Research Of Grid Resource Discovery Based On Mobile Agent

The idea of grid comes from Electric Power Grid. The users only need to know what they want without concerning the position and the provider of the resources. In nature, grid is new phase of the distributed technology. Its objective is to combine Internet into a virtual very large-scale super-computer system in order to share all computing resource, storage resource, data resource, information resource, knowledge resource, equipment resource and etc. And the resources are dynamic, heterogeneous, autonomous and distributed widely. Thus the free resources will work fully and the efficiency in use will be progressed.Grid resource discovery is the base of sharing and cooperating resources. The result of discovery is a serial of resources which match the need. The success of grid computing mainly depends on the resource discovery to special task. And resource discovery is a basal service of grid computing. With the continuous increase of grid resources, it is an important research project that how to accurately and effectively discover the resources needed in very large-scale resource library.Mobile agent is mixed-blood of distributed technology and agent. Its autonomy, mobile, and cooptation very meet to the need to accurately and effectively discovery grid resources which are dynamic, heterogeneous, autonomous and distributed widely. This article introduces the concept, character and architecture of grid, and explains the interrelated technology, character of mobile agent. According to the character of mobile agent, grid environment's character: autonomy, heterogeneity and distribution wide, and routing-transfer mechanism,it constructs a grid resource discovery model based on mobile agent, and introduces the design idea, goal and whole framework of the model. The working mechanism and character is introduced in detail.Comparing with 3 existed search arithmetic, It analyzes Time-complexity and Space-complexity of the model, and proves its advantage. Compare with globus's MDS, analyze character of the model, and judge wholly. Then it has a experiment, and analyzes the result data, proves the model's feasibility.At last, this article sums up the character of the model and expects the next work.