Frequency-Agility-Based Neural Network with Variable-Length Processing for Deceptive Jamming Discrimination.

Area of Science:

• Radar Systems Engineering
• Artificial Intelligence in Defense
• Electromagnetic Compatibility

Background:

• The rise of the low-altitude economy and Unmanned Aerial Vehicles (UAVs) increases the importance of Integrated Sensing and Communication (ISAC).
• Low-altitude platforms like UAV swarms are susceptible to deception jamming in complex electromagnetic environments.
• Existing multistatic radar systems struggle with processing slowly fluctuating targets and adapting to dynamic electromagnetic conditions.

Purpose of the Study:

• To develop a robust method for discriminating true UAV targets from deception jamming in multistatic frequency-agile radar systems.
• To address the challenges posed by slowly fluctuating targets and complex electromagnetic environments.
• To enhance the protection of low-altitude UAVs against sophisticated jamming techniques.

Main Methods:

• Leveraging the rapid amplitude fluctuation characteristic of frequency-agile radar echoes for target analysis.
• Developing a neural network approach for deep feature extraction from true and false target echoes.
• Proposing a neural network-based variable-length processing method for deception jamming discrimination.

Main Results:

• The proposed method effectively exploits deep-level echo features to improve true vs. false target discrimination.
• Significant improvements in discrimination probability were observed, particularly for slowly fluctuating UAV targets.
• The model demonstrates practical deployment capability by processing variable pulse counts after fixed-count training.

Conclusions:

• The developed neural network method offers a feasible and effective solution for deception jamming discrimination in multistatic frequency-agile radar.
• The approach enhances the resilience of ISAC systems supporting low-altitude UAV operations.
• The variable-length processing capability makes the method highly adaptable to dynamic mission scenarios.