In order to solve the problem of serious information loss when ambiguity function is used as input of radar emitter recognition deep learning model,a special input method of color mapping four-dimensional view is proposed.Firstly,the ambiguity function is mapped to the red-green color space,and then the input of the recognition model is formed by splicing the three views and the main ridge.This input method is more consistent with the distribution of ambiguity function and the characteristics of deep network image process-ing.The simulation results show that when the special color four-dimensional view is used as the input of AlexNet network,the accuracy is 88.71％when SNR=-6dB,which is 14.61％higher than other input methods.The new input method significantly improves the accuracy and robustness of the identification model and can be used as an input method for the intelligent identification system of radar emitter.
关键词
深度学习/雷达信号识别/模糊函数/色彩映射/特征提取
Key words
deep learning/radar signal recognition/ambiguity function/colormap/feature extraction
维普
万方数据