Cognitive Jamming Game for Dynamically Countering Ad Hoc Cognitive Radio Networks

Abstract

This paper demonstrates that game theory policies can adapt in realtime to enable a cognitive radio network and a cognitive jammer to engage in a game for control of the spectrum. Learning occurs as optimized actions are selected and played out in a dynamic RF propagation environment. System specific resource parameters are optimized for both the cognitive radio network and cognitive jammer. The goals of the two systems are different, the CRN seeks to be spectrally efficient while the jammer seeks to minimize data throughput. It is quixotic to model this interaction as a zero sum game, therefore a non-zero sum game is played. Additional flexibility of the jammer is considered when compared with previous works. Estimated performance results are shown.