短时三谱分析在通信电台个体识别中的应用

Identification of Individual Radio Transmitters Using STFT Tri-spectrum

提出一种利用短时Fourier变换理论构造三谱的新方法,用类似于普通三谱的短时Fourier三谱来分析信号的局部频谱特征,推导了短时三谱的数学形式,并用短时三谱系数提取信号频率相位特征。通过仿真实验分析了窗函数及其参数的选择,并运用支撑向量机(SVM)完成信号的识别。实验结果表明,此方法与普通双谱特征相比能够更好地反应通信电台个体的细微差异,具有更强的分类能力。

A novel method of using STFT tri-spectrum is proposed for identifying individual radio transmitters by means of analyzing the local spectrum features. The algorithm of STFT trispectrum is derived, and the phase feature of signal fre- quency is extracted by STFT tricoherence. Then the selections of window function and its parameters are discussed through the computer simulation, and support vector machine （SVM） is applied to realize individual transmitter identification. The ex- perimental results show that this method can achieve a higher classification rate compared to the common hi-spectrum method for its better performance in reflecting the nuances of individual radio transmitters.