Service Composition in Cyber-Physical-Social Systems

Abstract

Cyber-physical-social systems comprise physical, cyber, and social worlds with various resources as services. The operation and configuration of these services require service composition approaches that can be used to integrate essential components in these worlds, and organize and share relevant information as needed. Although the present service-oriented architecture can effectively support service composition, there still exist many challenges such as high time cost and low reliability for cyber-physical-social systems. In this paper, we propose a fast and reliable service composition approach to integrate the physical network, cyberspace, and social network. In this approach, first, skyline component computation is performed to reduce the solution space; then, the coefficient of variation is employed to filter the components with high quality of service (QoS) fluctuation and finally the best optimal components are selected by maximizing the fitness function based on users' end-to-end QoS requirements. We applied our approach to real-world dataset and the results showed that our proposed approach has better reliability as well as lower time cost than other approaches.