Study On Elevation/Azimuth Multichannel SAR Imaging

Higher resolution and wider mapping swath mean that SAR image can provide more information,which has been the constant pursuit in the development of SAR system.Due to the minimum antenna area constraint,it is difficult to realize high resolution and wide swath imaging simultaneously in the traditional single-input and single-output SAR system.Highresolution and wide-swath azimuth multichannel system and elevation multichannel system make it possible.However,channel error is inevitable in the actual work of multichannel systems.Before the array signal processing of multichannel system,it is necessary to estimate and to compensate the channel error,otherwise the beam can not be formed ideally.For the azimuth multichannel system,ghost targets of moving targets will appear in the final image due to the extra frequency offset introduced by their radial velocities,so moving targets need special processing.This paper concentrate on the study of the imaging technology of high-resolution and wide-swath azimuth multichannel SAR and elevation multichannel SAR,and the shortcomings of some of the existing algorithms are analyzed.The main contents of this paper are summarized as follows:In the second chapter,the geometric relationship between the platform and targets and the basic signal model are introduced.For the azimuth multichannel system,the moving target model is especially introduced,and the difference between the moving target signal and the static scene signal is analyzed,then two classical azimuth spectrum reconstruction algorithms and two channel error estimation algorithms are introduced.For the elevation multichannel system,the digital beamforming technology used to expand the mapping bandwidth is introduced.The following contents are discussed from these basic models.In the azimuth multichannel system,the Doppler shift introduced by the radial velocity of moving target is the reason why it can not be reconstructed correctly with the scene,resulting in ghost targets.In order to form the peak and zero at the correct Doppler point during reconstruction,it is necessary to estimate the frequency offset of a moving target,that is,the radial velocity estimation.In the third chapter,two radial velocity estimation methods are proposed.The first one operates in the data domain.Combined with the subspace theory,the frequency offset of the moving target is regarded as the rotation factor from the base of the stationary scene subspace to the base of its subspace.A group of projection matrices are constructed for the sampled Doppler cells to project the moving target signal into multiple subspaces,and the radial velocity estimation is obtained from the energy of the projected results.The influence of clutter and noise on the method and the time complexity of the method are analyzed.Finally,the method is verified by simulation data and airborne measured data.The second method operates in the image domain,aiming to obtain the real position information and the radial motion parameter information of the moving target directly from the image of the moving target and its ghost targets.It does not need to start the process from the echo data again.This part deduces the signal expression of the moving target and its ghost targets in the image after focusing,and analyzes the position of the moving target and ghost targets on this basis.Transforming the images of the moving target and its ghost targets into the Doppler domain,the radial velocity can be estimated by analyzing the weighted gain in reconstruction.What is more,the real position of the moving target is obtained.The influence of gain curve fitting order,clutter and noise on the method is analyzed.Finally,the simulation data and the airborne measured data are used to verify the method.In order to focus on the moving targets in the data and obtain the image without ghost targets,a complete set of azimuth multichannel marine moving target ghost target suppression scheme is proposed in Chapter 4.The processing steps include clutter suppression,detection,moving target data extraction,moving target imaging,ghost target suppression and moving target relocation.A clutter suppression method is proposed for the moving target detection of azimuth multichannel system with small number of channels.This method also combines subspace theory.Prior platform flight speed and channel arrangement information are used to construct the expected clutter signal subspace and its projection matrix.Projecting the received data of each channel into the clutter subspace,and then the clutter suppression results are obtained by subtracting the projecting results from the received data.Due to the Doppler frequency offset of radial velocity,the subspace of each moving target is different from the clutter subspace.The result of subtraction is the signal component of each moving target signal which is different from the clutter subspace.The clutter suppression is completed and the moving target signal is highlighted.The blind velocity value of the method and the influence of residual channel error on the method are analyzed.Then,the algorithm used in each step of the scheme is introduced,and the reason why there is no ghost target in the deramp azimuth focused image of undersampled single channel data is introduced.The relationship between the number of ghost targets in the high-resolution and wide-swath image,the reconstructed ratio and the ambiguity number is analyzed.Finally,the scheme is verified by simulation data and dual channel satellite measured data.Elevation multichannel system can expand the mapping bandwidth and ensure the gain of the scene echo,but the channel mismatch problem inevitably exists in the receiving channel,which makes the synthetic beam gain not meet the expectation.In Chapter 5,a digital beamforming method combined with channel equalization is proposed,which combines the ground digital elevation information and platform geometry as the prior knowledge to calculate the direction of the incoming wave.First the received signal model of the channel with errors is introduced.The amplitude and phase errors of the signal are extracted by division and interference,respectively,and the noise is suppressed by adaptive weighted rotary mean filter.The channel equalization is combined into the process of digital beamforming after range pulse compression,and the method is verified by airborne measured data.The differences between elevation channel error estimation and azimuth channel error estimation are also analyzed.