Jamming Recognition Based on Adaptive Feature-Focusing Convolutional Neural Network for Agile Cognitive Radar

Area of Science:

• Radar Systems Engineering
• Artificial Intelligence in Signal Processing
• Cognitive Radar

Background:

• Deep neural networks are crucial for radar jamming recognition in cognitive radar systems.
• Radar waveform parameter agility, an anti-jamming technique, challenges conventional CNN-based jamming recognition by altering signal features.
• This creates a trade-off between effective jamming recognition and anti-jamming agility.

Purpose of the Study:

• To develop a novel deep learning approach for robust radar jamming recognition.
• To overcome the limitations imposed by radar inter-pulse parameter agility on jamming identification.
• To enhance the adaptability of jamming recognition systems to dynamic radar environments.

Main Methods:

• Proposed an Adaptive Feature-Focusing CNN (AFF-CNN) incorporating a pre-trained AFF module.
• The AFF module maps agile radar parameters to adaptive feature scales, calibrating deviations caused by parameter agility.
• Utilized time-domain high-resolution range profiles (HRRP) and time-frequency domain short-time Fourier transform (STFT) data with a lightweight 1D-2D feature fusion CNN for single-pulse signal recognition.

Main Results:

• The AFF-CNN demonstrated superior recognition accuracy compared to five other approaches.
• The proposed method showed enhanced generalization capability in adaptive scenarios.
• Effective adaptation to inter-pulse agility in radar systems was confirmed through simulations.

Conclusions:

• The AFF-CNN effectively addresses the challenge of radar parameter agility in jamming recognition.
• The adaptive feature-focusing mechanism significantly improves the network's performance in dynamic environments.
• This approach offers a promising solution for robust jamming recognition in advanced cognitive radar systems.