摘要
服务选择是网格工作流的重要环节。针对现有W eb服务组合方法对服务质量(QoS)贪婪索取而不利于网格资源有效利用的问题,提出基于商品市场的"按需"网格服务选择方法:根据面向工作流全局的评价模型,以服务使用者需求为约束条件,将服务选择建模为0-1多维背包问题;并引入经济学中的"价格—供需关系"改进了服务选择模型。仿真实验表明:基于商品市场的服务选择均衡了系统负载,有利于网格资源的高效利用。
Services selection is important in Grid workflow, To deal with the inefficient resources utilization brought by greedy competition for QoS in current Web services composition methods, an "on-demand" Grid services selection method based on commodity market was put forward. Based on global evaluation model for Grid workflow, services selection was modeled as a 0-1 multi-dlmensional knapsack problem by transforming the service requester's requirements into constraints. Furthermore, the selection model was improved with economic "price-demand relationship". Finally, experiments show that the proposed services selection method can bring load balance, which is helpful to the utilization of Grid resources.
出处
《计算机应用》
CSCD
北大核心
2007年第12期2973-2975,共3页
journal of Computer Applications
基金
国家"十五"
"211工程"公共服务体系建设资助项目(219899004)
关键词
服务选择
商品市场
网格工作流
0-1多维背包问题
services selection
commodlty-market
grid workflow
0-1 multi-dimensional knapsack problem