物联网学报 ›› 2022, Vol. 6 ›› Issue (1): 91-100.doi: 10.11959/j.issn.2096-3750.2022.00258

• 理论与技术 • 上一篇    下一篇

边云协同场景下基于强化学习的精英分层任务卸载策略研究

方娟, 叶志远, 张梦媛, 史佳眉, 滕自怡   

  1. 北京工业大学信息学部,北京 100124
  • 修回日期:2021-12-29 出版日期:2022-03-30 发布日期:2022-03-01
  • 作者简介:方娟(1973− ),女,博士,北京工业大学信息学部教授、博士生导师,中国计算机学会会员,主要研究方向为高性能计算、边缘计算、大数据技术等
    叶志远(1996− ),男,北京工业大学信息学部硕士生,主要研究方向为移动边缘计算
    张梦媛(1997− ),女,北京工业大学信息学部硕士生,主要研究方向为移动边缘计算
    史佳眉(1997− ),女,北京工业大学信息学部硕士生,主要研究方向为移动边缘计算
    滕自怡(1997− ),女,北京工业大学信息学部博士生,主要研究方向为移动边缘计算、高性能计算
  • 基金资助:
    国家自然科学基金资助项目(61202076);北京市自然科学基金资助项目(4192007)

Research on elite hierarchical task offloading strategy based on reinforcement learning in edge-cloud collaboration scenario

Juan FANG, Zhiyuan YE, Mengyuan ZHANG, Jiamei SHI, Ziyi TENG   

  1. Faculty of Information Technology, Beijing University of Technology, Beijing 100124, China
  • Revised:2021-12-29 Online:2022-03-30 Published:2022-03-01
  • Supported by:
    The National Natural Science Foundation of China(61202076);The Natural Science Foundation of Beijing(4192007)

摘要:

随着5G的发展以及应用程序功能的丰富,应用程序对终端设备的计算能力提出了更高的要求,为了提高终端设备对应用程序的计算能力,降低任务的处理时间,针对移动边缘计算环境,提出了一种边云协同的任务卸载方式,并设计了基于强化学习的精英分层进化算法(RL-EHEA, elite hierarchical evolutionary algorithm combined with reinforcement learning)进行卸载决策,使多个具有依赖关系与截止时间的任务对计算资源竞争。结果表明,与遗传算法(GA, genetic algorithm)和精英遗传算法(EGA, elite genetic algorithm)相比,RL-EHEA能缩短任务的处理时间,得到更优的资源分配策略。

关键词: 移动边缘计算, 任务卸载, 边云协同, 进化算法, 串行任务

Abstract:

With the development of 5G and the enrichment of application functions, applications have put forward higher requirements on the computing capabilities of terminal devices.In order to improve the computing capabilities of terminal devices on applications and reduce the processing time of tasks, it is aimed at mobile edge computing environments, a task offloading method for edge-cloud collaboration was proposed,and an elite hierarchical evolutionary algorithm combined with reinforcement learning (RL-EHEA) was designed to perform offloading decisions, so that multiple tasks with dependencies and deadlines compete for computing resources.The simulation experiment results show that, compared with genetic algorithm (GA) and elite genetic algorithm (EGA), RL-EHEA can shorten task processing time and obtain better resource allocation strategy.

Key words: mobile edge computing, task offloading, edge-cloud collaboration, evolutionary algorithm, serial task

中图分类号: 

No Suggested Reading articles found!