Jamming-Resilient Dynamic Spectrum Access for Cognitive Radio Networks

Many spectrum sensing and access schemes have been proposed in recent years for opportunistic spectrum access (OSA). Most of them model the spectrum sensing and access problem as a partially observed Markov decision process (POMDP), where the channel statistics are assumed to be known to the secondary users (SU). However, few of them have considered the attackers who can launch effective jamming attacks by exploiting the same information that SUs have. In this paper, we study the problem of anti-jamming dynamic multichannel access in cognitive radio networks (CRNs) as a multiple-armed bandit (MAB) problem, where the transmitter and receiver adaptively choose arms (i.e., sending and receiving channels) to operate based on the common acknowledgment information. The proposed scheme enables the transmitter and receiver to hop to the same set of channels with high probability. Our simulation results show that the proposed scheme is resilient to various jamming attacks, and the SU transmitter and receiver can develop common knowledge on channel availability through online learning.