End-to-End Interrupted Sampling Repeater Jamming Countermeasure Network Under Low Signal-to-Noise Ratio.

Area of Science:

• Electrical Engineering
• Signal Processing
• Radar Systems

Background:

• Interrupted sampling repeater jamming (ISRJ) poses significant challenges to radar systems by generating spurious targets.
• Existing ISRJ countermeasures often degrade real target signal integrity, particularly in low signal-to-noise ratio (SNR) environments.
• These limitations lead to reduced detection performance and compromised radar system effectiveness.

Purpose of the Study:

• To propose a novel deep learning approach for effective ISRJ suppression.
• To address the limitations of existing methods in preserving real target signals under low SNR conditions.
• To enhance radar detection performance in the presence of ISRJ.

Main Methods:

• Development of a complex-valued encoder-decoder network (CVEDNet) for end-to-end ISRJ suppression.
• Utilizing signal decomposition and a stacked dual-branch structure with multi-scale dilated convolutions.
• Employing a multi-domain combined loss function integrating waveform and range-pulse-compression information.

Main Results:

• CVEDNet successfully suppresses ISRJ signals without compromising real target echo fidelity.
• The method demonstrates superior performance in low SNR conditions compared to conventional techniques.
• Validated improvements in jamming suppression capability, echo fidelity, and detection performance indicators.

Conclusions:

• The proposed CVEDNet offers an effective and robust solution for ISRJ suppression in radar systems.
• This deep learning approach preserves essential target signal characteristics, enhancing radar reliability.
• The method shows significant promise for improving radar performance in challenging electronic warfare scenarios.