Reinforcement Learning Based Anti-Jamming Cognitive Radio Channel Selection

Dynamic spectrum management (DSM) models and cognitive radio (CR) technology are presented as promising solutions to the spectrum scarcity and under-utilization problems. However, the CR efficient exploitation of the spectrum can be limited by the jamming attack. In this paper, we use the spectrum sensing and the learning abilities of the CR to solve this problem. The proposed algorithm enables the CR to pro-actively avoid the jammed channels. We present a suitable model to the channel selection problem and we enhance the proposed solution through cooperation between two cognitive radio nodes. Simulation results prove the performance of the proposed solution compared to other solutions and against different jamming strategies.