Estimation efficiency, accuracy and robustness improvement by exploiting the geometry information in SAR-GMTI system

Suffering from ambiguous estimation or heavy computation complexity load, the radial velocity estimation of moving targets becomes the bottleneck of the synthetic aperture radar-ground moving target indication system. In order to improve the radial velocity estimation efficiency, we have proposed an efficient Radon transform (RT) estimation (ERTE) method by using the never exploited geometry information, which performs well in high SNR scenarios but bad for low SNR. Focusing on these, we propose the least square RT estimation (LSRTE) method to improve the estimation accuracy by utilizing the geometry information of multiple RT results. Given the geometry information determining measurement error, we modified the LSRTE into a weighted LSRTE (WLSRTE) method to improve the estimation robustness and accuracy. Experiments results validate the effectiveness of the proposed methods, and the proposed methods perform much more accurate and efficient than the conventional method.