Study On New Techniques For ISAR Imaging Of Aerospace Targets

Aerospace targets include aerial targets and space targets. The aerial targets are flying in the atmosphere, such as air-planes and airships; and the space targets are orbiting in the extra-atmospheric space which is at least 100km beyond the earth surface, such as man-made satellites and space debris. Inverse synthetic aperture radar (ISAR) imaging is able to provide motion parameters and structural features of aerospace targets, thus makes great contributions to target classification, recognition, and cataloging. Therefore, ISAR imaging of aerospace targets plays a vital role in national security and development. Although techniques for ISAR imaging have been fully developed in recent years, there still remain many problems due to the particularity of aerospace targets. This dissertation studies new techniques for ISAR imaging of aerospace targets from three aspects, i.e. imaging of targets with rotating parts as well as spinning targets, imaging from sparse bands and sparse apertures, and imaging of group targets. The relevant work is supported by the National Science Foundation of China (No.60802081 and No.60890072) and the National High Technology Research and Development Program of China (No.2008AA8080402).The main content of this dissertation is summarized as follows.The first part introduces fundamentals of ISAR imaging briefly and summarizes the available basic imaging techniques.The second part focuses on echo separation and ISAR imaging of targets with rotating parts. The imaging model is created and a procedure for image acquisition of both the rigid body and rotating part is proposed. For echo separation of the rigid body and rotating parts, two methods are proposed, which are based on the complex-valued empirical mode decomposition (CEMD) and low chirp rate matching. For two-dimensional imaging of the rotating parts, the methods based on real-valued inverse Radon transform (RIRT) and complex-valued inverse Radon transform (CIRT) are proposed respectively. Additionally, their performance is analyzed and compared; and possible problems in practical use are discussed and effective solutions are provided. The validity of the proposed methods is proved by simulated and measured data.The third part focuses on narrow-band ISAR imaging of spinning targets. The narrow-band complex-valued back-projection method is proposed and its resolution is analyzed. For azimuth down-sampled echoes of a fast spinning target, an imaging method based on orthogonal matching pursuit (OMP) is proposed according to the theory of compressed sensing (CS). By this means, the radar ability in imaging and recognition is improved. Finally, this method is extended to wide-band radar imaging of fast spinning targets.The fourth part focuses on wide-band, three-dimensional (3D) imaging of spinning space targets. Analysis indicates that the phase and envelope of the spinning scatterer are sinusoids in the range-slow time domain. Then, the wide-band complex-valued back-projection method is proposed. Due to coherent accumulation, this method is high in resolution and is robust to additive noise. Based on its electromagnetic feature, the signal model of a smooth precession cone is derived, and a dynamic electromagnetic simulation method is introduced. According to the sinusoidal characteristics of its echoes, the 3D image of the cone is obtained from simulated data via the wide-band complex-valued back-projection method. To achieve image scaling of the 3D image, a method based on the bistatic ISAR configuration is proposed.The fifth part is contributed to ISAR imaging utilizing sparse bands and sparse apertures. For sparse band imaging of space targets in high speed motion, the influence of a large velocity on range and azimuth images is studied and a compensation method is proposed. Then, an effective method is proposed for sparse band imaging based on gapped data filling and scatterer parameter estimation. For sparse apertures, a practical imaging procedure is proposed based on characteristics of the two data missing patterns, i.e. the gapped data and the random missing data. For the gapped data, a new method for range alignment is introduced. According to the theory of sparse signal reconstruction, different imaging methods are chosen according to the mutual coherence of the over-complete dictionary. For a small mutual coherence, the imaging method based on basis pursuit (BP) is proposed. For the gapped data with a large mutual coherence, the gapped data amplitude and phase estimation (GAPES) method is applied to azimuth imaging. Simulated and measured data have proved the effectiveness of the proposed methods.The sixth part focuses on ISAR imaging of group targets. A parametric method is proposed for imaging of targets moving in a formation with constant accelerated rectilinear motion. This method can achieve accurate and well-focused imaging due to parametric motion compensation. Additionally, the computational complexity of parameter estimation is reduced since there is no need to compensate the first-order phase term of each sub-target respectively. For spinning group targets with uniform motion, an effective imaging method combined with motion compensation is put forward. Besides, this method is able to estimate the angle dependence of the scatterer's back-scattering coefficient, thus can provide a more comprehensive description of the target.