Study On Polarimetric Radar Imaging And Target Feature Extraction

Sponsored by the State Key Development Program for Basic Research of China (973) of National Security (New Mechanism and New Methods of Radar Automatic Target Recognition), the Major Program of the National Natural Science Foundation of China (Radar Polarization Information Acquisition and Processing) and National Defense Pre-Research Foundation of China (Target Polarization Character of High-resolution Radar), this dissertation mainly focuses on the research of polarimetric radar imaging and target feature extraction. The research topics including the polarimetric super-resolution imaging, feature extraction of target scattering center and perform analysis of the extraction, SAR imaging formats under the typical jamming background. The most significant point is the idea and method, which is brought out based on the coherent polarization scattering model and high-resolution polarization vector signal processing, is a breakthrough compared with the traditional multi-channel processing method. This method can not only estimate the feature parameters more correctly, but also resolve the difficult problems in the traditional method including association processing of center locations and polarizations more effectively. So it is more effective and has better performance comparing with the traditional method. Considering the complex electromagnetic environment in the future warfare, the PolSAR imaging method under the random polarization jamming and active decoy condition is also studied.Chapter 1 introduces the current development of the radar polarization theory and methodology, especially the imaging and feature extracting techniques of polarimetric radar. The major research topics of this dissertation are explained in this chapter.Chapter 2 studies the feature parameter estimation of the fully polarized scattering center and analyses its performance. Firstly, the explicit expressions of the Cramer-Rao bounds (CRB) of fully polarized scattering centers are presented for the first time. Moreover, the CRBs of amplitude, phase and phase difference between two components of the scattering matrix are derived and expressed in closed forms for several typical cases together with their relation with the radar system parameters (the band width, SNR and polarization etc). Thereby, two novel methods of fully polarized scattering center extraction and parameter estimation, which called P-MUSIC and P-ESPRIT, are proposed. The validity of these two methods is proved by the experimental results based on simulated and real data.In Chapter 3, the parameter estimation method of the fully polarized scattering center under the dispersive condition is studied and its performance is analyzed. The CP-GTD can represent the frequency dispersion characteristics and the full-polarization characteristics of the target scattering center simultaneously. A novel scattering center extraction and parameters estimation method is proposed based on the CP-GTD. The performance of this method is evaluated using coherence and incoherence index and testified by the simulation and testing data. The superiority of the CP-GTD is verified in theory. The relation between the estimation performance of the model parameters and the band width, SNR and polarization etc is studied. This study provides a meaning theory guide to the engineering practice.Chapter 4 focuses on the problems of the fully polarized super-resolution imaging of SAR/ISAR and the target feature extraction. Firstly, the two-dimensional CP-GTD (2D CP-GTD) is established and the methods of super-resolution imaging and features extraction for polarimetric ISAR are studied based on 2D CP-GTD. In the polarimetric complex imaging space, the super-resolution imaging and target scattering center extracting methods of polarimetric SAR (PolSAR) are proposed based on the coherent polarization idea point scattering model. These methods are validated using simulation data, data measured in microwave anechoic chamber and practice measured PolSAR complex imaging data. Unlike the traditional method, these methods can estimate coherent scattering matrices of scattering centers accurately by overleaping the step of scattering center association. They also can estimate the elements of the coherence polarization scattering matrix simultaneously.In Chapter 5, the polarimetric radar imaging method under the random polarized suppression jamming and the active decoy condition is studied. The concept and method of serial polarization transform (PT) is proposed to suppress the random polarized blanket jamming. The suppression method is proposed based on the PT filter, which makes a breakthrough on the suppression effect of the traditional polarization filter. The satisfied anti-jamming imaging result is realized based on this method using the practice SAR data. The PolSAR image method, which can realize the active decoy distinguish and cancelling, is proposed to suppress active-decoys jamming. This imaging method is realized by phase compensation in Doppler domain. Its validity is proved by good imaging results under active decoy condition using the simulation and experimental SAR data.