Grid Resource Discovery Algorithm Based On Relation Grid And Optimized-stop Theory

Grid is a completely shared and cooperated environment. It's a feasible and effective solution to the problem of sharing network resources in huge and complex computing devices. The highly distributed of grid, the isomerous, dynamic, diversity and autonomy behaviors of grid resources bring great challenge to the problem of resource discovery.This paper proposes solution to the problem on the base of relation grid and optimal stop theory. We introduce small-world model into structure model of resource discovery, and construct society relation grid in the foundation. We introduce optimal stop theory in resource discovery algorithm to discover resource fast and ensuring the quality of resources.This paper also presents methods of user classifying and data retrieval. The former resource discovery algorithms only consider the resources satisfying condition and rarely consider the quality of resources themselves. RGOSRD (Resource Discovery Based on Relation Grid and Optimized-Stop) algorithm is proposed to improve the quality of resources. It involves the attribute of the nodes and the cost. The cost between two nodes indicates the osculation relation of them. The introduce of cost improves the efficiency of resource discovery.The result in emulator experiment indicates that RGOSRD algorithm can greatly condense the search space. It not only improves the efficiency of resources discovery, but also increases the general value of achieved resource.