New Researches in Electronic Defense Systems

New Researches in Electronic Defense Systems

Introducing a method for recognition of Low Probability of Interception Radar signals based on deep learning

Document Type : Original Article

Authors
Esmaeil Kogani 1
Mohammad Hosein Sedaghi 2
1 Department of Electrical Engineering, Sahand University of Technology, Tabriz, Iran. E-mail: kogani621@gmail.com
2 Department of Electrical Engineering, Sahand University of Technology, Tabriz Iran
10.22034/joeds.2026.565820.1108
Abstract
In this paper, a suitable neural network is designed to increase the detection and classification of signals emitted from military targets. The proposed algorithm presents a novel method for target signal detection based on artificial intelligence. After generating two datasets of radar signals, the necessary preprocessing is performed on these data using the time-frequency transformation method and spectrograph images related to these data are generated. Then, by applying the dataset images to the model, the training, validation, and testing processes are performed. The results obtained show that the designed model is able to automatically and optimally extract radar signal features at different levels, and by selecting the best and most effective features, it increases the accuracy and reduces the error in signal detection and classification. The target detection accuracy is more than 97 percent, the detection error in the training data is 2 percent and in the validation data is 1 percent. By optimizing the detection accuracy, the target detection speed has increased acceptably.
Keywords
Artificial Intelligence
Deep Learning
Convolutional Neural Networks
Emerging and Novel Technologies
New Researches in Electronic Defense Systems
Volume 4, Issue 13
Spring 2026
Pages 1-12