An Efficient Service Selection Approach with Time-Dependent QoS

In advanced service-oriented computing, complex applications are usually specified as abstract business processes. The execution of these applications requires the selection of a set of services to invoke abstract business tasks. With the growing number of alternative services of each business task that differ in their QoS, the selection of the best combination of services that satisfies business process constraints and end-to-end users´ requirements becomes a complex decision problem. Current selection approaches consider only static QoS and ignore the fact that QoS values can depend on temporal properties. In this paper, we propose a novel service selection approach considering time-dependent QoS attributes. The proposed approach introduces two search space reduction mechanisms that combine QoS and temporal constraints. The application of these mechanisms improves the performance of the selection process which is demonstrated by experimental results.