Research On Multi-Channel High-Resolution And Wide-Swath SAR Imaging And Motion Error Compensation Technology

Synthetic Aperture Radar(SAR)is widely used in military reconnaissance,resource exploration,disaster warning,and mapping because of its ability to perform microwave remote sensing imaging in all-day and all-weather.High-resolution wide-swath(HRWS)SAR can obtain high-resolution and wide-swath images at the same time,which can improve imaging quality and efficiency and is the goal that SAR has been striving to develop.However,conventional single-channel SAR systems are difficult to obtain both high resolution in azimuth and wide swath in range because of the conflicting constraints of a low pulse repetition frequency(PRF)to avoid range ambiguity and a high PRF for azimuth high resolution.To solve this problem,the azimuth multi-channel SAR system has been studied in recent years.This system can avoid range ambiguity by transmitting a lower PRF signal and use azimuth multi-channel to receive signals simultaneously.It can increase azimuth spatial sampling frequency by rearranging multi-channel signals to avoid azimuth ambiguity.However,in practice multi-channel SAR systems inevitably suffer from non-uniform sampling,resulting in azimuth ambiguity.In addition,due to the inconsistency of multiple channels and antenna position deviation,the multi-channel system has errors such as amplitude error,phase error and baseline error,which lead to the performance degradation of the reconstruction algorithm.The spatial variation of the phase errors and multi-channel motion errors are difficult to compensate for,resulting in degraded image quality.Therefore,this thesis conducts an in-depth study on the problems of nonuniform sampling,channel errors and motion errors in multi-channel SAR imaging.The main contents and innovations are as follows:1.In order to solve the problem of azimuth nonuniform sampling,image domain reconstruction methods based on backprojection are proposed.A reconstruction model based on time-domain interpolation is first derived,and an image domain reconstruction method(IDR)based on time-domain backprojection(BP)is proposed.The method uses multi-channel echo data to construct sub-image echo data and uses the BP algorithm to process sub-image echo data.After that,it uses the reconstruction coefficient to weight the sub-image in the image domain to obtain the unambiguous image.The advantage of this method is that reconstruction is performed in the image domain,simplifying the reconstruction process to image fusion and facilitating the updating of reconstruction results.It can be applied to HRWS SAR imaging of maneuvering trajectory that satisfies certain constraints.The experimental results show that IDR can effectively suppress azimuth ambiguity.In particular,compared with the interpolation reconstruction algorithm and transfer function algorithm,the proposed IDR can effectively achieve the reconstruction of curved motion trajectory,expanding the scope of application.In order to improve the efficiency of the IDR algorithm,the frequency-domain BP based image domain reconstruction(FDBP-IDR)method is proposed.The method first constructs the sub-image echo spectrum,then calculates the backprojection index of the image spectrum and obtains the image spectrum of each sub-image by backprojection in the frequency domain.Finally,the image spectrum of the sub-image is transformed into the image domain and the reconstruction is performed.The method uses the frequency domain BP principle to complete the backprojection of the echo data in the frequency domain,and reduces the calculation of the backprojection index based on the feature that the index is the same for each sub-image spectrum.The simulation experimental results show that compared to IDR,FDBP-IDR improves the computational efficiency while maintaining image quality.2.For the problem of reconstruction quality degradation due to channel amplitude,phase and baseline errors,a sub-image local area minimum entropy reconstruction(SILAMER)method is proposed.The problem of degradation of IDR reconstruction quality when channel amplitude,phase,and baseline errors are present simultaneously is analyzed and the SILAMER method is proposed.The method first uses a backprojection algorithm to generate the sub-image.An optimisation estimation model is then used to estimate the optimal reconstruction coefficients and channel compensation phases by minimizing the reconstruction image entropy in local regions of the sub-image.The choice of local regions of each sub-image as input to the estimation algorithm rather than the whole region greatly improves the estimation efficiency.Finally,the estimated values are used to weight the sub-image and correct the phases to obtain an unambiguous image.The experimental results show that the proposed SILAMER achieves unambiguous reconstruction for multi-channel SAR with multiple errors.Moreover,the method can achieve curved trajectory SAR reconstruction with multiple errors and still performs well under the condition of low signal-to-noise ratio.3.To solve the problem of multi-channel SAR motion errors,a high-precision motion errors compensation method based on sub-image reconstruction(SI-MEC)is proposed.The spatial variation of the phase errors caused by platform motion errors in HRWS SAR imaging and the difference in motion errors between channels are first analyzed.These factors reduce the performance of the conventional motion errors compensation methods.Therefore,SI-MEC is proposed for the multi-channel SAR motion errors compensation.First,the method performs primary imaging and selects the location of the strong points in multiple regions of the image.A motion errors estimation model is developed which estimates the motion errors of the platform by maximizing the intensity of the strong points.To complete the motion errors compensation for each channel,the equivalent phase centre(EPC)corresponding to each channel sub-image is then corrected according to the multi-channel geometric structure relationship,and sub-image imaging is performed.Finally,the reconstruction is completed by weighting the sub-images to obtain a high-quality image.The method uses multiple strong points in multiple regions to estimate the motion errors,which can obtain a more accurate EPC correction value,and then combines with the BP algorithm to achieve the phase errors compensation for each pixel,which improves the ability to compensate for spatial variation phase errors.In addition,the method compensates for the different motion errors of each channel before reconstruction,which improves the ability to compensate for differential motion errors between channels.The experimental results show that the proposed SI-MEC method has the lowest image entropy compared with PGA and BP-FMSA,and improves the HRWS SAR imaging quality when suffering from motion errors.