The objective of the thesis is to propose a middleware for service composition and execution, and demonstrate its application in pervasive computing environments such as MANETs. The middleware should consider the dynamicity of services and the context of user and his/her environment to choose an appropriate service composition and execution plan.
In this thesis, we have proposed a middleware called ConAMi (Context-Aware Service Composition Middleware) that demonstrates the application of service composition in content adaptation. The ConAMi middleware performs content adaptation by composing services distributed in the environment. The composite service is derived from the contextual information (user’s preferences, device capacities, network condition) and services description (availability, functionality, execution time). ConAMi executes the composite service in a mesh-based manner which enhances the performance of the system related to network overhead and execution time. The middleware is integrated with a fault detection and recovery mechanism to overcome the failure of composite service execution caused by several kinds of faults such as network disruption, service discovery failure, service execution failure, etc.