Modeling Web Service Selection for Composition as a Distributed Constraint Optimization Problem (DCOP)

Abstract

During development of a Service-oriented Application, some software pieces could be fulfilled by the connection to Web Services. A list of candidate Web Services could be obtained by making use of any service discovery registry, which are then selected and integrated into the application. However, when it comes to a distributed system, multiple functional and non-functional constraints arise from the interaction between several service requesters and providers, particularly when composing different services. To overcome with such constraints, in this work we propose to model service selection and composition scenarios as Distributed Constraints Optimization Problems (DCOP).We propose different modeling approaches and develop representative examples to be solved through different DCOP algorithms. Also, we analyze the impact of possible extensions to the model in the computability of the problem.