process to generate the optimal trajectories and promote the coverage maximization. Furthermore, we utilize NOMA technology to provide stable data rate gains and allocate more power to weaker users under optimizing energy consumption so as to maximize the overall system throughput. Simulation results demonstrate that the proposed joint optimization algorithm achieves excellent convergence performance in UAV based integrated air-ground-space 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

Joint UAV Trajectory Optimization and Task Offloading in Integrated Air-Ground Networks

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Yuanwei Zhang,
  • Jiaqi Shuai,
  • Weichang Wen,
  • Haixia Cui

摘要

As an essential component of the integrated air-ground-space networks, the unmanned aerial vehicle (UAV) is indispensable for the mobile edge computing (MEC) processing. In this paper, considering the highly dynamic nature of information exchange, we apply the multi-agent reinforcement learning (MARL) for the UAV-based communication networks and introduce a low-complexity K-means algorithm for the pre-cluster/partition users to improve the system performance. We devise a general reward and punishment strategy within the multi-agent reinforcement learning process to generate the optimal trajectories and promote the coverage maximization. Furthermore, we utilize NOMA technology to provide stable data rate gains and allocate more power to weaker users under optimizing energy consumption so as to maximize the overall system throughput. Simulation results demonstrate that the proposed joint optimization algorithm achieves excellent convergence performance in UAV based integrated air-ground-space networks.