Study Of Airborne SAR Moving Target Detection And Parameter Estimations

SAR moving target detection (SAR-GMTI) technology can realize moving target detection, speed estimations and location with high resolution and has an important significance in the field of both military and civilian. Airborne SAR has great flexibility, many developed countries in the word are trying to seek advanced moving target detection methods,developing airborne SAR moving target detection system with high performance.This dissertation mainly studies the SAR moving target detection and parameter estimations. The specific contents of this paper are as follows:1. SAR moving target geometric model is established,and SAR moving target echo is analyzed; the influence of moving target for SAR imaging is analyzed from the aspects of moving targets influence Doppler centroid and Doppler frequency rate respectively; the range migration of moving target is analyzed.2.The ground clutter of SAR and constant false alarm processing are researched; K clutter model is thoroughly analyzed;Dual channel SAR system of slow moving target detection model is established; DPCA algorithm and ATI algorithm are researched; DPCA algorithm and ATI algorithm performance are analyzed;the influence of channel error is researched; compared the detection performance of DPCA algorithm and ATI algorithm under the condition of different signal-to-noise ratio(SNR), different signal-to-clutter ratio (SCR) and different speed; Get a conclusion that DPCA is affected by the noise obviously,while ATI is affected by the clutter obviously,and the theoretical analysis is presented.3.Fractional fourier transform(FrFT) is analyzed, reseached how to use the FrFT detect LFM signal and estimate its parameters;an algorithm combined DPCA and FrFT is presented; a three-channel DPCA geometric model of the moving target detection is established; use FrFT and peak amplitude ratio to estimate target parameter after CFAR determines the existence of moving targets; multiple moving targets detection method is studied, verify the effectiveness of the algorithm under noise and clutter.