基于差分-变分模态分解与全局信息分析网络的辐射源个体识别方法

Approach of Specific Communication Emitter Identification Combining Differential Variable Modal Decomposition and Global Feature Analysis

为了解决基于希尔伯特黄变换(HHT,Hilbert-HuangTransform)辐射源个体识别方法中的模态混叠分解不充分以及低信噪比下效果较差的问题,本文将信号处理与深度学习相结合提出了一种新的辐射源个体识别方法。首先,对信号进行差分处理,并通过变分模态分解得到对应的模态分量;接着,对各模态分量进行希尔伯特变换得到希尔伯特谱;最后,针对希尔伯特谱的稀疏性特点,本文运用改进的全局信息分析模块对其进行全局细微特征提取。本文实验采用ORACLE公开数据集对所提方法进行性能测试,实验结果表明,该方法识别性能优于4种现有的基于希尔伯特黄变换的辐射源识别方法,其不仅有较低的计算复杂度,而且在5 dB信噪比下有着90%以上的识别效果。

To solve the problem of inadequate decomposition between modal components in the Hilbert-Huang transform(HHT)based special emitter identification(SEI),a new SEI method combined signal processing and deep learning was proposed.Firstly,the raw signal was differenced and the corresponding intrinsic modal components(IMF)were obtained by variational modal decomposition(VMD).Then,the Hilbert spectrums were obtained by Hilbert transform of each IMF.Finally,for the sparsity of the Hilbert spectrum,this paper invoked the global context block and improved it to further extract the global subtle features from the Hilbert spectrum.The performance of the proposed method was tested using ORACLE public dataset,and the experimental results showed that the recognition rate of the proposed method is better than four existing methods of SEI based on Hilbert spectrum,which had low computational complexity and had more than 90%recognition rate at 5 dB signal to noise ratio(SNR).