QoS-Aware Path Finding and Load Balancing in Service-Composition

Abstract

Service-composition is an effective way to satisfy users' complex and multi-functional requirements in the current cloud environment. Aiming at the problem that the current large-scale service-composition is inefficient and service system is instable caused by load imbalance, this paper proposes a path finding and multi-path service-composition method based on QoS. Firstly, the services with the same input and output parameters are classified into one service class. Secondly, according to the input and output parameters matching, irrelevant services are removed and effective services sets and parameter source tables are obtained. Finally, according to the user’s requests and the QoS constraints, multipaths of service composition are generated, which satisfy the goal of load balancing. Simulation results show that our proposed methods not only can satisfy the user’s requests, but also find more choices for the execution of service composition.