QoS-based Selection of Multi-Granularity Web Services for the Composition

Existing methods for QoS-aware services composition only consider web services whose service class is specified in the process definition as the candidate. However, there may exist some services that could also accomplish partial goal of the web service composition but their service classes do not appear in the process definition. In this paper, we propose a new QoS-aware service composition approach, which expands the choice for services selection by allowing web services of various granularities to be available for selection. A method based on Mixed Integer Linear Programming (MILP) is proposed to solve the QoS-based Multi-Granularity Service Selection Problem (QMGSSP), which optimizes the user-defined objective and meets the end-to-end QoS constraints as well. Experiments show the effectiveness of our approach.