COMPRESSIVE SENSING SAR IMAGE RECONSTRUCTION BASED ON A PSEUDORANDOM 2-D SUBSAMPLING MEASUREMENT MATRIX

Along with increasingly intense desire to achieve super high-resolution images, synthetic aperture radar (SAR) is facing more severe technical challenges such as sampling, storage and transmission of massive data as well as high complexity of hardware. Compressive sensing (CS) theory, which utilizes the signal sparsity, can implement accurate image reconstruction from an extremely less amount of measurements than what is typically considered necessary in Nyquist-Shannon sampling theorem. In this paper, the detailed CS SAR image reconstruction model and process is presented in comparison with back-projection algorithm (BPA). Meanwhile, a pseudorandom 2-D subsampling measurement matrix is redesigned by considering the antenna pattern weighting and linear frequency modulation (LFM) signal. The validity and performance of CS technique using this matrix is exhibited by sparse scene simulation results with multi-point targets.