网格计算中基于改进蚂蚁算法的任务调度研究  

Research on Task Scheduling Based on Improved Ant Algorithm in Grid Computing

作　　者：林晓娴[1] 王维欢[1] 

机构地区：[1]西北师范大学数学与信息科学学院,甘肃兰州730070

出　　处：《计算机技术与发展》2011年第6期24-27,31,共5页Computer Technology and Development

基　　金：甘肃省科学技术攻关计划项目(2GS047-A52-002-04)

摘　　要：文中提出了一种以蚂蚁算法为基础的改进算法,用以解决网格环境下的任务调度问题。首先从蚂蚁算法的基本思想出发,结合网格环境下任务调度的特点,逐步改进了资源信息素的初始化处理、局部更新及全局更新方式,并针对网格环境下的任务给出信息素的概念;然后,综合考虑资源信息素和任务信息素两方面的需求,提出了一种新的任务选择资源机制;最终,提出了一种基于蚂蚁算法的、改进的网格任务调度方法。通过仿真实验的结果分析表明:基于蚂蚁算法的、改进的网格任务调度方法实现了网格环境下任务的有效调度问题,并使系统获得较好的负载平衡度。An improved ant algorithm has been presented in order to solve the problem of task scheduling in the grid environment.Firstly,a new style of the initialization processing,the local update and the global update of resource＇s pheromone has been introduced which based on the ant algorithm and the task scheduling in grid environment.And,the conception of pheromone for the task in grid environment has also been given.Secondly,a new selective mechanism of task resource has been showned by considering the pheromone of resource and task synthetically.Finally,an improved ant algorithm for the task scheduling in the grid environment has been obtained.It indicates from the simulation experiments that the effective scheduling of tasks and the load balancing of system has been implemented by the improved ant algorithm of the task scheduling in grid environment.

关 键 词：改进的蚂蚁算法 任务调度 网格计算 

分 类 号：TP393.01[自动化与计算机技术—计算机应用技术]