Resource Searching Mechanism In Distributed Resource Space Model

Resource Space Model (RSM) is a new resource management model. It adapts to the next generation of intelligent network which is semantic-oriented and knowledge-oriented. Distributed Resource Space Model (DRSM) is the application of RSM on the Internet. One of the major objectives of DRSM is to share the mass heterogeneous data in a distributed model. Searching resources is a prerequisite for the sharing of resources. So, it becomes a very meaningful research to find the searching mechanism based on the features of Distributed Resource Space Model. To solve this problem, this dissertation puts its emphasis mainly on the research of the resource searching mechanism used in the practical DRSM environment.The main contributions of this paper are listed as follows:1. In the paper, the current searching methods have been researched, introduced the standard of remarking the searching method. To analyze the characteristics of Distributed Resource Space Model, the paper studies in depth the theory of Resource Space Model and Distributed Resource Space Model. Based on the characteristics of Distributed Resource Space Model, this paper proposes the Chord algorithm based on the matching of similarity and the improved Tell-flooding algorithm. In the new algorithms, the searching structure is changed and resources matching based on the resource similarity is used to search resource. Analysis shows that the optimized search methods effectively reduce the amount of search information.2. The paper proposes Searching Mechanism of DRSM based on hybrid network structure. With the merit of structured-Chord and unstructured-Gnutella, Distributed Resource Searching Model divides Distributed Resource Space Model into reasonable units, to scale down the whole model. In the unit, the model used by structured-Chord uses Chord algorithm based on the similarity matching to search resources; And between units, the structure uses the unstructured-Gnutella and an improved Tell-flooding algorithm to find resources. The efficiency of locating resource with the resource locating algorithm is highly increased by choosing nodes purposefully. In the searching process, the concept of similarity computation of resource and matching information based on resource coordinates domain are introduced, to greatly reduce the resource redundant, and faster query and more accurate search. Analysis demonstrates that the resource searching method in the paper has advantage of generating less selecting messages, cutting down the total costs on maintaining information of node and needing less time to search information.