earticle

영어

Secure problem has become a major concern in spectrum sensing. PUE attack is a common attack in spectrum sensing. To defend PUE attack, an anti-PUE attack method in spectrum sensing based on fractal dimension is proposed. It detects PUE attack by identifying the modulation type of the received signal using SVM classifier. Sevcik fractal dimension in frequency domain (SFDF) and Higuchi fractal dimension (HFD) of the received signal are adopted as the characteristics for classification by SVM classifier. So the task of anti-PUE attack can be carried out in spectrum sensing. Besides the parameters for spectrum sensing, i.e. SFDF and HFD, no other parameter is required to be calculated, which will decrease the calculation amount and calculation time. The Numerical results show that, the proposed method can effectively detect the PUE attack. When SNR is larger than 10 dB, its PUE detection probability can reach 1. Even when SNR is low to -10 dB, the PUE detection probability is larger than 0.97.