移动边缘计算中利用BPSO的任务卸载策略  

Task offloading strategy using BPSO in mobile edge computing

作　　者：汪小威[1] 林宁 胡玉平 WANG Xiao-wei;LIN Ning;HU Yu-ping(School of Information Engineering,Nanning University,Nanning 530200,China;Information Science School,Guangdong University of Finance and Economics,Guangzhou 510320,China)

机构地区：[1]南宁学院信息工程学院,广西南宁530200 [2]广东财经大学信息学院,广东广州510320

出　　处：《计算机工程与设计》2021年第12期3333-3341,共9页Computer Engineering and Design

基　　金：广东省自然科学基金项目(2016A030313717);南宁市科技局基金项目(20181189-4);广西南宁市邕宁区科技攻关基金项目(20160312A);南宁学院2019年度教授培育工程基金项目(2019JSGC12)。

摘　　要：为提高移动边缘计算任务卸载方案的性能,提出一种移动边缘计算中利用BPSO的任务卸载策略。构建三层移动边缘计算(MEC)网络架构,移动设备根据任务情况进行本地计算,或者将其卸载至边缘计算节点与云服务器;根据MEC网络中的计算模型、通信模型设计计算卸载目标,即任务最优分配、节点负载均衡,使计算任务得到及时、有序、高效的分配;利用二进制粒子群(BPSO)算法对优化目标进行求解,得到最优卸载策略,实现能量消耗最小且时延最短,系统整体负载最为均衡。实验结果表明,所提策略能量损耗最小且系统整体负载性能明显提升。To improve the performance of the mobile edge computing task offloading scheme,a task offload strategy using BPSO in mobile edge computing was proposed.A three-layer mobile edge computing(MEC)network architecture was constructed.Mobile devices performed local computing according to the task situation or unloaded them to edge computing nodes and cloud servers.According to the computing model and communication model of MEC network,the goal of computing unloading was designed,that was,task optimal allocation and node load balancing,so that the computing tasks were allocated timely,orderly and efficiently.The binary particle swarm optimization(BPSO)algorithm was used to solve the optimal unloading strategy to achieve the minimum energy consumption and the shortest time delay,and the most balanced load of the system.Experimental results show that the energy consumption of the proposed strategy is minimum and the overall load performance of the system is significantly improved.

关 键 词：任务卸载 移动边缘计算 BPSO(二进制粒子群优化算法) 系统整体负载 能量损耗 物联网 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]