Performance analysis of composite web services using Stochastic Automata Networks over IP network

The emergence and success of IP networks make the development of new services like voice, multimedia or web services possible. One of the major problem of these type of application is that IP networks is not designed to provide Quality of Service (QoS) in an end to end architecture. Thus the performance evaluation of these applications is important to guarantee a high level of QoS. We focus in this study to evaluate the performance evaluation of Web services which is the most difficult to achieve because of the complexity of their architectures. The classical mathematical formalisms such as queueing networks can not be used in this kind of problems due the synchronization events and the state space explosion. In this paper we propose to use an adequate and efficient modelling tool: the Stochastic Automata Network (SAN) to evaluate the performance of Web services architecture over IP networks. Numerical results using PEPS are given for considered performance measures.