文章基本信息

标题：Modeling Server Load Balance in Cloud Clusters Based on Multi-Objective Particle Swarm Optimization
本地全文：下载
作者：Cao Lijun ; Liu Xiyin
期刊名称：International Journal of Grid and Distributed Computing
印刷版ISSN：2005-4262
出版年度：2015
卷号：8
期号：3
页码：87-96
DOI：10.14257/ijgdc.2015.8.3.09
出版社：SERSC
摘要：Load balancing is one of the hotspots in cloud computing research. Typically, the objective of workload balance in cloud environment is to assign tasks to proper virtual machines and physical servers with consideration of computing capability and communication cost. In this work, we propose to leverage PSO based algorithm to dynamically balance the workload of physical cloud servers. The problem is formulated as an optimization of the best solution of task assignment with the objectives of minimizing the average workload of all servers in cloud clusters, the deviation of the workload, and the migration cost between servers. Moreover, in order to avoid the low diversity of particles searching and low convergence speed, we employ a multi-swarm PSO method and introduce communication between swarms by random restructuring. Our extensive experiments show that our modified PSO method is efficient in balancing the workload of cloud servers.
关键词：Cloud computing; Load balancing; Particle Swarm Optimization (PSO)