缓存辅助移动边缘计算的任务卸载与资源分配联合优化策略  

Joint Optimization Strategy of Task Offloading and Resource Allocation for Cache-assisted Mobile Edge Computing

作　　者：赵婵婵[1] 郭晓敏 海晓伟 李晨浩 韩国英 武文红[1] ZHAO Chan-chan;GUO Xiao-min;HAI Xiao-wei;LI Chen-hao;HAN Guo-ying;WU Wen-hong(College of Information Science and Engineering,Inner Mongolia University of Technology,Hohhot 010080,China;College of Economics and Management,Inner Mongolia University of Technology,Hohhot 010051,China)

机构地区：[1]内蒙古工业大学信息工程学院,呼和浩特010080 [2]内蒙古工业大学经济管理学院,呼和浩特010051

出　　处：《科学技术与工程》2023年第9期3812-3819,共8页Science Technology and Engineering

基　　金：内蒙古自治区高等学校科学研究项目(NJZY22382,NJZY22374);内蒙古工业大学科学研究项目(BS201936)。

摘　　要：为了减少资源受限的移动边缘计算场景下任务卸载和资源分配过程中的能量消耗,提出缓存辅助的动态卸载决策和计算、通信、缓存多维资源分配的联合优化策略。该策略根据任务流行度制定缓存服务,通过控制用户设备的发射功率优化通信资源分配,并结合计算卸载合理利用服务器的计算资源。提出最小化时延和能耗的均衡优化目标,设计基于深度强化学习的优化求解算法。最后,通过仿真实验验证所提策略的有效性,结果表明该策略在计算资源和缓存容量约束条件下能展现较优性能。In order to reduce energy consumption when tasks offloading and resource allocation in resource-constrained mobile edge computing scenarios,a cache-assisted dynamic offload decision-making and resource allocation joint optimization strategy for multidimensional resources including computing,communication,and caching resources was proposed.The caching service was formulated according to the popularity of the task,the communication resource allocation was optimized by controlling the transmission power of the user equipment,and the computing resources of the server were reasonably utilized in combination with the computing offload.A balanced optimization goal of minimizing delay and energy consumption was proposed,and an optimization algorithm based on deep reinforcement learning was designed.Finally,the effectiveness of this strategy was verified by simulation experiments.The results show that the strategy can exhibit better performance under the constraints of computing resources and cache capacity.

关 键 词：移动边缘计算 深度强化学习 任务卸载 缓存 资源分配 

分 类 号：TP399[自动化与计算机技术—计算机应用技术]