Applying Particle Swarm Optimization to Quality-of-Service-Driven Web Service Composition

Web service composition is a very important task in service-oriented environments. The composition of services has to be based not only on functional but also on non-functional properties. In particular, the selection of the services should be performed at run-time rather than at design-time in order to adjust to changes in the environment that are due to the volatile nature of service-oriented environments. An optimization technique is necessary to perform the composition of web services based on Quality of service (QoS) parameters. Many different methods have been used to address the service composition problem, in particular, many linear programming methods have been used to optimize the process of service composition. The proposed work is different from related work in two aspects: (1) two meta-heuristic methods based on Particle Swarm Optimization (PSO) are introduced to address the optimization problem, and (2) several workflow requests are being processed simultaneously. The experimental results show that the hybrid PSO method in particular performs very well in service-oriented environments.