High-Speed UAV Swarms Detection via Coherent Integration and GTE-Based Super-Resolution Method

The radar detection of high-speed Unmanned Aerial Vehicle (UAV) swarms has gained a lot of popularity in military field. However, the high-speed and high-density features of UAV swarms could introduce challenges in accurately detecting them, including low Signal-to-Noise Ratio (SNR), Range Migration (RM), Doppler Frequency Migration (DFM), insufficient angular resolution, and false alarms. This paper provides a comprehensive approach by combining a long-time coherent integration method with a super-resolution method. Firstly, Second-order Keystone Transform and General De-chirping Process (SKT-GDP) as well as beamforming is applied to integrate signal in range-doppler-space dimension, so the SNR can be enhanced. Then, Frequency-Selective Reweighted Atomic-norm Minimization (FS-RAM) gridless method is modified to efficiently estimate the precise direction of targets in a single angular unit. Thirdly, the source number estimation accuracy is improved through employing a designed strategy to eliminate ghost targets, i.e., Ghost Target Elimination (GTE), following with the CLEAN process to detect the rest targets. Finally, the effectiveness and detection accuracy of the proposed method is verified through several simulations.