Abstract
In order to reach the intelligent recognition, the deep learning classifiers adopted by radar waveform are normally trained with transfer learning, where the pretrained convolutional neural network on an external large-scale classification dataset (e.g., ImageNet) is used as the backbone. Though transfer learning could effectively avoid overfitting, transferred models are usually redundant and might not generalize well. To eliminate the dependence on transfer learning and achieve high generalization ability, this paper introduced neural architecture search (NAS) to search the suitable classifier of radar waveforms for the first time. Firstly, one of the innovative technologies in NAS called differentiable architecture search (DARTS) was used to design the classifier for 15 kinds of low probability intercept radar waveforms automatically. Then, a method with an auxiliary classifier called flexible-DARTS was proposed. By adding an auxiliary classifier in the middle layer, the flexible-DARTS has a better performance in designing well-generalized classifiers than the standard DARTS. Finally, the performance of the classifier in practical application was compared with related work. Simulation proves that the model based on flexible-DARTS has a better performance, and the accuracy rate for 15 kinds of radar waveforms can reach 79.2% under the −9 dB SNR which proved the effectiveness of the method proposed in this paper for the recognition of radar waveforms.
Highlights
In modern electronic warfare, the classification of radar waveforms is one of the pivotal technologies in radar countermeasures and reconnaissance systems
It can be seen from above figures that the flexibleDARTS is superior to the standard differentiable architecture search (DARTS) whether in searching speed or in training stability. e standard DARTS requires 38 epochs to complete the search even results might be very unstable. e flexible-DARTS only requires 17 epochs to complete the search, and the results are obviously stable. e searching results proves that the addition of the auxiliary classifier can enhance the stability of the search time
In order to solve the dependence on transfer learning, this paper introduces neural architecture search into the recognition of radar waveforms, using differentiable architecture search (DARTS) to design the recognition model
Summary
The classification of radar waveforms is one of the pivotal technologies in radar countermeasures and reconnaissance systems. It is an important basis for judging the threat of enemy weapons [1, 2]. With the application of various new radar systems based on low probability of intercept (LPI) technology, traditional classification could not meet the needs of actual electronic warfare any more. Researchers convert the waveform into two-dimensional time-frequency image by Choi–Williams distribution (CWD) time-frequency analysis [3] or other techniques and send it to different models to achieve continuous upgrading of recognition capabilities. Compared with other neural networks [5,6,7], the convolutional neural network (CNN) has a better performance in the processing of image, including radar and sonar images, facial images, and hand gesture images [8,9,10]. erefore, it has been widely used in the recognition of radar waveforms [7, 11,12,13,14,15,16,17,18,19,20]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
