基于Bagging-CNN雷达信号分类方法

Radar Signal Classification Method Based on Bagging-CNN

针对雷达辐射源信号在低信噪比情况下分类精度不高、分类结果具有偏向性,设计了一种基于集成学习结合深度学习的雷达信号分类方法。首先利用模糊函数特征对雷达信号进行特征提取,对提取的图像特征进行预处理得到网络训练的数据集。然后构造一个集成多个卷积神经网络分类模型对雷达信号进行识别。采用5种不同的雷达信号进行验证,实验结果表明,相比于单个卷积神经网络,采用集成学习和卷积神经网络结合的分类方法有助于提高低信噪比信号的分类效果。

In order to solve the problem of low classification accuracy and biased classification results of radar radiation signals at low SNR,a radar signal classification method based on integrated learning and deep learning was designed.The fuzzy function feature was used to extract the radar signal,and the extracted image feature was processed to obtain the network training data set.Then an integrated multiple conventional neural network classification model is constructed to identify radar signals.Five different radar signals are used for verification.The experimental results show that compared with a single conventional neural network,the classification method combining integrated learning and conventional neural network is helpful to improve the classification effect of low SNR signals.