Research On Extended Function And Service Selection Method Of Enterprise Service Bus

For the appearance of Semantic Web services technology and meeting the demand of personalized service selection, the research on extending and improving functions of the ESB platform has become the hot toppic in the field of enterprise application integration. The thesis presents a systematic, in-depth research on how to increase the support for Semantic Web Services operating and improve the existing function of Web service selection for the ESB platform. And emphatically discusses the method on combining the Semantic Web Services platform with ESB platform and the architecture, and the Web services selection algorithm based on QoS and its application. The main research and innovation of our job are described as follows:1) In the thesis, we make a deep improvement study of the linear weighted algorithm currently used for service selection, and propose a computing method for the QoS attribute value of Web service. The improvements adopted by our method about the linear weighted service selection alogrithm mainly included:eliminating the difference of QoS attributes value tendency; the selection of the fixed value; simplifying the standardization process, and so on. At the same time, we also present a subjective and objective hybrid assigned weight method based on the coefficient of variation algorithm, to reduce the technical complexity of setting the QoS attributes’ weight for Users. The experiment results show the algorithm we proposed is correct and feasible.2) Setting the weight of the QoS criteria increases the difficulty of using Web services selection function, at the same time is not convenient to extend the QoS model. So we proposed a Web service selection method based on Grey Relational Analysis, the most important feature of our method is no need to specify the weight and can directly evaluate Web services based on the QoS data. Finally, the comparing experiments verified that algorithm is valid to solve the QoS-based Web service selection.3) The research on Web service composition selection algorithms is the hot topic in the field of service composition. The thesis also gave a preliminary discussion and useful attempt about that. We present a Web service composition selection method based on the Dijkstra algorithm, and highlight how to model the problem of Web service composition selection as the shortest path selection problem in DAG. The experiment results show that our method is feasible.4) On the basis of the relevant research work of Semantic Web Services technology and service selection algorithms presented in the thesis, we proposed a service bus platform-EFSB, extended the support for Semantic Web Services operating and improved service select function, focused on the choice of the platform’s integration mode; the design of the platform’s communication model and conceptual model; and described the architecture and core components of the EFSB platform. In addition, the semantic annotations for traditional Web services and service matching based on semantic information were described in detail.5) Finally, we designed and implemented the application system prototype using the EFSB platform as the core, and emphasized the design and development of the service selection component basing on the relevant service selection algorithms proposed in this thesis.