首页 | 本学科首页   官方微博 | 高级检索  
     检索      

一种面向工作站网络的系统负载预测方法
引用本文:李庆华,郭志鑫.一种面向工作站网络的系统负载预测方法[J].华中科技大学学报(自然科学版),2002,30(6):49-51.
作者姓名:李庆华  郭志鑫
作者单位:华中科技大学计算机科学与技术学院
基金项目:国家高性能计算基金资助项目 (0 0 3 0 1),国家高技术研究发展计划资助项目 (863 3 0 6 ZD11 0 1 6)
摘    要:提出一种应用人工智能技术方法解决基于PVM环境的负载平衡问题,采用多项负载指标表示结点的负载情况,总利用BP算法,预测未来的负载情况,解决了采用多项负载指标所带来的系统开销大的问题,在任务分配与任务执行时,应用人工智能技术预测各结点的负载情况,找到较佳的分配方案,实验结果证明,本算法较通常的算法在性能上有很大的提高。

关 键 词:工作站网络  系统  负载预测  计算机
文章编号:1671-4512(2002)06-0049-03
修稿时间:2001年11月5日

Approach to load prediction of networks in workstations
Li Qinghua Guo Zhixin Prof., College of Computer Sci. & Tech.,Huazhong Univ. of Sci. & Tech.,Wuhan ,China..Approach to load prediction of networks in workstations[J].JOURNAL OF HUAZHONG UNIVERSITY OF SCIENCE AND TECHNOLOGY.NATURE SCIENCE,2002,30(6):49-51.
Authors:Li Qinghua Guo Zhixin Prof  College of Computer Sci & Tech  Huazhong Univ of Sci & Tech  Wuhan  China
Institution:Li Qinghua Guo Zhixin Prof., College of Computer Sci. & Tech.,Huazhong Univ. of Sci. & Tech.,Wuhan 430074,China.
Abstract:This paper presented an artificial intelligence (AI) strategy to resolve load balancing problems under the PVM. Multi load indexes were used to present the load of each node in this algorithm. The BP algorithm was used to predict load of the node and solve the problems caused by the adoption of multi load indexes. The AI technique was employed to accurately predict load of each node and the better allocation for PVM tasks in their initial assignment and in execution time and achieved dynamic load balancing. The experimental results demonstrated that this algorithm performed more effectively than conventional approaches did.
Keywords:computer  networks of workstations  load balancing  load prediction
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号