Knowledge management for context-aware, policy-based ubiquitous computing systems

Abstract

Ubiquitous computing systems depend on a distributed intelligence that relates the context of the task being performed to the available system resources and services. Currently, this is hampered by the problems inherent in heterogeneous devices and technologies used to manage network resources and services, and the associated lack of a network language lingua franca that all systems can use and understand. This paper describes a new approach for representing, using and managing knowledge for ubiquitous computing systems, based on a novel combination of extracting knowledge from models and ontologies to form such a lingua franca. An extensible context model is used to select applicable policies to govern system behaviour; as context changes, policies change, which in turn causes system behaviour to change accordingly.