Research On Grid Resource Discovery Based On Fuzzy Clustering

Grid is a hotspot in distributed computing field at present. The real and specific problem that underlies the grid concept is coordinated resource sharing and problem solving in dynamic, multi-institutional virtual organizations. Grid resource discovery is the precondition of grid resources sharing which is the core of grid research. With the goal of improving the efficiency of resource discovery, the resource discovery method is researched thoroughly, and the problem is discussed deeply ground works of states.Firstly, aimed at resource management model of poor adaptability and distributed resource management model of low efficiency, this paper puts forward a resource reorganization model based on administrative domains according to some features of resources in grid environment, such as geographical dispersion, dynamic change, and heterogeneous system. Using the distribution and extension of P2P network and the sort idea in the economy administrator, we establish a layered model partitioned the grid resource organization system into P2P layer, resource region layer and resource layer, and transformed the grid resource as logical node according to the resources classification.Secondly, resource consolidated description is the foundation of sharing because of the resource heterogeneous. Unifying description is the foundation of sharing the grid resource.Unifying resource information could not only help understand the resource attributes and related information but also facilitate computer to dispose and search the grid resource.Finally, we provide a simulation system using the algorithm we proposed with the GridSim toolkit at the multi-layer overlay network resource organization model. And we do some analysis with the tradition flooding resource discovery algorithm in order to verify the validity and useableness of the algorithm.