Research And Design Of A Composite Web Service QoS Simulation Tool Based On OWL-S

With the development of Web Service, composite Web Service has been widely employed. However, composite Web Service may be composed of different levels sub-services from the internet, the unclearly details of remote Web Service and unstable internet environment make it is hard to maintain the performance of service. In some industries such as Military, Financial industry who have very strict demand on the Quality of Service(QoS), a service failure may contribute to a immeasurable demage. Therefore, it has been a hot spot in current research that to improve the QoS of composite Web Services. In contrast to traditional Fault-tolerant software system, relationship between Web Services is loosely coupled. Moreover, there are many accessible redundant Web Services on the Internet. The Fault-tolerant mechanism is a major approach to achieve high quality of service.In this paper, we make a study of the analysis of composite Web Service QoS simulation. Considering the existing Composite Web Service Model based on Markov chain model cannot describe the parallel structure in composite Web Service, we presented a Binary Tree Model named OWL-CDT to describe a composite Web Service and a QoS Model which contains Operation profile, reliability, availability, response time and Fault-tolerant strategy for analyzing QoS of Web Services. Based on OWL-CDT and QoS Model, we designed a Composite Web Service QoS Simulation Tool based on OWL-S, a ontology language which is popular in growing Semantic Web. Users could transform a composite Web Service Architecture which is described by OWL-S or user defined on Custom Architecture Panel of this simulation tool into a OWL-CDT, and analyze QoS of this composite Web Service based on simulation.There are two kinds of Fault-tolerant Strategies: time redundancy strategy such as the famous fault-tolerant strategy Retry Strategy and space redundancy strategy. In this paper, we presented six kinds of fault-tolerant strategy based on the combination of traditional fault-tolerant strategy, and make research on the effects on composite Web Service QoS by deploying fault-tolerant strategy. Moreover, in order to verify the effectiveness of Simulation Tool designed in this paper, we designed a composite Web Service named EC2 Financial Management Service(EFMS) and deployed it to Amazon Elastic Compute Cloud(Amazon EC2). By comparing the QoS test results of EFMS in real internet and simulation results, we can verify the effectiveness of Composite Web Service QoS Simulation Tool designed in this paper.