Research On High Resolution ISAR Imaging And Scaling

Inverse Synthetic Aperture Radar (ISAR) imaging technique has the advantages of long-distance applications, all-weather, and all-day/night, which significantly enhances the capability of modern radar information awareness and data acquisition. Therefore, ISAR technique plays an important role in both military and civilian fields.With the increasing of application requirements, ISAR imaging is gradually developed from the coarse resolution to high resolution, and from single channel, polarization to multiple modes, and from single sensor signal processing to multi-sensor signal fusion processing. Along with the development of radar systems and phased array techniques, modern multi-function ISAR not only has the capability of high resolution imaging, but also needs to simultaneously complete more other tasks, such as wide swaths observation, multi-targets tracking and scaling. In practice, however, the time and resources used for wideband imaging for a single target are limited, leading to short or sparse aperture generally. Therefore, combining the characteristics of the modern radar system and imaging applications, it is necessary to conduct the research on ISAR accurate phase error compensation, high resolution imaging and scaling technique to improve the ability of radar detection and target recognition.Without increasing the complexity of the existing radar system, this dissertation focuses on studying new techniques to improve the resolution, robustness, and flexibility of ISAR using sparse signal processing methods. Three key aspects for ISAR have been studied, including ISAR cross-range scaling, precision motion compensation procedure with sparse-aperture measurement, multiple channels three dimension high resolution imaging. The relevant work is supported by National Basic Research Program of China (973 Program, No.2010CB731903), National Science Foundation of China (No. 61301280 and No.61001211) etc.The main contents of this dissertation can be summarized as follows:The first part is the basic theory of this dissertation, which provides the necessary background knowledge for the subsequent chapters. This part first introduces the principles and summarizes the traditional approaches of ISAR imaging. Then, the Compressive Sensing (CS) based high resolution ISAR imaging algorithm is presented by combining the three major elements of the CS theory.The second part studies the algorithm of cross-range scaling for ISAR imaging with short aperture data and under low Signal Noise Ratio (SNR). We propose an effective algorithm to solve the cross-range scaling for ISAR imaging during a short Coherent Processing Interval (CPI) under low SNR. The relationship between the two ISAR images obtained at different aspect angles is analyzed firstly, and then high resolution imaging is performed using Weighted Compressive Sensing (WCS) procedure based on the sparsity characteristic of the ISAR image, which can encourage signal components while suppress noise. Then the initial estimation of the Rotation Angle Velocity (RAV) is realized by the correlation between two polar images on the basis of the characteristics of 2-D fast Fourier transform (2-D FFT) and polar mapping. Finally, accurate estimation of the maximum correlation position is conducted using WCS. The estimation precision and efficiency of RAV are improved, and thus the rescaled ISAR image can be implemented.The third part studies the algorithm of phase adjustment and high resolution ISAR imaging with sparse apertures. Based on sparse apertures observation model, we propose a weighted eigenvector autofocus algorithm that precisely corrects the phase error for sparse apertures ISAR data. In the approach, dominant range cells are firstly chosen according to the normalized amplitude variance criteria. Then, different weights are added to each dominant cell according to their SNRs, which can effectively encourage the contribution of the range cell with high SNR and suppress noise support. Moreover, to improve the estimation precision, the estimation and compensation of the Doppler frequency and phase error are processed in an iterative manner. Finally, according to Bayesian Compressive Sensing (BCS), a sparsity-constraint optimization model is established using the Maximum a Posteriori (MAP) estimator, and high resolution ISAR imaging under sparse aperture observation can be reconstructed with high precision.The last part focuses on three-dimensional Interferometric ISAR (InISAR) imaging algorithm with limited pulses by exploiting joint sparsity. For maneuvering targets, long CPI observation results in a serious time-varying Doppler modulation that is difficult to achieve high quality three-dimensional (3-D) InISAR images. As a result, we propose a novel 3-D InISAR imaging algorithm with limited pulses for maneuvering targets by exploiting joint sparsity. In the approach, we first correct the range alignment and phase adjustment using combined processing approach, and then compensate the range difference between the two echo signals received by different antennas, and the registered ISAR images could be obtained. Then, we establish a joint sparsity-constraint optimization model to reconstruct super-resolution ISAR images. By exploiting the joint sparsity of the multi-channel echo signals, the proposed algorithm effectively improves the recovery precision of scattering centers while preserves the cross-channel information during the resolution enhancement process. Consequently, the accurate interferometric phase and high quality 3-D images of maneuvering targets can be achieved via the conventional interferometry technique. Moreover, the application of fast Fourier transform (FFT) implementation simplifies the determination of optimization and further promotes the computational efficiency of the proposed method.