A novel method for estimating the baseband Doppler centroid of conventional synthetic aperture radar

The Doppler centroid is a crucial parameter for the purpose of implementing azimuth processing for synthetic aperture radar (SAR) data. In this paper, we propose a scheme for estimating the Doppler centroid of conventional SAR. Virtual multi-channel SAR data can be generated by sub-sampling the original SAR data in azimuth. Consequently, each channel data is aliased and the spectrum components within each Doppler bin can be viewed as sources from different known directions. Then the Capon spectral estimator can be utilized to estimate the antenna pattern. Based on the estimated antenna pattern, we can obtain the baseband Doppler centroid of the original SAR data. Finally, experiments are provided to validate the effectiveness of the proposed algorithm.