High-Resolution And Wide-Swath Multi-Channel SAR And Moving Target Imaging Theory And Methods

With superiority in high-resolution and wide-swath (HRWS) observation, ground moving target indication (GMTI), digital elevation model (DEM) generation, interference suppression etc., multi-channel synthetic aperture radar (MC-SAR) system has attracted much attention in radar imaging and remote sensing domain, and becomes a research focus at home and abroad. HRWS observation of earth is one of the important trends for the future development of MC-SAR systems, especially the space-borne MC-SAR system. For the conventional single-channel SAR system, in order to achieve a wide-swath imaging, the radar transmitter pulse repetition frequency (PRF) should be low enough to avoid the serious range ambiguity. However, to obtain a high azimuth resolution, which indicates a wide Doppler bandwidth, the PRF should be high enough to ensure that the Doppler spectrum is ambiguity-free. This contradiction is called "Minimum Antenna Constraint". Fortunately, the MC-SAR combined with digital beam-forming (DBF) has been adopted to resolve this contradiction effectively and achieve HRWS imaging, where the system transmits chirp signals with a low PRF to avoid the serious range ambiguity, and all of the channels receive the echo simultaneously.To satisfy the new trend of HRWS observation and solve the key problems in MC-SAR imaging, this dissertation studies new techniques for MC-SAR imaging in some aspects, i.e. channel-calibration, Doppler parameters estimation, squint mode imaging processing, clutter suppression and imaging processing for moving target, interference suppression etc.. The relevant work is supported by the "973" Program (No. 2010CB731903), the National Science Foundation of China (No.60802081, No. 60890072 and 61222108) and the Doctoral Foundation (No.200807010002).The main content of this dissertation is summarized as six parts, which are shown as follows:(1) The echo model of MC-SAR system and channel mismatch modelIn Chapter 2, the echo model of MC-SAR system is studied. Firstly, the echo of MC-SAR system is modeled on the ideal cases. Taking the channel mismatch for the real MC-SAR system, i.e. gain-phase error, range sampling time error as well as the error of antenna position measurement etc. into consideration, we discuss the channel mismatch with MC-SAR echo. At the same time, the error of antenna position measurement is divided into the error of azimuth baseline measurement and the range-variant channel mismatch in phase. Then, the available channel-calibration approaches are introduced, which include subspace projection based channel-calibration approach and cross-correlation coefficient based channel-calibration approach etc. In addition, the azimuth ambiguity-free Doppler spectrum reconstruction approaches are discussed, i.e. transfer function reconstruction approach and space-time adaptive processing (STAP) reconstruction approach etc. Additionally, a multi-Doppler-direction restriction ambiguity suppression approach is proposed. The airborne real measured ScanSAR data, which is acquired by a seven-channel in azimuth SAR imaging system working at X-band, is utilized to demonstrate the performance of the new Doppler ambiguity suppression algorithm.(2) Channel-calibration techniques for range-variant channel mismatch of MC-SAR systemIn Chapter 3, the channel-calibration approaches for range-variant channel mismatch are discussed during the HRWS SAR imaging processing. At first, a subspace based range-variant channel mismatch calibration method is proposed for the MC-SAR system. During the channel-calibration, the mismatch between the channels, which results from the gain-phase error and the range sampling time error, is corrected by the coarse-calibration processing in the range frequency domain. Then, the error of along azimuth baseline measurement is estimated and the data are transformed into the range time domain and azimuth Doppler domain (RD domain). Considering the range-variance components in the residual phase error, the data are processed in blocks along range and the error of every sub-block data is estimated. After that, a fitting and filtering operation is implemented along range to get the estimated values of the phase error of all sub-blocks. The range-variant phase error is then compensated using its estimated value. The processing results of airborne real measured seven-channel SAR data are utilized to demonstrate the performance of the proposed channel-calibration method.Another range-variant channel-calibration is also proposed in Chapter 3, which is based on local maximum-likelihood weighted minimum entropy (LML-WME). The proposed algorithm is implemented in two steps:(a) the timing uncertainty in each channel and most of the range-invariant channel mismatches in amplitude and phase have been corrected in the pre-processing of the coarse-compensation; (b) After the pre-processing, there is only residual range-dependent channel mismatch in phase. Then, the retrieval of the range-dependent channel mismatch in phase is achieved by a LML-WME approach. The simulated multi-channel in azimuth SAR data experiments is adopted to evaluate the performance of the proposed channel-calibration algorithm. Then, the real measured airborne five-channel HRWS SAR data is utilized to demonstrate the effectiveness of the proposed approach.(3) Squint mode MC-SAR imaging processing techniqueIn Chapter 4 of this dissertation, the echo model of squint mode MC-SAR system is introduced. Then, a squint mode MC-SAR imaging approach is proposed, which is based on Doppler centroid estimation and a two-step focus processing. For the Doppler centroid estimation, a robust entropy-based Doppler centroid (REDC) estimation approach is proposed, where the baseband Doppler centroid is estimated by combining the cross-functions between neighboring channels in range-frequency domain. Then, an entropy-based Doppler ambiguity number resolving approach is proposed and the bisection algorithm is utilized to accelerate the searching process. During the imaging process, the linear range cell migration correction (RCMC) is firstly employed to correct the skew Doppler spectrum for each channel echo. After the azimuth signal reconstruction, a two-step focus approach is utilized to focus the image, the first step is to process the azimuth-invariance components and the second step is to correct the azimuth-variance component brought by the linear RCMC. Some simulation experiments are taken to verify our proposed algorithm. In addition, the experimental results of some real MC-SAR data also show that the proposed method works well.(4) Moving target imaging technique for HRWS MC-SAR imaging mode and DPCA conditionIn Chapter 5 of this dissertation, the moving target echo model of MC-SAR system for displaced phase center antenna (DPCA) condition is studied and a novel moving target imaging approach based on local maximum-likelihood minimum entropy (LML-ME) is proposed. The proposed approach utilizes the wide bandwidth characteristics of the transmitted signal (multiple wavelengths) to estimate the moving target velocity. Firstly, there is a phase mismatch (PM) between azimuth channels for the echo of a moving target, which depends on range frequency. In order to correct PM, an algorithm based on LML-ME is proposed. The linear dependence of PM on range frequency is employed to estimate the target velocity. Secondly, after azimuth signal reconstruction using the DPCA condition and compensation of PM for a moving target, the estimated target velocity is utilized to implement the linear RCMC and the Doppler centroid shifting. Then, the quadratic range cell migration (QRCM) is corrected by the keystone transform processing. After that, the focused moving target image can be obtained using the existing azimuth focusing approaches. Theoretical analysis shows that no interpolation is needed. The effectiveness of the imaging algorithm for moving target is demonstrated via simulated multi-channel SAR data and real measured ship five-channel SAR data.(5) Robust clutter suppression and moving target imaging technique for HRWS mode MC-SAR systemIn Chapter 6 of this dissertation, a robust clutter suppression and moving target imaging approach for HRWS mode MC-SAR system is proposed. Firstly, the clutter and moving target echoes are transformed into the range compression and azimuth chirp Fourier transform (CFT) frequency domain (RC-CFT), i.e., coarse-focused images are obtained, when the clutter echoes of each channel are with azimuth Doppler ambiguity. Considering that moving targets are sparse in the imaging scene and that there is a difference between clutter and a moving target in spatial domain, a series spatial domain filters are constructed to extract moving target echoes. Then, using the extracted moving target echo, two groups of signals are formed and slant-range velocity of a moving target can be estimated based on baseband Doppler centroid estimation algorithm and multi-look cross correlation Doppler centroid ambiguity number resolving approach. After the linear RCMC and azimuth focus processing, a well-focused moving target image can be obtained. In addition, the proposed clutter suppression and imaging approach are not only adapted for uniformly displaced phase center sampling, but also for the non-uniform sampling cases. Some simulation experiments are taken to demonstrate our proposed algorithms. Finally, some real measured data results are presented to validate the theoretical investigations and the proposed approaches.(6) Interference suppression technique for MC-SAR imaging processingIn Chapter 7 of this dissertation, the interference signal characteristic is analyzed for the MC-SAR system. Then, an interference suppression approach for SAR based on time-frequency transform is proposed. The goal of the proposed approach is to suppress the narrowband interference (NBI) and wideband interference (WBI) in MC-SAR system by a non-parametric method. The short-time Fourier transform (STFT) is utilized to estimate the instantaneous frequency of the MC-SAR echo data with interference. In the STFT domain, the instantaneous frequency spectrum is represented by wavelet, and then the designed filter with the constant false alarm rate (CFAR) filters the corresponding wavelet coefficients of the interference components. In addition, the proposed algorithm is robust for time-varying NBI and WBI. The performance of the proposed approach is evaluated by the simulated and measured data, and the effectiveness is demonstrated.Another interference suppression approach for deception interference is also proposed, which is based on digital beam-forming (DBF) technique. Since there are some differences between scene echo and deception interference signal for the MC-SAR system in the spatial domain, the DBF approach can suppress the deception and hold the energy of scene echo. The effectiveness of the proposed approach is validated using some simulation experiments.