| Inverse Synthetic Aperture Radar (ISAR) imaging technology has played an irreplaceable role in both military and civilian fields, mainly attributed to its capabilities of all-weather, all-day, long-distance work, high-resolution and etc. ISAR is mainly applied to image and identify the non-cooperative targets in airspace, aerospace and sea. With the increasing of application requirements, the manners of data acquisition of ISAR systems developed to multi-function, multi-mode, multi-view and multi-polarization. The complicated working patterns challenge the current high-resolution imaging. Current high-resolution radar imaging techniques face several problems, i.e. motion compensation, two dimensional coupling and polarization information integration, and target backscattering field description.This dissertation focuses on studying new techniques to improve resolution of polarimetric ISAR and orthogonal frequency division multiplexing (OFDM) ISAR with compressive sensing (CS) and electromagnetic scattering theory. The relevant work is supported by National Basic Research Program of China (973 Program, No.2010CB731903), National Science Foundation of China (No.61301280) and others.In this dissertation, we put our efforts to improve the resolution of ISAR imaging under different working patterns. The main content of this dissertation can be summarized as follows:1. ISAR super-resolution imaging for short-aperture and limited bandwidth dataWhen wideband ISAR acquires short-aperture and limited bandwidth data, the non-cooperative target motion approximates stationary and the imaging signal processing is simple, which can effectively overcome the difficulties of resource constrained time and frequency to a single target of multi-function radar. However, imaging with short-aperture and limited bandwidth data might affect the resolution of image, which should be considered in the imaging processing. In the first part of the third chapter, we extend the CS-based short-aperture ISAR imaging approach to the united two-dimensional one, which fully and effectively utilizes the two-dimensional coupling information by modeling the two-dimensional super-resolution imaging as an l1 norm optimization problem. The super-resolution image can be achieved through a fast iterative method combined with fast Fourier transform (FFT), the conjugate gradient algorithm (CGA) and the Hadamard product. Results approve that the CS-based two-dimensional ISAR super-resolution imaging can utilize the two-dimensional coupling information to reconstruct the super-resolution ISAR image with strong noise robustness.2. Polarimetric super-resolution ISAR imaging and autofocusingPolarimetric ISAR imaging is defined as using fully polarimetric radar to image artificial target, such as satellites, space shuttles, aircrafts, missiles, etc. The vector characteristics of electromagnetic waves bring rich redundancy and complementarity information of the target backscattering signal in different polarization channels. By analyzing backscattering characteristics based on polarimetric scattering mechanisms, we can extract important feature information of the target. The paper will be carried out to develop the fully polarimetric super-resolution ISAR imaging and autofocus method. First, combining the polarization signal characteristics, the two-dimensional super-resolution algorithm is extended to the polarimetric radar in the second part of Chapter 3. For the anisotropy of scattering centers, we present a method to obtain the super-resolution imaging while holding polarization information. By combining echoes in different channels, the method can insure the continuance of scattering centers in different polarization channels, which is beneficial for the extraction of the polarized scattering matrix. Then, in the fifth chapter, we propose a full polarimetric autofocusing algorithm for sparse aperture data. The combination of polarization information during autofocusing can effectively improve the estimation accuracy and strong noise tolerance. The experiments using the backhoe data show that the polarimetric super-resolution ISAR imaging and autofocus algorithm can utilize the redundant polarization information to improve the imaging resolution and accuracy of phase compensation.3. Super-resolution imaging based on attributed scattering center modelISAR imaging can precisely describe the electromagnetic scattering distribution of the echoes in specific azimuth angle and frequency. Under certain observational condition, there is a clear correspondence between target geometry and signal electromagnetic scattering distribution. In Chapter 4, the attributed scattering center model which can accurately describe the correspondence relation is simplified and a fast property scattering center model parameter estimation method is proposed to reconstruct the target geometry and characteristics. Then we utilize the estimated parameters of the scattering centers to extrapolate spectra, which can break the theoretical limit of resolution. During super-resolution imaging in azimuth dimension, by selecting appropriate center angle, the signal energy is concentrated in the main-lobe and a good focusing effect can be achieved. The final experiments show the algorithm can obtain obvious super-resolution results, in addition, overcome the incomplete and deformation target structure problems of the traditional point scattering center model super-resolution method and retain the geometric characteristics of the target.4. OFDM-ISAR imaging and motion compensationConventional pulse compression technology in ISAR imaging may cause high side-lobe in range dimension, i.e. the echo signal energy of one target scattering center may leak to the adjacent range cells, which will have an effect on resolution of ISAR imaging. Chapter 6 presents an orthogonal frequency division multiplexing (OFDM) ISAR imaging algorithm. Based on the special time-frequency structure, by inserting a sufficient length of cyclic prefix, the side-lobe in range dimension is effectively suppressed. The same as conventional ISAR imaging for non-cooperative targets, motion compensation is the key issue in the OFDM-ISAR imaging. In the second part of Chapter 6, we propose an OFDM-ISAR range alignment and autofocus method. Range alignment is used to compensate the envelope migration of the target translational motion, while maintaining the range side-lobe suppression characteristics. In the processing of autofocusing, the errors from range alignment and phase errors of translational motion are compensated. The simulation experiments show, by transmitting the OFDM waveform with sufficient length of cyclic prefix, the OFDM-ISAR can obtain the low side-lobe images and improve the ISAR imaging resolution.
|
PDF Full Text Request