Service Selection Based on Correlated QoS Requirements
Yanchun Wang, Qiang He, Dayong Ye, Yun Jin Yang · 2017
With the proliferation of service-based systems (SBSs), more and more functionally equivalent services with different Quality-of-Service (QoS) are emerging. System vendors can select from these services to create service compositions to fulfill users' multi-dimensional QoS requirements. In order to address the QoS-aware service selection problem, various approaches have been proposed in recent years. However, most of them assume that the users' QoS requirements are deterministic and different QoS dimensions of the requirements are independent of each other. In fact, correlations may exist between different QoS dimensions of users' QoS requirements, which make existing service selection approaches impractical. In this paper, we propose SCORE-QoS (Service Selection based on Correlated QoS Requirements), in which the correlated QoS requirements are represented with QoS correlation functions and integrated into the Constraint Optimization Problem that models the service selection process. Two types of correlated QoS requirements, with and without optimization goals, are considered in SCORE-QoS. The experimental results show that our approach can effectively and efficiently select services for composition based on correlated QoS requirements.