Pruning Based Service Selection Approach Under QoS and Temporal Constraints

Abstract

Dynamic selection of the best services to execute abstract tasks of business processes is very important. Indeed, it enables to cope with complex user's requirements that require the collaboration of several more elementary services. However, with the increasing amount of candidate services of each business task that offer different QoS (Quality of Service) values, the selection of the optimal combination of services becomes a very hard task. This problem is more complex when dealing with temporal properties of business processes associated with time-dependent QoS parameters that can change according to the execution time. Unlike static QoS which have been deeply studied in the existing service selection approaches, time-dependent QoS are insufficiently taken into consideration. In this paper, we are interested in the problem of service selection to satisfy a given business process while considering temporal properties associated to time-dependent QoS. The selection approach that we propose relies on a new service pruning approach that is applied prior to our selection algorithm to reduce the number of candidate services while guaranteeing that the optimal solution still be found.