A framework of synthetic aperture radar imaging based on iterative reweighted compressed sensing

Synthetic aperture radar system has an ability to generate high resolution image data. It requires high rate data acquisition, huge data storage, high power consumption and high onboard resources complexity. Compressive sensing offers solution to overcome those problems. A new framework of SAR image formation is proposed for sparse targets based on compressed sensing. It allows a compressed sampling with a very few of SAR echo samples required under Nyquist / Shannon theorem of both slow time and fast time radar signals and uses Iterative Reweighted Least Square Lp-minimization algorithm for image reconstruction in the noisy environment. Analysis was conducted by comparing the imaging results of SAR point target with different number of under sampling measurement and signal noise ratio level. The results showed that this approach is able to produce reconstructed sparse image with reduced side lobe and noise as compared to iterative reweighted L1 and respectively L2.