Research On Key Technologies Of Web Service Composition Based On Uncertain QoS

With the increasing diversification and personalized demands of users,Web service composition technology has been developed.Web services composition based on QoS is a process which selects a set of services sequence to complete the complex functions easily, the standard of process is the user demand for Web services QoS attributes. Research on Web services composition based on uncertain QoS is more close to the actual. Because users’demands for QoS attributes is uncertain and Web service provided by service providers is uncertain in the real network environment,too.With the development of Web service, there were more and more service providers and it is not avoiding that there were some services had the same function and different non-function attributes. Which service was selected from these services to join to the process of service composition was the first problem should be solved before the composition based on uncertain QoS. Based on the research about Web service composition driven by uncertain QoS, we did the following analysis and research:Building Web service experiment environment, testing Web service performance and establishing QoS model for Web service. In the Internet, finding a large number of free Web services with the same function is not easy. So we need build Web service environment as the requirements of service composition. Then, we test Web service performance and built the a descriptive model for the non-function aspects of Web service with the international standards and the actual situation.Enhancing the ability of service selection based on uncertain QoS. In Web service selection process, using language data to describe the QoS attributes, which is not suit for specific numbers, can express the users’ real wishes. And the language-based data can not be used to calculate. To solve this problem, the dissertation gives a Web services selection algorithm using fuzzy ordered weighted averaging(WSF). The algorithm use fuzzy multiple attribute group decision theory in Web service selection field, And apply the concept of triangular fuzzy number to solve Web service selection problem based on uncertain QoS attribute (describe with language).Enhancing the robustness of Web service composition model. Web Services composition model based on robust optimization theory take the open-changing Internet environment into account, considered the effects that Web Services QoS attributes uncertainty make on the composite services results.So, the dissertation gives a robust optimization model of Web services composition based on the shortest path to enhancing the robustness of Web service composition model.Design and implementation the Web service composition Management system based on uncertain QoS. The system is divided into two main modules:SOA simulation module and Web service composition module.SOA simulation module uses ASP technology to achieve a SOA simulation system based on Web. Web service composition module research the thought of separating functional and non-functional attributes, according to ASP.Net and WF technology, design and implementation the Web service composition Management system based on uncertain QoS. The system is the practical application of an experimental.Now, Web service composition based on uncertain QoS are the hot research point, but there are also some problem in theory and methodology. The models, and the algorithm mentioned in this dissertation had the important referenced value for the current research of Web service composition. The Web service performance Data set, and the Web service composition Management system had the important academic significance and big applicative prospects.