通信学报 ›› 2023, Vol. 44 ›› Issue (2): 185-197.doi: 10.11959/j.issn.1000-436x.2023025

• 学术论文 • 上一篇    下一篇

基于蜂窝网络的多无人机能量消耗最优化算法研究

夏景明1,2, 刘玉风3, 谈玲4   

  1. 1 南京信息工程大学人工智能学院,江苏 南京 210044
    2 南京信息工程大学江苏省大气环境与装备技术协同创新中心,江苏 南京 210044
    3 南京信息工程大学软件学院,江苏 南京 210044
    4 南京信息工程大学计算机学院、网络空间安全学院,江苏 南京 210044
  • 修回日期:2023-01-03 出版日期:2023-02-25 发布日期:2023-02-01
  • 作者简介:夏景明(1980− ),男,江苏南京人,博士,南京信息工程大学副教授、硕士生导师,主要研究方向为物联网应用和边缘计算等
    刘玉风(1996− ),女,河北衡水人,南京信息工程大学硕士生,主要研究方向为移动边缘计算和机器学习等
    谈玲(1979− ),女,江苏宜兴人,博士,南京信息工程大学教授、硕士生导师,主要研究方向为机器学习和移动边缘计算等
  • 基金资助:
    国家重点研发计划基金资助项目(2021ZD0102100);江苏省产学研基金资助项目(BY2022459)

Research on multi-UAV energy consumption optimization algorithm for cellular-connected network

Jingming XIA1,2, Yufeng LIU3, Ling TAN4   

  1. 1 School of Artificial Intelligence, Nanjing University of Information Science and Technology, Nanjing 210044, China
    2 Jiangsu Collaborative Innovation Center of Atmospheric Environment and Equipment Technology (CICAEET), Nanjing University of Information Science and Technology, Nanjing 210044, China
    3 School of Software, Nanjing University of Information Science and Technology, Nanjing 210044, China
    4 School of Computer Science, Nanjing University of Information Science and Technology, Nanjing 210044, China
  • Revised:2023-01-03 Online:2023-02-25 Published:2023-02-01
  • Supported by:
    The National Key Research and Development Program of China(2021ZD0102100);Jiangsu Province Industry University Research Fund(BY2022459)

摘要:

在一些复杂时变环境中,地面基站(GBS)可能无法协助处理无人机的计算任务,为此研究了一种基于数字孪生(DT)技术的移动边缘计算(MEC)蜂窝网络。考虑到多无人机效率,引入多只配备MEC服务器的高空气球(HAB)协助,在此基础上提出一个所有无人机能量最小化问题,并给出一种多无人机轨迹优化和资源分配方案。应用双深度Q网络(DDQN)解决多无人机与多HAB之间的关联问题;采用连续凸逼近技术(SCA)和块坐标下降算法(BCD)对多无人机轨迹和计算资源进行联合优化。仿真实验验证了所提算法的可行性和有效性。实验结果表明,所提算法使系统能量消耗降低30%,明显优于对比算法。

关键词: 无人机, 任务卸载, 数字孪生, 双深度Q网络, 连续凸逼近

Abstract:

In complex time-varying environment, the ground base station (GBS) may not assist the UAV.Therefore, a mobile edge computing (MEC) cellular-connected network based on digital twin (DT) technology was studied.Given the efficiency of multi-UAV, multiple high-altitude balloon (HAB) equipped with MEC servers were introduced.On this basis, an energy minimization problem for all UAV was proposed, and a multi-UAV trajectory optimization and resource allocation scheme was presented to solve it.The double deep Q-network (DDQN) was applied to handle the association between multi-UAV and multi-HAB, and the multi-UAV trajectory and computing resource allocation were jointly optimized by the successive convex approximation (SCA) and the block coordinate descent (BCD).Simulation experiments verify the feasibility and effectiveness of the proposed algorithm.The system energy consumption is reduced by 30%, better than the comparison algorithms.

Key words: UAV, task unloading, digital twins, DDQN, SCA

中图分类号: 

No Suggested Reading articles found!