基于信任模型的数据密集型网格任务调度算法研究  

On Data-intensive Task Scheduling Algorithm in the Grid Environment Based on Trust Model

作　　者：林剑柠[1] 吴慧中[1] 

机构地区：[1]南京理工大学计算机系,江苏南京210094

出　　处：《信息与控制》2006年第5期667-672,共6页Information and Control

基　　金：国防预研课题基金资助项目(51404020303BQ0220);江苏省博士后基金资助项目(KR20003)

摘　　要：针对数据密集型应用的调度问题,提出一种新的调度算法,在选择文件传输节点的同时考虑网络带宽和节点的信任度.针对传输文件时带来的传输节点负载不均的现象,采用基于sufferage思想的算法均衡负载.最后,通过实验证明该算法优于传统的M in-M in算法.*For the scheduling problems in data-intensive application, a new scheduling algorithm is proposed, which considers both the bandwidth and the trust level of the source sites when choosing the grid nodes. As for the unbalancing problems among many sites when transferring files, an algorithm based on the idea of sufferage is addressed to balance the download loading. The simulation experiment indicates that this algorithm is better than the traditional Min-Min algorithm.

关 键 词：调度 网格 数据密集型 信用模型 均衡负载 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]