Research On Distributed Multilevel Indexing Model For Decentralized Service Repositories

In recent years,with the rapid development of information technology,Cloud Computing,Internet of Things(Io T)and Peer-to-Peer(P2P)have been in further study and research.The application of these network technologies in various dimensions leads to an easier and more convenient service provision concepts.Thus,Service Computing obtains a richer development space.Service-Oriented Architecture methodology has been widely applied in different enterprises.Efficient service retrieval and composition and accurate service selection are two important factors.Then how to effectively retrieve service in distributed environment becomes current research hotspot.In this study,the foundational knowledge of Service Computing and Peer-to-Peer network are introduced and studied.Based on these existing theories,this study proposed a distributed index service repository model and distributed resource discovery algorithms.The main works and are as follows:1.A multilevel index service repository model,based on the theory of equivalence relations,is proposed in this study.This model not only improves the efficiencies of service retrieval and composition extremely,but also eliminates the redundancies at the maximum extent.Based on set computation concepts,the multilevel index service repository provides some basic service retrieval computing operations,including service retrieval,service addition,service deletion and service replacement.The service retrieval computing operations of multilevel index service repository model have been adapted to meeting the requirement of distributed service environment.Services are stored in various nodes in distributed system.The improved service retrieval computing operations have more powerful function,larger capacity and flexible distribution in different locations in distributed service environment.2.An efficient resource discovery algorithm based on Discrete Hash Tables(DHTs),is presented to enable efficient and effective retrieval services among different distributed repositories.Further researches have been carried out in different resource discovery algorithms and compared them with each other individually.By means of P2 P algorithms,storing structure of every node in distributed system has been turned to distributed repository.This improvement eliminates redundant information and enhances the time-management quality of the service retrieval process of the service repository architecture.Furthermore,according to the scale of service repository,the distributed multilevel index service repository model enables flexible deployment.3.Since plenty of factors in distributed service environment may have influence on the proposed model,some mainly influencing factors and conditions have been taken into consideration to compare distributed multilevel service repository model with inverted model in the form of experiment in this study,in order to prove the correctness and performance of multilevel index service repository model.The performance of the proposed model and the supporting algorithms have been evaluated in different distributed environment.Experimental results validate the effectiveness of our proposed indexing model and search algorithm.