Research on Radar Jamming Signal Classification and Suppression Method Based on Deep Learning
In this paper,a radar interference signal classification and suppression method is proposed un-der the application of deep learning method.Firstly,the automatic classification of radar interference signals is realized under the application of convolutional neural network.It is mainly divided into three layers,namely convolution layer,pooling layer and fully connected layer.After training,the classification model is obtained.Based on this,the design of the de-interference network is realized,and then the interference signal is applied in the reconstruction loss function.After the design is completed,the experimental research is carried out,and a total of 4 types of interference signals and radar signals are selected.The results show that the accuracy of this method in radar signal classification can reach 96.3％,and the effective signal can be retained.The signal-to-noise ratio can be increased to more than 5 dB,and the accuracy of data after noise reduction can reach 98.2％.It can be seen that this method can not only realize the automatic classification of radar interference signals,but also can effectively suppress them,and can achieve good noise reduction effect,which is more conducive to the implementation of radar signal processing.
deep learningconvolutional neural networkradar interference signalclassificationinfer-ence elimination
万方数据
维普
