Imaging Algorithm And Position Method Study On Bistatic/Multistatic SAR

Bistatic forward-looking synthetic aperture radar(BF-SAR)is a SAR system whose transmitter and receiver are spatially separated,in which the transmitter works in the side-looking mode,and the receiver in forward-looking one.It can not only perform two-dimensional(2D)imaging with high-resolution to forward area that breaks through the inherent limitations of the conventional monostatic SAR,but also extend the application of SAR system in terminal guidance period of missile.Besides,this separation improves the system's flexibility and electromagnetically concealing,and the radar cross section(RCS)of the target from different directions by adjusting the bistatic angle could be obtained which can enhance the ability of imaging processing and target extraction.Therefore,BFSAR has been an important research subject.Compared with conventional SAR signal,BFSAR signal has some new characteristics,such as,the loss of azimuth translation invariance,the couple relationship in range between transmitter and receiver,and vague definition of azimuth direction,which will make conventional imaging methods invalid.In this dissertation,based on the properties of BFSAR mounted on maneuvering platforms,some key points on imaging algorithms and system applications are discussed.This thesis specifically includes the following parts.1.Slant-distance model is the fundamental of SAR signal analysis and processing,whose accuracy of expression and explicit representation of physical parameters are closely related to subsequent design of imaging algorithm.Firstly,by taking bistatic translation invariant mode and bistatic translation variant mode as the standard,the non-approximate slant range model and the monostatic equivalent slant range model under bistatic translation invariant(TI)mode are summarized.Meanwhile,azimuth definition and equivalent conditions under this configuration are defined explicitly.The invalidation of azimuth translation invariance is explained by using monostatic translation variant mode.Secondly,aiming at the unavailability of definition about range and azimuth in translation variant mode,the traditional non-approximate slant range model with bistatic arbitrary configuration is used to explain this problem,and "point-to-point" non-approximate slant model and its corresponding high-order polynomial slant model are proposed.Next,in multistatic SAR systems with arbitrary curve trajectory,a fourth-order approximate expansion model is proposed under the smallest work unit,that is,bistatic SAR,and high-precision 2D spectrum is obtained by series inversion.And this part lays a theoretical foundation for design and application of subsequent imaging algorithms in frequency domain and beam domain.2.Different from the trajectory design methods of airborne or missile-borne BF-SAR,the resolution constraint in one single direction is not suitable for airplane-missile bistatic forward-looking SAR(AMBFL-SAR)in which there are the requests for high real-time performance and low maneuverability and the problem of great differences of dynamic performance between transmitter and receiver.Thus,an aircraft trajectory design method based on pre-planned trajectory of missile with constraints of 2D resolution and resolution cell area is first proposed.According to Doppler contribution of aircraft-missile platform under different missile-to-target distance and the influence of flight parameters on system resolution,the maximum rotation angle between aircraft and target is obtained under the assumption of uniform motion.Then,the trajectory of the aircraft is optimized by using orbital angle and initial squint angle of the aircraft as input variables to ensure that the projectile meets the imaging requirements in the whole trajectory.Finally,the effectiveness of the proposed algorithm is verified by simulation experiments.3.Aiming at the problems of range linear migration,two-dimensional spatial variation of azimuth frequency modulation term and image distortion caused by cross-track configuration in bistatic forward-looking SAR imaging during flat flight,an improved azimuth nonlinear chirp scaling algorithm based on azimuth focusing frequency point in frequency domain and inverse projection geometric deformation method are first proposed in this paper.Firstly,keystone transform is introduced to correct the most linear RCM.Then,different from existing methods,through imaging geometrical configuration,the proposed algorithm presents the method of building the spatial variant functions of Doppler parameters in terms of focus position to realize the equalization of the dependent azimuth modulation term.Finally,a back projection geometric deformation correction method corresponding to the imaging algorithm is proposed,which provides convenience for subsequent image matching and recognition.Simulation results and real data processing further validate the effectiveness of the proposed algorithm.4.For curve trajectory of BFSAR,there are few researches about wavenumber-domain algorithm.For this case,a modified Stolt interpolation imaging algorithm based on azimuth resampling technique is proposed.This method not only corrects the range dependence of RCM,but it also establishes the azimuth variable model of azimuth phase based on frequency focusing position.Firstly,the linear range cell migration correction in time domain is implement to realize the “no squint” procedure.Meanwhile,range linearization of 2D spectrum is accomplished by high-order polynomial fitting.For combining with motion compensation and avoiding zeros-padding operation caused by time-domain focusing,we choose to focus image in wavenumber domain,and solve the problem that azimuth frequency modulation term is variant spatially through azimuth resampling.The proposed wavenumber domain algorithm paves the way for the future research in BFSAR with curved trajectory.5.For the application of BFSAR system in terminal image-guidance image,a geometrical location method based on the R-D principle is proposed.First,construct the geometrical model between the transceiver platform and the target at the time of center of adjacent synthetic aperture.Then the position information of the target relative to receiving platform can be solved by combining the relationship between the target in the SAR image and the center point of scene.And according to the characteristic of BFSAR system,the “space-time-frequency” synchronization error models are built.What's more,we analyze in detail how the model influences the proposed algorithm,and deduce the constraints relationship between positioning error and synchronization error through simulation calculation.Simulation results show the validity of the positioning method and the proposed error models.