Design And Implementation Of QoS-Based Service Selection Mechanism

In recent years, the concept of service has been proposed as a mean of abstracting and wrapping diverse resources in open environment, which makes service composition a natural way for dynamic resource aggregation and flexible application integration. In order to optimize the QoS of the composition, the research on the QoS (quality of service) of composite service focuses on the QoS model, QoS evaluation and QoS-aware selection algorithm,. We argue that the users should always be at the center in the process of the application of the service composition techniuqes. Unfortunately, it is not the situation somehow today. We present a framework for optimizing service composition based on scenario, which will optimize the whole service compositon and improve user-aware QoS and user satisfiction degree.In order to improve the QoS of the whole web services composition process, a framework based on scenario is designed. QoS manager, the core component of the framework, is implemented. The contributions of this dissertation are as follows:1. A framework for optimizing web services composition based on scenario is designed. The framework make use of the process repository to store and manage the abstract plan, and reuse the abstract plan to improve the efficiency of the web service composition. It also uses the execution plan repository to maintain the execution plan, which is personalized composite service for the special users.2. The service selection algorithm is selected based on user scenario and preference, which make the selected one satisfying user requirements better. A heuristic algorithm, TP_HEU, is presented, which is fast for its response time and user preference for service composition in terms of execution time., TP_HEU assures the optimized the time of QoS, and can find the (sub)-optimal service composition in a fast way.3. The QoS manager is implemented, which is the core component of the farmework proposed. Firstly, it executes the service matching, and selects the service selection algorithm based on the user's scenario and the preference. Secondly, it selects the component service based on the QoS, and binds the component service to the abstract plan.