Design And Implementation Of SAR Imaging Algorithm For Micro Millimeter-wave Radar

Synthetic aperture radar(SAR)imaging is an active microwave remote sensing technology.Compared with other imaging sensors,SAR can work stably in harsh conditions such as night,smoke,rain,snow and fog,which is not limited by working time and environment.Meanwhile,with the properties of wide area detection and high imaging resolution,SAR has become an irreplaceable technology of modern remote sensing and mapping.Due to the capability of data acquisition and information perception,SAR has been widely used in military fields,such as battlefield monitoring,reconnaissance and strike,and high-maneuvering target imaging,and also in civil applications,such as resource surveying,terrain monitoring,and high-precision mapping.The traditional SAR imaging is mainly applied to spaceborne and airborne platforms,and is generally used in long remote scenes.Depending on the beam steering,polarization mode and the formation of antenna,SAR has been equipped with various functions and operating modes.Recently,with the development of semiconductor process and integrated circuits,the effective combination of SAR and micro millimeter-wave radar has derived many new applications,including UAV mapping,automotive assistant driving and security imaging.Because of the new application background,the imaging geometry,signal model and imaging processing of SAR have dramatically changed.Firstly,millimeter-wave SAR generally works at short range and wide viewing angle,which yields great challenges in imaging and focusing.Secondly,with the high observation efficiency,millimeter-wave radar is capable of video SAR imaging,which needs to be studied for the practical implementation.Finally,two-dimensional SAR images have geometric distortion problems,such as scene overlapping,so it is necessary to study three-dimensional SAR imaging algorithms to restore the real three-dimensional scenes.This thesis mainly focuses on development of micro millimeter-wave SAR imaging algorithm,including the high-resolution two-dimensional(2D)imaging,video SAR imaging and threedimensional(3D)SAR imaging.On this basis,the millimeter-wave SAR imaging algorithm and processing are packaged in software,and the micro-SAR imaging function is basically realized in engineering.The main research content of the thesis is summarized as follows:1.At the beginning,the application background and development history of SAR are briefly introduced,and the problems of traditional SAR imaging are pointed out.Then,the necessity of micro millimeter-wave SAR is illustrated.After that,we review the current research status of system development and imaging algorithm related to micro-SAR at home and abroad,and enumerate the main challenges faced by micro-SAR imaging.2.Starting with the basic theory of frequency modulated continuous wave(FMCW)SAR imaging,the imaging geometry and signal model of micro-SAR have been established.Two frequency domain imaging algorithms suitable to dechirp mode are deduced,and the imaging quality of point targets is analyzed using simulation experiments.With accurately estimating the doppler parameters of the echo data,we utilize the relationship between the doppler parameters and the motion parameters to invert the motion parameters of the SAR imaging model,which can be used to accurately compensate the motion errors.3.According to the requirements of video SAR imaging,a video SAR imaging algorithm based on time-domain back-projection(BP)algorithm is proposed.Firstly,the key steps in the processing of BP imaging are emphasized.Then,the frame refresh rate of video SAR is effectively improved by the overlapping sub-aperture method.The experiment of simulated data and measured data verify the feasibility of the algorithm.Finally,the proposed BP algorithm is successfully applied to circular SAR imaging,and the experimental results of public datasets have illustrated the effectiveness of the algorithm.4.Aiming at the geometric distortion of 2D SAR imaging,the 3D SAR imaging technology is studied.For micro-SAR,the third dimensional imaging in height direction is based on traditional2 D imaging processing.The digital beamforming algorithm suitable to uniform sampling and sparse signal processing method applied to non-uniform sampling are studied,and the experiments using simulated data and public dataset are performed to confirm the effectiveness of the proposed algorithm.5.The flowchart of millimeter-wave SAR imaging processing is cumbersome,which is not easy to maintain and widely transplanted to multiple platforms.Therefore,a millimeter-wave SAR imaging software suitable to vehicle platform is designed by hybrid programming technology of Python and Qt.Real-time and efficient SAR processing is realized by logic control and automatic data transmission.