一种基于相噪特性的辐射源指纹特征提取方法  

Emitter Fingerprint Feature Extraction Method Based on Characteristics of Phase Noise

作　　者：黄渊凌[1] 郑辉[1] 

机构地区：[1]西南电子电信技术研究所盲信号处理重点实验室,四川成都610041

出　　处：《计算机仿真》2013年第9期182-185,共4页Computer Simulation

基　　金：国家"863"高技术计划基金资助课题(2011AA7031014C)

摘　　要：辐射源指纹特征可用于实现对辐射源目标个体的识别。为了提升特征的区分能力和识别性能,提出了基于相位噪声特性的辐射源指纹特征提取方法。通过分析发射机频率源电路的等效数学模型,建立了描述发射机相位噪声特性的自回归-滑动平均(ARMA)模型,并提出通过ARMA参数估计构建辐射源指纹特征,从而完成辐射源个体识别。仿真结果证实了上述方法提取的特征可用于区分辐射源目标个体,且性能在中高信噪比情况下优于相噪功率谱特性的方法。结果表明,与依赖经验认识的经验参数特征方法相比,改进的相噪机理模型的指纹特征能更好地反映辐射源的本质差异,从而获得更好的识别性能。Emitter fingerprint features can be used to uniquely identify individual emitters.In order to improve the discriminability of features and the identification performance,an emitter fingerprint feature extraction method based on the characteristics of phase noise was proposed.The equivalent mathematical models of frequency generators were described,and the autoregressive-moving average （ARMA） models were used to characterize the phase noise （PN） generated by the transmitters.Fingerprint features were extracted through estimating ARMA parameters and used for specific emitter identification （SEI）.The simulation results validate that the proposed features can be used to uniquely identify individual emitters and its identification performance is superior to the power-density-based method in medium to high SNR situations.This suggests that the proposed fingerprint features derived from the phase noise model are more powerful to describe the essential differences between emitters than the features derived from experience knowledge,thus better identification performance can be attained.

关 键 词：辐射源指纹识别 相位噪声 自回归-滑动平均过程 特征提取 

分 类 号：TN911.23[电子电信—通信与信息系统]