Evolving Presentity-Based Context Schemas by Estimating Context Proximity

The definition of what constitutes context proximity has remained largely unexplored but accepted as being a fundamental issue towards realising an architecture of connected things. Existing solutions aimed at enabling context awareness are often undermined by their dependencies on centralized architectures limited with respect to their scalability. Our previous work proposed the use of the so called Context Schema; an encapsulated representation of the information points constituting the context of a presentity. Building such a schema requires support for determining set members limited by some metric; a proximity metric. In this paper, we propose an algorithm for estimating the context proximity among presentities, enabling complete schemata of entities relevant to, and expressing the current context of a presentity. Secondly we propose an extension of a gossiping algorithm to optimize the ability create schemata as one traverses a vast and dynamic connected things infrastructure.