DRFM Jamming Mode Identification Leveraging Deep Neural Networks
Jiannan Gao, Min Wang, Longbiao Chen, Bingwei Hui, Cheng Wang, Hongqi Fan · 2021 International Conference on Control, Automation and Information Sciences (ICCAIS) · 2021
In modern information warfare, radar must apply some anti-jamming technologies. Digital radio frequency memory (DRFM) and its technology can adapt to the complex threat signal environment and have an excellent jamming effect that other technical means in the past cannot match. Currently, traditional DRFM jamming identification is mainly based on the method of extracting and modeling radar signal features. It is unable to identify various types of jamming and lacks a large amount of radar data. In this work, we propose a three-phase framework to identify DRFM jamming. First, we design a hardware system based on PXI module to generate jamming signal data. Second, we perform range-doppler processing on the raw radar signal data to generate range-doppler map (RD-map) which can accurately represent jamming signals. Finally, we use deep learning to identify DRFM jamming and use pre-trained models to improve identification accuracy. We evaluate our proposed framework using real DRFM jamming data collected from PXI system. The valuation results confirm that our approach effectively identifies DRFM jamming, achieving an average accuracy of 84.4%.