Research On The Airborne Downlooking Three Dimensional Imaging SAR Of Unmanned Aerial Vehicle Based On Linear Array Of MIMO Technology

Airborne Downlooking Three Dimensional Imaging SAR of Unmanned Aerial Vehicle is anemerging technology that has drawn considerable attention. Through judiciously designed antennaarrays and diversities, ADL-3D-SAR has been shown to provide enhanced performance in detection,estimation, imaging and etc., so ADL-3D-SAR has significant potential for advancingthestate-of-the-art of modern radar. Aimed at the application of ADL-3D-SAR on the surface of theearth of the imaging target detection problem, ADL-3D-SAR imaging algorithms are studied in thisdissertation. The main contents fall into four sections:1.The research background and its significance of the ADL-3D-SAR based on linear array ofMIMO technology are introduced. the paper introduces its research situation is mainly from theMIMO-SAR concept, classification and research trends.2. ADL-3D-SAR based on linear array of MIMO technology is introduced. Mainly elaboratedthe ADL-3D-SAR main principle and the MIMO linear array design method which based on theprinciple of PCA technology.3.The signal model of ADL-3D-SAR based on single-transmit and multiple-receivelinear(STMR) array antennas is firstly derived in this paper. Secondly, a3D imaging algorithm fit forSTMR ADL-3D-SAR is proposed, and analytic expression is presented via a stepwise mathematicalderivation. Thirdly, the relation of received antenna elements spacing and elevation angular ambiguityis analyzed in this part. Finally, the signal model and3D imaging algorithm proposed in this paper aredemonstrated by simulation and The signal model of ADL-3D-SAR based on multiple-transmit andsingle-receive linear array antennas is conceived.4.According to ADL-3D-SAR two-dimensional top image intrinsically sparse structure, proposeda ship target detection algorithm of top image based on the scattering center. The top image （Due tothe present no conditions to obtain the top image so take approximate SAR image to instead of it）isfiltered by application of sparse dictionary learning method. through the scattering center modelconstruct the over-complete dictionary, the image target information project in two-dimensional spaceof frequency–azimuth, to achieve the image sparse representation. finally through clustering testmethod, the target pixel and the background pixel are classification. The experimental results showthat, the algorithm is better on detection effect of a certain threshold level target.5.Finally, the work of the dissertation are summarized, pointing out the shortcomings in the research process, and the next step research work is prospected.