Modified multiple-measurement vector model for SAR imaging

Abstract

The multiple measurement vectors (MMV) model can be used to effectively represent the jointly sparse signals. However, this model cannot be directly used in the scenario of synthetic aperture radar (SAR) imaging. The reason is that, the MMV model requires that the multiple sparse vectors have the common sparse pattern, while the SAR range profiles at different pulses have different sparse structures due to the range migration effect. In this paper, a modified MMV model is proposed for SAR imaging. The range migration effect is embedded into the proposed model, and a modified orthogonal matching pursuit (OMP) algorithm is developed to obtain the high-resolution range profile. Then the cross-range focusing is performed by traditional matched filtering operation. Simulations and experimental results are presented to demonstrate the effectiveness of the proposed model and algorithm.