ISAR Ship Imaging And Cross Range Scaling With Multipath And Sea Clutter Interference

Imaging of radar has been widely integrated in modern radar instrument as an important function and used to generate two dimensional high resolution images of targets in all-time, all-weather and long range conditions, which plays important roles in both military area including strategic defense, anti-satellite, etc. and civil applications such as radar astronomy, terrain mapping and disaster forecasting. Inverse synthetic aperture radar (ISAR) is such kind of high resolution radar, which can generate images of moving targets such as satellites, missiles, aircrafts, ships and so on. It has great significance in offshore reconnaissance, antisubmarine, anti-ship, offshore traffic control and anti-smuggling to image ship targets at sea. Many countries in the world are engaged in shorebased ISAR imaging of seaborne target, which is still of leading edge and a hotspot in the world. This paper mainly investigates some confronted proplems which consist of complexity of target's three-dimensional rotation, suppression of background interference of multipath effects and sea clutter, optimum imaging time selection and cross range scaling when imaging ship targets using ISAR technologies. New algorithms of signal processing are furtherly brought forward to solve the problems. Main research contents are listed as below:1. Principles and algorithms of ISAR imaging are briefly introduced and ship target modeling and its motion characteristics are researched. Ideal ISAR imaging of ships is simulated without any background interference, which lay groundwork for the following researches when background interference exists and ship images need further optimizing.2. Multipath effects on ISAR imaging are studied and a new suppression algorithm is put forward. Imaging results will be affected by multipath effects when ship target is illuminated by radar beams at low elevation angles. Besides real target, two homogeneous artifacts with different locations and amplitudes will be emerged at the same side of the real target image, which blurs the ISAR image and hinders target recognition. This part begins with ISAR point spread function (PSF) theory and analyzes the mechanism behind this artifacts phenomenon through echo modeling. Application of range-Doppler principles with motion compensation process intuitively reveals the cause for this phenomenon. According to whether multipath effects exist and relative positions between the artifacts and real image, ISAR images can be classified into four categories: no multipath, blurry multipath, clear multipath and separable multipath. As for Separable multipath image, an auto image segmentation algorithm based on cross range vertical projection difference (VPD) is proposed to eliminate influence of the multipath effects. As for blurry and clear multipath images, an estimation method for multipath effects parameters is brought forward based on the mechanism producing artifacts, which uses two-dimensional image entropy as the evaluation function. The parameters are precisely estimated when corresponding image entropy is the least. Then, Influence of multipath effects can be canceled through estimated parameters and the artifacts will be eliminated. Parameters of multipath effects are multivariate, so simulated annealing (SA) algorithm is applied and the optimum nonlinear multivariate parameters could be estimated.3. Influences of sea clutter on ship imaging are researched, and corresponding clutter suppression algorithm based on frequency domain CFAR is presented. ISAR imaging of ship target at sea will be interfered by background sea clutter. This part begins with the study on statistical characteristics of the temporal-spatial correlated K-distributed sea clutter, and based on this, simulation of sea clutter under specific radar parameters using spherically invariant random process (SIRP) are carried out; then, with the sea clutter adding to the target echo generated under the same radar parameters, sea clutter effects on ISAR imaging are analyzed from the perspectives of sea spikes, sea clutter power spectrum and correlation features. When shape parameter is smaller, ISAR imaging will be affected by sea spikes; when clutter power spectrum approaches that of the target, spectrum of the target and sea clutter will be overlapped; when sea clutter is correlated under low signal to noise (SNR), precision of ISAR motion compensation will be reduced. In order to eliminate the above effects of sea clutter, comparative analysis between time- and frequency-domain constant false alarm rate (CFAR) on K distributed sea clutter is carried out and frequency-domain greatest of CFAR (GO-CFAR) detection method is adopted to eliminate sea clutter effects on ship imaging and clean ISAR images are obtained at last.4. Optimum imaging time (OIT) selection and cross-range scaling problems are studied, new algorithms for OIT section selection and fast cross-range scaling are put forward. Target echo is becoming nonstationary signal because of the complex ship motions, which are usually periodic and non-uniform rotation, so its Doppler frequency is time variant. Further more, cross-range resolution and image projection plane of ISAR will be changed too, hence, OIT selection is necessary in order to generate well-focused images. At first, definition of OIT is given, after that, OIT center is determined using Doppler spread algorithm. Combined with OIT section determined by image contrast maximization algorithm (ICMA), OIT selection process is completed. Imaging with the selected data section, optimum ISAR image can be generated, not only well focused but also with higher resolution. Then, cross-range scaling algorithm based on fractional Fourier transform (FrFT) is used to determine the cross-range resolution of the obtained optimum image. The cross-range scaling algorithm is detailed as following: after range compression, the slow time ISAR echo sequence in the same range bin are linear frequency-modulated (LFM) signals, and the chirp rate of these LFM signal contain information of the target angular velocity. Characteristics of FrFT enables it excellent LFM signal detection performance, so target angular velocity in the imaging interval can be estimated quickly and precisely using FrFT combined with maximization of kurtosis, then total aspect angle can be calculated and the cross-range scaled ISAR images are obtained finally. Further more, Simulation and real data experiments validate the algorithms mentioned above.