An Investigation Of Ground Moving Target Detection Approaches For Wide-Swath Array Radar

Radar system is a key means to acquire information of targets and perceive the environments with its unique advantages,which plays an irreplaceable role in military and civil fields,such as early warning and surveillance,target tracking and recognition,remote sensing survey and mapping,resources monitoring,and so on.Ground moving target indication(GMTI)is one of the foremost tasks for radar system,and clutter suppression and target refocusing are dominated factors for improving the target detection performance.By introducing additional spatial degrees of freedom,the multichannel configuration can effectively improve the clutter suppression performance.However,in the real course of wide-swath data acceptance,the overlapping in the same range cell of echoes from different range regions due to the range ambiguity would complicate the clutter characteristics,which seriously deteriorates the clutter suppression and target detection performance.The traditional azimuth multichannel system focuses on the settlement of the azimuth ambiguity but not the resolving of range ambiguity and is hard to suppress the range ambiguous energy.At the same time,the complex motions of moving targets within a coherent integration interval,would cause the targets being severely smeared along range and azimuth directions,which degrades the target detection performance and the relocation accuracy,etc.Furthermore,the local heterogeneous background clutter and the low coherence of different channels would cause large amount of clutter residue,which futher degrades the target detection and parameters estimation performance.Aiming at the key issues of wide-swath ground moving targets detection and imaging in the array radar system,this dissertation mainly focuses on new techniques in aspects of range ambiguous clutter suppression,efficient target parameters estimation and refocusing,and robust clutter cancellation and target detection with channel errors.The main works of this dissertation are summarized as follows:1?Range ambiguity degrades the clutter suppression performance and incurs parameters estimation ambiguities in application of wide-area SCAN-GMTI.To deal with this issue,in this dissertation,an adaptive clutter suppression approach in data-domain based on element pulse coding(EPC)for multiple-input multiple-output(MIMO)radar system is proposed.The general signal model of EPC array radar is presented.The EPC array encodes the transmit pulses for each transmit element,which can introduce additional controllable degrees-of-freedom(DOFs)in coding dimension at the transmit elements.Then the distribution of range ambiguous echoes in transmit-receive spatial domains is studied.Besides,the echoes from different range-ambiguous regions can be well separated in joint transmit-receive spatial frequency domain by properly designing the EPC shift factor.In the sequel,the coding-angle-Doppler three-dimensional localization(3DL)technique is employed to suppress the range ambiguous clutters,which can evidently reduce the dimensions of the processor.Additionally,compared with the conventional clutter suppression methods,the proposed framework can effectively alleviate the nonhomogeneity of the clutter resulted from the range ambiguity.Numerical simulation experiments are provided to verify the effectiveness of the developed framework for range ambiguity suppression.2?To handle the issues of waveform separation and range ambiguity in the wide-swath SARGMTI application,we propose a coherent EPC-MIMO GMTI approach in image-domain.With the Doppler frequency shift property,the transmit waveform separation can be achieved at the receive end using digital beamforming technique.In the sequel,a set of transmit filters are employed to extract the echoes of each ambiguous region independently.Then the azimuth deramp operation is applied to the extracted data to focus the target energy of the desired region while the residual target energy of other regions is still smeared due to the mismatched reference function.After that,the adaptive matched filtering algorithm is adopted to suppress the clutter and detect the moving target.Compared with the traditional MIMO GMTI methods,the developed framework avoids the inter-symbol interferences of code diversity and the low coherence of frequency diversity,and increases the detection and imaging swath width by resolving the range ambiguity.Numerical simulation experiments are presented to demonstrate that the developed framework can obtain good results for waveform separation and range-ambiguous clutter suppression.3?Ground moving targets may be smeared in a SAR image due to the range migration,Doppler frequency migration and azimuth spectrum split caused by target unknown motion parameters.To handle these problems,we propose a fast non-searching method based on azimuth-deramp,second-order keystone transform(SOKT)and phase difference(PD).First,the azimuth-deramp is conducted,which is capable of eliminating the azimuth spectrum splitting.In the sequel,SOKT based on the scaling principle is applied to further correct the residual range curvature.Then,PD is performed to eliminate the range walk and estimate the motion parameters,which does not require the prior knowledge of Doppler ambiguity number.Finally,the moving target can be well focused with the estimated parameters.Compared with the conventional methods,this method is free of search and has high computational efficiency,which is suitable for the application of real time processing.Also,the proposed mehtod can solve the residual Doppler defocus problem resulting from spectrum split and along-track velocity,and thus obtain improved parameter estimation accuracy.Both simulated and real SAR data are processed to demonstrate the effectiveness of the proposed method.4?The range migration including range walk and range curvature may not be dealt with in one processing and the residual high-order range migration cannot be precisely compensated for the traditional methods.In addition,the residual range migrations caused by the different ambiguity components require to be accurately compensated one by one,which may greatly increase the system complexity.To tackle these issues,we propose a fast non-searching method based on range frequency axis reversal transform.With the symmetry and the equal interval sampling property of range frequency variable,a range compressed signal is initially transformed into the range frequency domain and then range frequency axis reversal transform is applied to directly compensate the range migration via multiplying the signal in the range frequency domain by its reversed data.The signal energy concentrates on one range bin when transformed into range time domain.Then the non-linear transform and generalized scaled Fourier transform are applied to estimate the motion parameters of the moving target.Finally,the moving target can be well focused after compensation.Compared with the existing methods,the presented algorithm can accurately eliminate the range migration including range walk and range curvature without the residual compensation error in one processing,which can better satisfy the real-time target focusing requirement.Also,the range migration caused by the velocity ambiguity components of multiple fast-moving targets can be solved simutaneously without searching procedure.Both numerical simulation experiments and real SAR data processing results demonstrate that the proposed algorithm has satisfied performance for target refocusing.5?In reality,the heterogeneous clutter as well as the residual channel errors resulting from image coregistration,channel balance and the terrain elevation phase compensation procedures,will remarkably decline the clutter suppression performance and the estimation performance of the target radial velocity.To address these issues,a robust radial velocity estimation algorithm is proposed.Based on the joint-pixel signal model,the joint-pixel normalized sample covariance matrix(JPNSCM)is employed to mitigate the effect of heterogeneous clutter,and then the shift vector determined by JPNSCM is used to modify the ideal target steering vector,which yields the actual target steering vector.Then,the adaptive matched filtering(AMF)algorithm is adopted to estimate the target radial velocity.Compared with conventional clutter suppression methods based on single-pixel model,the proposed framework makes full use of the coherence information contained in the adjacent pixels,and thus the problem of increasing degrees of freedom of clutter caused by non-ideal factors could be well tackled.Both simulated and real data are processed to demonstrate the robustness of the proposed method.