Detecting Multi-Channel Wireless Microphone User Emulation Attacks in White Space with Noise

In this work, we study a special kind of primary user emulation (PUE) attack, named wireless microphone user emulation (WMUE) attack in white space cognitive radio networks. In WMUE attacks, a malicious user emulates wireless microphone (WM) signals in order to block sencondary users. Existing work on WMUE attack detection deals with single channel senario. Although multi-channel WM (MCWM) systems are common, detecting WMUE attacks under a multi-channel setting in noisy environments has not been well studied and the existing solution for single channel case cannot be directly applied. In a practical multi-channel WM system, the audio signals on different channels mix with each other and are contaminated by noises, which introduce great challenges on WMUE attack detection. We propose a novel multi-channel WMUE attack detection scheme which is based on the cross-correlation between the demodulated FM signal and the acoustic signal. The audio interferences, audio noises, and RF noises are all resisted by the cross-correlation. To reduce computation complexity, we propose a 1.5-bit FM demodulator whose outputs are represented by only 0, 1 and -1. Moreover, we set up a MCWM system and developped a hardware based prototype to evaluate the performance of the proposed scheme. Experimental results show that, the proposed scheme can effectively detect multi-channel WMUE attacks within 0.25 second with detection rate larger than 0.9 and false alarm rate lower than 0.1 under low SNR.