基于改进粒子群优化的移动云应用卸载决策  

Mobile cloud application offloading decision-making based on improved particle swarm optimization

作　　者：李廷元[1] LI Ting-yuan(School of Computer,Civil Aviation Flight University of China,Guanghan 618307,China)

机构地区：[1]中国民用航空飞行学院计算机学院,四川广汉618307

出　　处：《计算机工程与设计》2020年第9期2543-2551,共9页Computer Engineering and Design

基　　金：国家民航局科技创新引导基金项目(MHRD20140214);民航局和国家自然基金委民航联合基金项目(U1333113)。

摘　　要：为优化移动云计算中应用执行延时和移动设备能效,提出基于粒子群优化的移动云应用卸载决策算法。利用预定义和随机粒子群方法混合生成粒子群的初始种群;为预定义的预留粒子群设计一种基于汉明距离函数的适应度函数,更好衡量粒子差异;为丰富种群多样性,利用杂交繁育生成变异粒子。改进粒子进化操作可以更合理地获取应用分割的最优解。仿真结果表明,改进算法在能耗、效率及综合权重代价指标上均表现更好。To solve the application execution delay problem and to improve the energy-efficiency of mobile devices,a mobile cloud application offloading decision based on particle swarm optimization was presented.The hybrid method of predefined and random chromosomes were used to generate initial population of particle swarm.A specific fitness function based on Hamming distance function was designed for the predefined reserved population,which better measured the difference between particles.The crossbreeding was used to generate the population crossover to enrich individual species.The improved particle evolutionary operations could obtain the best solution of application division reasonably.The simulation conclusions show that the proposed algorithm preforms better on the energy consumption,the efficiency and the overall weight cost.

关 键 词：移动云 粒子群优化 应用卸载 延时 汉明距离 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]