Reliability Composite Service Recommendation Model

With the economic globalization and the rapid development of network information technology, various types of services emerge in endlessly. And the current researches of service composition are mainly focus on the description of the combination service and its coordination mechanism. It is important to study the unified management of service platform and the Qo S measure research. To address above problems, this thesis mainly does research works as following several aspects:1) In view of traditional integrated service management platform, its service registry management and service quality management are implemented based on service monitoring platform. And high processing performance requirements of monitoring platform, which is unfavorable for practical use. This thesis proposes a reliable composite service recommendation model based on third-party supervision. On the basis of the traditional service management model, this reliable composite service recommendation mode adds three function: service monitoring, composite service and backup storage. Among them, the service monitoring module is responsible for the real-time monitoring the quality of the service resources; Composite service module is responsible for the selection and combination of service test, which can reduce the complexity of the system and reduce the overall costs; backup storage mechanism is used to improve the reliability of system and reduce system risk. This thesis focuses on composite service technology research, it proposes a process driven composite and decomposition services model, and realizes the orderly combination of services.2) In order to well support the model, this thesis introduces the model of the recommended the main part of the service composition process, using DAG chart to describe the process of service composition abstraction for the current several more commonly used to study the workflow scheduling algorithm. According to the time sequence analysis, the advantages and disadvantages of several kinds of algorithms are introduced. The algorithm time fraction too much problem, a fragmentation issues in a certain extent, can reduce the time reliable RCLWS hierarchical algorithm, because the algorithm is based on the critical path algorithm, and it then will be assigned to a layer of nodes with the same depth, priority of the nodes in the critical path. The simulation experiments are done to verify that the method can effectively reduce the time pieces in the node.3) Through the analysis of two examples to illustrate the feasibility and superiority of the proposed model and algorithm in this thesis.