Deep Reinforcement Learning Based Anti-Jamming Using Clear Channel Assessment Information in a Cognitive Radio Environment

Abstract

Jamming as a type of denial of service attack has proved to be destructive to communication systems. This paper investigates and implements an anti-jamming scheme in a dynamic jamming environment. In our study, we utilize the clear channel assessment (CCA) information available in the MAC layer of a standard IEEE wireless device. Consequently, we eliminate the need for additional equipment to obtain the raw spectrum information. This contrast existing works which need a priori knowledge of the jamming patterns or employ raw spectrum information. The CCA information of all available spectrum channels is utilized as input states to train a double deep q-network (DDQN) agent online to mitigate the effects of jamming. Numerical results show that the proposed anti-jamming approach is effective in different jamming scenarios.