Research On Liner Array Three-dimensional SAR Imaging Technology

As a new radar imaging system, liner array three-dimensional synthetic aperture radar(LASAR) can reflect the three-dimensional scatter information of targets without distortion. Limited by the Nyquist sampling theorem and affected by the grating lobe, the traditional LASAR imaging techniques apply only to uniform array, and need array elements spacing little as possible which leads a great number of elements. What’s more, for the Rayleigh limit and high side lobe, the resolution of the traditional methods is always low. For the above shortcomings, this dissertation study the new imaging methods based on compressed sensing theorem and modern spectral estimation technology using a spot of array elements. The simulation proves that these new methods can effectively improve resolution of cross-track direction, reduce the number of array elements and expand the no fuzzy imaging range.The main contents is as follows:(1).This dissertation introduces the geometrical model and echo of the downward-looking LASAR. The spatial resolution and its spatial variation characteristic is analyzed. Lastly, liner measurement matrix model of echo is constructed.(2).This dissertation gives two non-uniform array configurations: co-prime nested array and random distribution array. Then a new imaging method based on the iterative adaptive approach(IAA) is given, and simulation proves that the method can effectively solve the problems of traditional methods. Then, taking noise into consideration, a improved IAA is applied to LASAR imaging and achieve good results.(3).On the basis of sparse characteristics of three-dimensional scene, this dissertation uses compressed sensing theorem to restructure the cross-track direction scattering coefficient of the targets. A imaging method based on threshold based OMP algorithm is proposed. The method can avoid the imaging error caused by the sparse degree error.(4).The dissertation gives the expression of radar echo in real number field, and using BCS algorithm and GPSR algorithm, which have a better performance, to three-dimensional imaging in LASAR system with non-uniform array. The simulation proves the feasibility of the method.(5).At last part, the dissertation compares the performance of above imaging methods, included the speed, limit resolving power and anti-noise performance, from the simulation results we obtain that the threshold based OMP method has a fastest speed, and GPSR method has a best comprehensive performanc e.