移动边缘计算中无人机三维轨迹和计算卸载的联合优化策略研究  

Research on Joint Optimization Strategy of UAV 3D Trajectory and Computation Unloading in Mobile Edge Computing

作　　者：彭振春 王涛[1] 刘含 朱耀辉 雷文静 

机构地区：[1]上海大学通信与信息工程学院,上海200444

出　　处：《工业控制计算机》2024年第5期93-95,共3页Industrial Control Computer

基　　金：国家自然科学基金资助项目(61671011)。

摘　　要：无人机辅助的移动边缘计算逐渐成为新一代移动通信网络的研究热点。针对联合无人机三维轨迹和卸载策略优化的路径规划问题进行研究,通过联合优化无人机3D飞行动作和用户卸载方式考虑无人机能效最大化。仿真实验表明,与其他基准算法相比,提出的基于深度强化学习的联合无人机三维轨迹和卸载策略优化的路径规划算法(DRL-PPO3DTUO)收敛更快,能获取显著性能增益。Mobile edge computing assisted by unmanned aerial vehicles has gradually become a research hotspot for future mobile communication networks.The UAV coverage is maximized by jointly optimizing UAV 3D flight action and user unloading mode.In this paper,deep reinforcement learning based on path planning algorithm for the optimization of joint UAV three-dimensional trajectory and offloading strategy based on is proposed converges faster and speed of solving higher than other benchmark algorithms.Significant performance can be obtained.

关 键 词：深度强化学习 移动边缘计算 无人机 路径规划 

分 类 号：V279[航空宇航科学与技术—飞行器设计]