What a Cognitive Radio Network Could Learn From a School of Fish

In recent years, various types of control algorithms have been proposed for cognitive radios (CR), ranging from al-gorithms coordinated by centralized control to ones coordi-nated in a distributed manner. These algorithms, however, all require communication to either peer nodes or a master node, thus creating communication overhead and potential vulnerability. We introduce a new class of control algorithms to the area of CRs derived from observations of emergent design in nature. Specifically, we introduce an algorithmic approach based on swarm behavior to the task of config-uration management in CR networks. Without requiring the exchange of information among peers or a central au-thority, CRs equipped with such an algorithm are able to globally optimize the configuration of a CR network in the presence of interference and jammers, while only relying on local information, thus providing a fast and efficient way for configuration management especially for large networks