An architecture for Cognitive Radio ad-hoc network nodes

Cognitive Radio (CR) is considered to be the future of wireless communication. However, for a practical realization of CR, there is a need for an integrated, extensible and flexible architecture with well-defined interfaces and support for real-time adaptations. This article presents a novel architecture for CR nodes operating in an ad-hoc network scenario. Based on a highly reconfigurable software defined radio, it incorporates a Cognitive Resource Manager used for optimization, decision making and learning as well as a database-driven knowledge base called Distributed Resource Map. In contrast to other research in this field, this work focuses on practical feasibility, combines multiple hot topics in the CR domain and considers them holistically. We present implementation details and conclude the paper with intermediate results obtained in our networking testbed.