A context awareness model enhanced with autonomic features

The increasing complexity of the context sensitive systems, and the difficulties of their management, administration and adaptation have headed us towards the necessity of integrating self-* autonomic computing paradigms (self-configuring, self-healing, self-optimizing and self-protecting) into the development of context sensitive pervasive systempsilas functional components. This paper introduces and defines the concepts of isotropic context space, context granule and context model entropy as the basic features to formally describe and evaluate the RAP context model autonomic characteristics. The paper also propose a methodology for enhancing the RPA context model with self-configuring and self-healing autonomic properties. The self-configuring property or context adaptation is obtained by detecting / configuring / integrating new context resources / actors into the context specific model. The context model self-healing properties are obtained by continuously monitoring the real context for detecting broken context policies and executing compensating actions.