Tradeoff between spoofing and jamming a cognitive radio

Recent studies show that spectrum sensing in cognitive radio exposes vulnerabilities to adversaries. An intelligent adversary can launch sensing disruption in the sensing duration, by putting spoofing signals in allowable bands to prevent secondary users from accessing. In the meantime, the adversary can also attack secondary users by traditional jamming, once they access the spectral bands and start transmission. Both attacks can significantly degrade the performance of a cognitive radio system. In this paper, we address the design of an energy constrained intelligent adversary. More specifically, a global optimization problem is formulated, to optimally distribute its energy between spoofing and jamming, so that the average sum throughput of the secondary users is minimized. To simplify the computation complexities, we divide our optimization into a 2-step problem: first optimally spoof and then optimally jam. Numerical results show that, to induce the worst effect on the average sum throughput of the secondary users, there is a tradeoff between spoofing and jamming: 1) when spoofing and jamming capabilities are comparable, the optimal attack is a combination of partial-band spoofing and partial-band jamming; 2) when spoofing is more effective, a spoofing only strategy is required; 3) when jamming capability dominates, a jamming only attack should be adopted.