基于小波熵的辐射源指纹特征提取方法

发布时间：2018-06-20 01:53

本文选题：辐射源识别 + 小波熵　；参考：《数据采集与处理》2014年04期

【摘要】：在对辐射源信号进行小波分析的基础上,提出一种基于小波熵的辐射源指纹特征提取方法。首先计算辐射源信号的功率谱,对功率谱进行连续小波变换,提取不同尺度下小波系数的熵特征作为辐射源信号指纹特征。识别分类器采用概率神经网络,对20部手持机进行识别实验,并与传统矩形积分双谱进行对比。实验结果表明,该方法能够把辐射源信号的时频特性信息通过小波系数的熵特征映射到特征向量中,从而实现对辐射源个体的有效识别,而且该特征参数对噪声干扰不敏感,在信噪比为20dB时,系统识别率达到95%以上,在信噪比为5dB时系统识别率仍优于80%,验证了所提方法的有效性。
[Abstract]:On the basis of wavelet analysis of emitter signal, a method of fingerprint feature extraction based on wavelet entropy is proposed. Firstly, the power spectrum of emitter signal is calculated, the power spectrum is transformed by continuous wavelet transform, and the entropy feature of wavelet coefficients is extracted as fingerprint feature of emitter signal at different scales. The recognition classifier uses probabilistic neural network to carry out recognition experiments on 20 handsets and compares it with the traditional rectangular integral bispectrum. The experimental results show that the time-frequency characteristic information of the emitter signal can be mapped to the eigenvector by the entropy feature of the wavelet coefficient, so that the individual of the emitter can be effectively identified, and the characteristic parameter is not sensitive to noise interference. When the SNR is 20 dB, the recognition rate of the system is over 95%, and the recognition rate is still better than 80 when the SNR is 5 dB. The validity of the proposed method is verified.
【作者单位】：解放军理工大学通信工程学院;
【分类号】：TN911.6

【相似文献】