基于蚁群算法的云计算联盟资源调度  

Resource Scheduling Method Based on Ant Colony for Cloud Computing Federation

作　　者：陈冬林[1] 姚梦迪[1] 桂雁军 陈玲[1] 

机构地区：[1]武汉理工大学经济学院,湖北武汉430070 [2]福建新奇特车业服务有限公司,福建福州350002

出　　处：《武汉理工大学学报（信息与管理工程版）》2014年第3期337-340,373,共5页Journal of Wuhan University of Technology：Information & Management Engineering

基　　金：国家自然科学基金资助项目(71172043;71072077);国家科技支撑计划基金资助项目(2013BAH13F01;2012BAH93F04);中央高校基本科研业务费专项资金资助项目(2012-IB-063;2012-IB-060);中央高校基本科研业务费专项资金资助武汉理工大学自主创新优秀博士论文培育基金资助项目(2013-YB-017)

摘　　要：针对目前云计算联盟的架构和单云环境下资源调度的研究缺少对云计算联盟下的资源调度问题的研究情况，建立了由云用户、云服务供应商和云联盟协调器组成的云计算联盟资源调度模型，为达到云供应商利益最大化，设计了任务-虚拟机-数据中心的调度算法，利用蚁群算法进行模型求解，并通过Cloudsim仿真软件证实了该算法的合理性，验证了供应商资源的数据中心负载率在60％～80％之间时达到均衡，并可获得最大利益。Currently studies on cloud computing are focused on the architecture of cloud computing federation , or on the resource scheduling method under the single cloud environment .There are few studies about resource scheduling for cloud computing federation .To solve this problem , a resource scheduling model for cloud federation was proposed .It consisted of cloud users, cloud service providers and cloud federation coordinator .Furthermore , the scheduling algorithm of tasks-visual machinedata centers was designed to maximize the profits of providers .Then the ant colony algorithm was applied for model solution .Its rationality was confirmed by Cloudsim .It was verified that the resource load rate of data centers will achieve a balance between 60%-80%, and the providers will achieve the most profit .

关 键 词：蚁群算法 云计算联盟 资源调度 供应商利益最大化 负载均衡 

分 类 号：TP316[自动化与计算机技术—计算机软件与理论]