Inverse Synthetic Aperture Radar Imageing Based On Compressed Sensing

Recently, a new theory known as Compressed Sensing(CS) is proposed. It indicatesthat a sparse signal or a compressible signal can be recovered with a high probability fromthe measurements less than usual. Inverse synthetic aperture radar(ISAR) image has fewernumber of scatterers which can be considered as a scene with sparsity, that meets theapplication conditions of compressed sensing.In the thesis, the theory and algorithms of CS based ISAR imaging is discussed andcombined. The main contents are summarized as follows:Firstly, the compressed sensing theory is described briefly, then the sparsereconstruction algorithms of compressed sensing complex domain is researched mainly,and the performance of several common matching pursuit algorithms is compared. Then arandom selection strategy based on matching pursuit algorithm is proposed, thesub-dictionary of reconstructing signal is composed by selecting randomly some atomsfrom the entire dictionary, and avoiding the process of using the entire dictionary duringeach iteration, experiments demonstrate the effectiveness of the algorithm.Secondly, the two-dimensional inverse synthetic aperture radar imaging algorithm isstudied, the formation of two-dimensional imaging is analyzed by basing on the inversesynthetic aperture radar signal model, obtaining a2-d ISAR imaging under two differentforms: a small-bandwidth small-angle and a wide-bandwidth large-angle. Then, theinverse synthetic aperture radar imaging Range-Doppler algorithm is researched. Theprocess flow of the Range-Doppler algorithm is described in detail, and motioncompensation is performed to the compressed image, that achieves a one-dimensional andtwo-dimensional ISAR imaging, and imaging results from point targets and the ship targetunder traditional Range-Doppler algorithm.Finally, compressed sensing reconstruction algorithm for inverse synthetic apertureradar imaging is researched, establishing the sparse representation models of echoes afteranalyzing the signals’ features. Then, a2-d inverse synthetic aperture radar imaging basedon compressed sensing is composed by the signal reconstruction algorithm combined with sparse representation models of echoes and non-correlation measurement matrix. Throughthe experimental results of the simulation data processing and compared to the traditionalISAR imaging algorithm verify the effectiveness of the proposed algorithm.