人工蜂群算法在云计算系统中的应用  

Application of Artificial Bee Colony Algorithm in Cloud Computing System

作　　者：倪中华[1] NI Zhong-hua(Department of Computer Science,Shangqiu Polytechnic College,Shangqiu 476000,China)

机构地区：[1]商丘职业技术学院工程技术学院,河南商丘476000

出　　处：《数学的实践与认识》2020年第24期275-284,共10页Mathematics in Practice and Theory

基　　金：河南省科技攻关项目(142102210231)。

摘　　要：在云计算系统中,现有的研究工作无法解决服务器的故障率问题,并且没有提供故障恢复机制.针对云环境,提出了一种基于优先权的采用人工蜂群优化算法的资源分配与调度技术.首先,侦察蜂通过监视过去的资源利用率和所分配虚拟存储器的大小来预测服务器的工作负载和用户的资源需求.利用这个预测的工作负载,可以估计每个服务器的预期完成时间.然后根据期限和资源需求对请求资源的任务进行分类,再根据工作负载和预期完成时间对服务器进行分类,将每个任务分配给相应的服务器类别.最后该方法在Java的CloudSim环境中实现,并与现有技术在资源利用率、资源成功分配百分比、错过的截止日期百分比、服务器平均工作负载等方面进行了比较.In the Cloud computing systems,the existing works fails to address the failure rate of servers and no mechanism has been provided for fault recovery.In this paper,we propose a Priority based Resource Allocation&Scheduling Technique using Artificial Bee Colony(ABC)Optimization(PRAS-ABC)for cloud environment.Initially,the work load of server and resource requirements of users are predicted by the scout bees by monitoring the past resource utilizations and size of the allocated VM.With this predicted workload,the expected completion time of each server is estimated.Then the tasks requesting the resources are categorized based on the deadline and resource requirements.Then based on the work load and expected completed time,the servers are categorized.Then each category of task will be allocated to the respective category of servers.The proposed approach is implemented in CloudSim environment of Java and compared with existing techniques in terms of resource utilization,percentage of resources successfully allocated,percentage of missed deadlines,average work load of server etc.

关 键 词：云计算系统 人工蜂群算法 资源分配 虚拟机 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] TP3[自动化与计算机技术—控制科学与工程]