基于离散粒子群优化的云工作流调度  

Cloud workflow scheduling based on discrete particle swarm optimization

作　　者：杨玉丽[1,2] 彭新光[1] 黄名选[3] 边婧[1,4] 

机构地区：[1]太原理工大学计算机科学与技术学院,太原030024 [2]运城学院公共计算机教学部,山西运城044000 [3]广西教育学院,南宁530023 [4]山西职工医学院网络信息中心,太原030012

出　　处：《计算机应用研究》2014年第12期3677-3681,共5页Application Research of Computers

基　　金：国家自然科学基金资助项目(61262028);山西省留学基金资助项目(2009-28);山西省自然科学基金资助项目(2009011022-2)

摘　　要：针对云工作流调度问题面临的安全威胁,首先采用云模型量化任务与虚拟机资源的安全性,通过安全云相似度衡量用户对任务所分配虚拟机资源的安全满意程度;然后建立考虑安全性、完成时间和使用费用的云工作流调度模型,并提出基于离散粒子群优化的云工作流调度算法;最后对所提算法进行仿真实验。实验结果表明,与同类算法相比,该算法在安全效用值、完成时间、使用费用和负载均衡离差方面具有较好的性能表现。In order to solve the security threats that workflow scheduling problems faced in cloud computing environments,firstly,this paper quantified the security of tasks and virtual machine resources using the cloud model,and measured users’ satisfaction with the security of tasks assigned to the virtual resources through the security cloud similarity. Then,it defined a cloud workflow scheduling model considering security,completion time and cost constraints,and proposed a cloud workflow scheduling algorithm based on discrete particle swarm optimization. At last,it verified the proposed algorithm with the simulation experiment. Compared with other similar algorithms,simulation results show that the algorithm has better comprehensive performance in the security utility value,completion time,cost and load balance.

关 键 词：云工作流 云模型 离散粒子群算法 进化方程 

分 类 号：TP393[自动化与计算机技术—计算机应用技术] TP309[自动化与计算机技术—计算机科学与技术]