QoS-aware service composition with user preferences and multiple constraints

Article type: Research Article

Authors: Liu, Fagui | Deng, Dacheng^{; *}

Affiliations: School of Computer Science and Engineering, South China University of Technology, Higher Education Mega Center, Guangzhou 510006, P.R. China

Correspondence: [*] Corresponding author. Tel.: +86 188 1918 9640; E-mail: [email protected].

Abstract: With the growing application of services on the basis of physical networks and Internet of Things devices, service composition based on quality of service (QoS) has attracted wide attention. However, QoS-aware service composition with multiple QoS constraints is time consuming work. Given the evident user preference for QoS, we propose a service composition framework and a QoS-aware method incorporating local and global optimization models. Candidate services are preprocessed and assigned different priority levels based on local optimization. The priority weight of quality attributes is obtained by a fuzzy analytic hierarchy process and the approximate optimal composite service is derived by global optimization for multiple QoS constraints. Simulations verify the proposed method outperforms traditional integer programming (IP) and genetic algorithm, particularly for large scale datasets, providing near optimal solutions with significantly less time cost.

Keywords: Network service, service composition, user preference, quality of service, multiple constraints

DOI: 10.3233/JHS-160542

Journal: Journal of High Speed Networks, vol. 22, no. 3, pp. 193-204, 2016