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  

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

作　　者：方娟 叶志远 张梦媛 史佳眉 滕自怡 FANG Juan;YE Zhiyuan;ZHANG Mengyuan;SHI Jiamei;TENG Ziyi(Faculty of Information Technology,Beijing University of Technology,Beijing 100124,China)

机构地区：[1]北京工业大学信息学部,北京100124

出　　处：《物联网学报》2022年第1期91-100,共10页Chinese Journal on Internet of Things

基　　金：国家自然科学基金资助项目(No.61202076);北京市自然科学基金资助项目(No.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。With the development of 5 G 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.

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 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]