Towards Dynamic Reconfiguration of a Composite Web Service: An Approach Based on QoS Prediction

Service-oriented architecture provides the ability to combine several web services in order fulfil a user-specific requirement. In dynamic environments, appearance of unforeseen events can destabilize composite service (CWS) and affect its quality. To deal with these issues, must be dynamically reconfigured. Dynamic reconfiguration may enhanced by avoiding invocation degraded predicting QoS for candidate service. this paper, we propose method based on HMM (Hidden Markov Model) states predict imminent degradation prevent partner values. PSO (Particle Swarm Optimization) SFLA (Shuffled Frog Leaping Algorithm) are used improve prediction efficiency HMM. Through extensive experiments real-world dataset, WS-Dream, results demonstrate that proposed approach achieve better accuracy. Moreover, carried out case study where revealed outperforms state-of-the-art methods terms execution time.