An Approach to Resource and QoS-Aware Services Optimal Composition in the Big Service and Internet of Things

Recently, various types of services belonging to different domains intertwine together and constitute the Big Service. In the Big Service and Internet of Things, services' optimal composition is a key technology to create value-added services to satisfy users' complex requests. However, massive services that possess the same functionalities but different quality of services (QoSs) are emerging on the Internet. Moreover, the online performance of many online services is determined by their distribute resources. Therefore, the expected performance of a composite service depends on the creation of the optimal composite service that can meet end-to-end quality requirements while ensuring that component services have sufficient resources to support their successful execution. To this end, resource and QoS-aware services' optimal composition (RQ-SOC) becomes an important issue in the Big Service and Internet of Things. Moreover, with the evolution of service industries, service features in various service domains (SFSD) (priori features, correlation features, and similarity features) are gradually formed. These SFSD have great influences on the RQ-SOC. Thus, to effectively solve the RQ-SOC problem, this paper first defines the SFSD and describes the important influences of SFSD on the RQ-SOC. Then, the improved artificial bee colony (ABC) algorithm for RQ-SOC is proposed, and a resource checking operator based on the analysis of the mutual relations between services and resources is presented. Third, the resources checking operator is integrated into the improved ABC to solve the RQ-SOC problem effectively. Finally, the experimental results show that the proposed method for RQ-SOC is feasible and effective.