计算机工程与应用 ›› 2008, Vol. 44 ›› Issue (21): 138-141.DOI: 10.3778/j.issn.1002-8331.2008.21.038

• 机器学习 • 上一篇    下一篇

基于移动Agent的网格计算资源管理模型设计

曾正军1,舒万能2   

  1. 1.东华理工大学 核工程技术学院,江西 抚州 344000
    2.中南民族大学 计算机科学学院,武汉 430074
  • 收稿日期:2008-04-30 修回日期:2008-06-03 出版日期:2008-07-21 发布日期:2008-07-21
  • 通讯作者: 曾正军

Grid resource management model based on mobile agent

ZENG Zheng-jun1,SHU Wan-neng2   

  1. 1.College of Nuclear Engineering and Technology,East China Institute of Technology,Fuzhou,Jiangxi 344000,China
    2.College of Computer Science,South-Central University for Nationalities,Wuhan 430074,China
  • Received:2008-04-30 Revised:2008-06-03 Online:2008-07-21 Published:2008-07-21
  • Contact: ZENG Zheng-jun

摘要: 网格计算是当前高性能计算领域的一个研究热点,由于网格计算的资源具有分布性,异构性,自治性,动态性等特点,因而其资源管理比一般系统的资源管理具有更大的复杂度。文中提出了一种基于Agent的网格计算资源管理模型,并且采用了混合遗传模拟退火算法作为调度策略,满足了网格对调度系统可扩展性和全局最优调度的需求。实验结果证实了该方法的有效性。

关键词: 网格计算, 移动Agent, 资源管理, 任务调度, 混合遗传模拟退火算法

Abstract: Grid computing is a hot research area in distributed and parallel computing field.Resources in grid are of distributive,heterogeneous,autonomous and dynamic,thus managing the resources is complex work.This paper presents an agent-based model of the resource management and uses the hybrid genetic simulated annealing algorithm(HGSAA) in the scheduling policy,which serves the need of the grid resource management scheduler well.Simulation experiment is given to show the modeling procedures and efficiency of the HGSAA.

Key words: grid computing, mobile agent, resource management, task scheduling, hybrid genetic simulated annealing algorithm