Toward A Cognitive Radio Architecture: Integrating Knowledge Representation With Software Defined Radio Technologies

The ultimate vision of cognitive radio technology encompasses many capabilities including autonomous execution of tasks that today require manual intervention. A conventional radio when operating in a particular communications mode always follows the same procedure and either succeeds or fails at a given task. A cognitive radio, by contrast, can have a knowledge-driven differential-response capability; that is, it can use knowledge of radio technology and policy, representations of the goals, and other contextual parameters to reason about a failed attempt to satisfy a goal and to identify alternative actions that would achieve the goal. We have built a prototype simulation framework for a cognitive radio that exhibits this capability in various scenarios. Based upon this experience, this paper proposes a general architecture that merges knowledge representation technologies (both ontologies and rules) with the processing structures of existing software defined radio technology to enable this capability as well as form a foundation for other cognitive abilities.