基于瞬时频率畸变特性的FSK电台指纹特征提取

FSK Radio Fingerprints Extraction Based on Distortions of Instantaneous Frequency

为解决FSK电台个体识别问题,需要从辐射源信号上提取特征构成辐射源指纹。通过分析构成FSK电台的各个模块的畸变特性,阐明了FSK电台指纹产生机理,建立了基于瞬时频率的指纹信号模型,并根据信号模型设计了基于辅助参数的最小二乘算法以完成FSK频率畸变特性参数的估计,从而构建指纹特征完成对FSK电台的识别。对仿真信号和实际信号的识别测试试验表明,该算法具备辐射源个体识别能力,对实际环境下的4个FSK电台的识别率大于96%,优于双谱类方法。

Radio frequency fingerprints （RFFs） are extracted from radio frequency （RF） signals to uniquely identify individual FSK radios. In this paper, the module imperfections in FSK transmitters are analyzed to show how RFFs are generated. A signal model about the instantaneous frequency is constructed to describe the RFFs. Then the least square method with auxiliary vmiables is used to estimate the parameters which characterize the distortions in instantaneous frequency of FSK signals. The RFFs are constructed using the estimates and used to identify individual FSK radios. The tests for simulated signals and real-world signals show that the proposed method is able to uniquely identify individual radios. Specifically, the identification accuracy for 4 FSK radios in practical scenarios exceeds 96%, which outperforms the bi-spectrum approaches.