Detecting identity-spoof attack based on BP network in cognitive radio network

In the cognitive radio network (CRN), all the cognitive user opportunistic access the spectrum, so it can improve the spectrum utilization. As a smart spectrum sharing technology, cognitive radio can change its transmitter parameter based on interactive the environment where it operates. So the MAC address can be easy spoofed in the open air. An attacker can exploit this character to preempt spectrum resources and bring the negative impact to the network. In this paper, we propose a scheme of detecting identity-spoof attack based on Back Propagation (BP) network. The scheme extracts the fingerprint feature of cognitive user to get the received signal strength (RSS). The extracted Eigen values can be used as the input vectors of the BP neural network which are trained by the honest users´ RSS. The honest users and the attackers can be divided into two categories by the trained BP network to detect an identity-spoof attack. Simulation result shows that our scheme can effectively detect identity-spoof attack with a low false alarm rate and miss alarm rate.