移动边缘计算网络中基于DQN的能效性卸载决策及无线资源分配  

Energy-Efficiency Offloading Decision and Wireless Resource Allocation Based on DQN in Mobile Edge Computing Networks

作　　者：高云 郭艳艳[1] 卫霞[1] GAO Yun;GUO Yanyan;WEI Xia(College of Physics and Electronic Engineering, Shanxi University, Taiyuan 030006, China)

机构地区：[1]山西大学物理电子工程学院,山西太原030006

出　　处：《测试技术学报》2022年第1期54-59,共6页Journal of Test and Measurement Technology

摘　　要：本文研究多种移动设备、单个移动边缘计算(Mobile Edge Computing,MEC)服务器网络场景下,移动设备计算任务的动态卸载决策及资源分配问题.在移动设备的计算任务队列稳定与时延限制、及最大发射功率约束等条件下,建立以系统长期平均能耗最小化为优化目标,实现任务卸载决策、计算资源分配、无线信道及发射功率分配的优化模型.根据优化参数之间的相关性,将模型简化为信道和功率分配的联合优化问题.利用深度Q学习(Deep Q-network,DQN)方法,实现功率与时延约束下计算任务队列长期稳定的资源分配算法.仿真结果表明,本文所提出的算法能够有效地提高系统的能效和数据处理率.In this paper,the dynamic uninstallation decision and resource allocation of Mobile Computing tasks were studied in the scenario of multiple Mobile devices and a single Mobile Edge Computing(MEC)server.Under the conditions of task queue stability,delay limitation and maximum transmitting power constraint of mobile devices,the optimization model of task unloading decision,computing resource allocation,wireless channel and transmitting power allocation was established to minimize the long-term average energy consumption of the system.According to the correlation of optimization parameters,the model was simplified to a joint optimization problem of channel and power allocation.Then,the deep Q-network(DQN)method was used to realize a resource allocation algorithm for long-term stability of the computing task queue under the constraints of power and delay.Simulation results show that the proposed algorithm can effectively improve the energy efficiency and improve the data processing ratio of the system.

关 键 词：移动边缘计算 资源分配 DQN 能效性 

分 类 号：TN92[电子电信—通信与信息系统]