QoS-Aware Service Composition in Mobile Cloud Networks

Abstract

Content delivery through cloud networks has gained popularity due to its effectiveness and reliability. Much progress has been made to allow for the easy integration of heterogeneous systems to compose services on the fly. Despite all efforts, the task of fast service composition has made the assumption of the presence of static searchable and updatable central repositories within cloud data centers. With today's dynamic and mobile environments, service providers' movements are unpredictable, thus a rejuvenated service composition mechanism for cloud environments is needed. This paper presents a cloudlet-supported service composition solution in cloud networks that can find a suitable mechanism to discover mobile media processing functions and seamlessly integrate them into media delivery sessions. Furthermore, the established sessions are monitored by a novel client-side QoS measurement collection mechanism and adapted dynamically to network, user, and service providers' changing conditions. Simulation results showcase the effectiveness of the presented solution in terms of composed service stability and QoS measurement collection accuracy.