基于谱包络分形特征提取的仿冒攻击检测

Counterfeit attack detection based on spectral envelop fractal characteristic extraction

以辐射源电子器件本身存在的寄生调制作为指纹特征,改进分形维数算法。将幅度谱信号的盒维数与信息维数各自的均值与方差之比求和,记作新的分形维数。针对八组辐射源实测信号的仿真表明,主用户和仿冒攻击用户的新分形维数区别明显,可用于仿冒攻击检测。

The fractal dimension algorithm is improved by taking the spurious modulation in radiation source electron device as fingerprint feature.Add up the ratios between the means and the variances of the box dimension and the information dimension of the magnitude spectra signal accordingly,thus,the sum can be taken as a new fractal dimension.A simulation is carried out with eight sets of radiation source actual measurement signals,and the results show that,the new fractal dimension is distinctively different between each master user and each counterfeit attack user,and therefore,it can be used for counterfeit attack detection.