An interval-based fuzzy ranking approach for QoS uncertainty-aware service composition

Web services are a major technique for building loosely coupled distributed systems. QoS-aware service composition concerns the selection of candidate services while maximizing the overall Quality of Service (QoS). However, these approaches fail to take the uncertainty of QoS into account. The resultant QoS of composite service observed by users is therefore unstable and subject to risks of failing composition by end users. In this paper, we propose a QoS Uncertainty-aware Web Service Composition approach based on a novel QoS interval model. It performs evaluation under a trade-off between profit and stability of QoS in a fuzzy way, and selects out composite service among the numerous solutions with an overall more stable QoS. A Multi-criteria Decision Analysis (MCDA) method and improved Genetic Algorithm is introduced in this paper. Based on real-world dataset QWS, simulations validate the correctness of the interval model and demonstrate the performance of composite framework. A comparative study with traditional tentative model is also discussed briefly.