Research Of Web Service Selection Based On QoS

With the widespread use of computers in all walks of life and the use of Web servicesalso penetrating into every corner of life, Web services play an important role. With the rapiddevelopment of Web services, the requirements of the users of Web services is not limited toachieving a certain function. More and more users begin to pay attention to non-functionalrequirements, which are also called QoS properties of services, such as service uptime,reliability, and so on. At the same time, the selection of services based on QoS also developsand a large number of studies focus on the selection algorithm based QoS model.Classic Web services model does not include the QoS information. QoS attribute value ofWeb services is dynamic, such as service run time. The static property value can not reflectthe real situation of the services very well. Therefore, this paper proposes a Web serviceselection model in which the QoS values are dynamically updated. The model is called theQoSB-WSA model. In the QoSB-WSA model, QoS information processing functions and anew dynamically-updating-QoS method are included. The method is called HDB-QoS-DPMmethod. The HDB-QoS-DPM is based on Pareto’s law and statistical regularity. The methodcan make the dynamic updates by processing QoS feedback data of the services from usersand providers. The experiment shows that the Web service selection model based QoSdynamic update is of the feasibility and correctness. By comparing XM method, Bayesmethod, M-Bayes method and HDB-QoS-DPM method proposed in this paper, theHDB-QoS-DPM has better effect than other three methods in this model.