Super-resolution ISAR imaging via statistical compressive sensing

Developing compressed sensing (CS) theory has been applied in radar imaging by exploiting the inherent sparsity of radar signal. In this paper, we develop a super resolution (SR) algorithm for formatting inverse synthetic aperture radar (ISAR) image with limited pulses. Assuming that the target scattering field follows an identical Laplace probability distribution, the approach converts the SR imaging into a sparsity-driven optimization in Bayesian statistics sense. We also show that improved performance is achieved by taking advantage of the meaningful spatial structure of the scattering field. To well discriminate scattering centers from noise, we use the non-identical Laplace distribution with small scale on signal components and large on noise. A local maximum likelihood estimator combining with bandwidth extrapolation technique is developed to estimate the statistical parameters. Experimental results present advantages of the proposal over conventional imaging methods.