Critical Aware Community Based Parallel Service Composition Model for Pervasive Computing Environment

Abstract

Service composition in pervasive computing environments is needed to provide best quality of service. Services need to be discovered at run time and composed together for best possible user scenarios. The need for identifying the best services among the service nodes is essential in pervasive computing systems as the environment of operation can change rapidly. Pervasive computing demands systems that are scalable, adaptive, fault tolerant and can work in heterogeneous environments. Hence an adaptive method that takes into account the environment is the need of the hour. In this work, a dynamic parallel composition model to compose the best matched services is proposed for the pervasive computing environment exhibiting the quality of service and contingency management properties. The model ensures that the highest quality of service conditions is fulfilled. Facilities for contingency management ensure efficient fault tolerance and failure recovery. The proposed model uses the community framework for grouping the service nodes and composing the services provided by the nodes. This ensures that resultant composition mechanism is dynamic in nature to adapt to the service nodes failure without compromising the quality of service with better fault error recovery time. The model has been validated experimentally and the results show considerable promise. The work is unique in its extensive mechanisms for modeling the pervasive computing environment, failure handling, fault tolerance and best quality of service parameters.