A self-organizing and self-configuration algorithm for resource management in service-oriented systems

With the ever increasing deployment of service-oriented distributed systems in large-scale and heterogeneous computing environments, clustering and communication overlay topology design has become more and more important to address several challenging issues and conflicting requirements, such as efficient scheduling and distribution of services among computing resources, reducing communication cost between services, high performance service and resource discovery while considering both inter-service and inter-node properties and also increasing the load distribution and the load balance. In this paper, a four-stage hierarchical clustering algorithm is proposed which automates the process of the optimally composing communicating groups in a dynamic way while preserving the proximity of the nodes. The simulation results show the performance of the algorithm with respect to load balance, scalability and efficiency.