High resolution SAR imaging with efficient azimuth compression in chirp scaling principle

Chirp scaling algorithm (CSA) is one of the most popular algorithms in radar imaging, due to its excellent focusing ability and implementation simplicity. However, such superior performance, especially the resolution capability is greatly restricted by the number of observed measurements. Specially, if the measurements are reduced, images with high sidelobes and lower resolution are formed. The recent developed compressed sensing (CS) as well as its application in radar imaging demonstrates that, if the sparsity constraint is cooperated in the imaging technique, high resolution imaging quality can be still achieved, even using limited measurements. In this paper, a new chirp scaling azimuth compression technique based on CS theory is proposed, which exploits the sparse property of the measurements in azimuth dimension, and the sparsity constraint is combined with the CSA to perform all the azimuth processing steps. Comparisons were performed between the proposed algorithm using data under-sampled in the azimuth dimension at different ratios and the traditional CSA using full data set at different signal-to-noise ratio (SNR). Results show that the proposed CS based algorithm has better performance than the traditional algorithm even with using low percentage of the azimuth data and also indicate that the proposed algorithm is robust with the existence of low SNR.