| The airborne long-range battlefield surveillance radar system can provide real timereconnaissance and surveillance over a vast area on the ground or the sea, and hence isone of the most important sensors that are used for modern military applications. Theradar’s two working modes, i.e., synthetic aperture radar (SAR) ground moving targetindication (GMTI) mode or wide-area surveillance (WAS) GMTI mode, offer thedetection of moving targets in the battlefield. The WAS-GMTI mode, especially turnsout to be a particularly effective way to detect, locate and track moving targets, and thusis becoming more and more important in modern warfare. Recently, most militarypowers in the world have been engaging in developing airborne battlefield surveillanceradar systems, and a great number of universities and research institutes are developingpractical GMTI techniques and exploring novel moving target detection and locationalgorithms with high efficiency.Based on the airborne WAS-GMTI research development of home and abroad, thisdissertation aims to improve moving target detection and location, and additionally togive in-depth discussion on WAS-GMTI signal processing method. More specifically,there were four key areas that we wished to address. First, this dissertation containssome basic materials on signal model of WAS-GMTI mode. Then, more importantly,this dissertation consists of a channel phase error estimation method, two imagestitching algorithms for ESA mode and MSA mode, and two moving target relocationalgorithms, all verified through real data processing. The main content of thisdissertation is summarized as follows.1. Chapter2provides the reader with the observation models and mathematicalpreliminaries required to understand the WAS-GMTI mode. The purposes of Chapter2are to define important concepts and notations in WAS-GMTI, to describe WAS-GMTIdata acquisition geometry, and to represent more specially signal processing techniqueswhich form the basis for the proposed algorithms carried out in the subsequent chapters.2. The channel phase error presented in radar will affect clutter suppression anddecline the parameter estimation accuracy as well. To deal with this problem, anestimation method of channel phase error is proposed in Chapter3for the phased arrayantenna. First the clutter of different channels is processed using the interferometrictechnique, after which the channel errors are estimated with the known scan angles. Toreduce the influence of noise and enhance the estimation accuracy, two strategies areintroduced for phase error estimation. On one hand, the independent and identically distributed (i. i. d.) range cells are proposed to be chosen as estimation samples. On theother hand, clutter of multiple wave positions is employed to further average theestimation error. Processing results with measured data validate that the proposedmethod can compensate the channel phase error very well and enhance the ability ofmoving target relocation significantly.3. Based on the discussion about observation models and mathematicalpreliminaries, this dissertation contains descriptions and analyses of the image stitchingalgorithms of Doppler beam sharpening (DBS) image for ESA system and MSA system,respectively. Chapter4is concerned with image stitching in ESA. By stitching Dopplerbeam sharpening (DBS) images together the main-lobe DBS images obtained from allthe beam positions, a wide swath ground image is obtained. In practice, however, thereare some difficulties in stitching the DBS images due to the nonideal movement(variation of the velocity vector and the attitude) of the radar platform. To deal withthese problems, a novel method is proposed, which can stitch the DBS imageseffectively. The inertial navigation system (INS) information is employed to estimatethe instantaneous position of the aircraft and the radar beam direction, both of which arethen transformed to the original reference coordinate so that the corrected parameters ofthe image can be obtained. Finally, by using these parameters, the image stitching isachieved. Processing results of measured data show that the proposed method not onlycompensates the motion error excellently, but also significantly outperforms thetraditional algorithm in stitching performance.4. Chapter5deals primarily with high resolution image stitching algorithm forDBS images in MSA system. The current image stitching algorithms are primarilysuited to process electronically steered array (ESA) data. However, these algorithmswill give rise to degradation of the performance for mechanical scanning antenna(MSA). In this paper, we develop an effective mulitsubbeam combination algorithm thatapplies to the MSA. For MSA whose beam is steered with physical antenna motion, thesuccessive pulses of adjacent subbeams are temporally correlated, and thus can becoherently integrated to generate a DBS image. The most prominent contributions of theproposed method are mulitsubbeam combination according to MSA geometry, whichguarantees that a high resolution DBS image can be obtained, and good data selection inthe way of Doppler spectrum combination. Results from real data show that theproposed method is computationally quite efficient for high-resolution imaging andsignificantly outperforms the traditional algorithm.5. The system factors that affect the performance of moving targets relocation can be examined through channel imbalance and antenna array misalignment etc. Thisdissertation derivates and developes two algorithms by which a moving target can beaccurately relocated in the SCAN-GMTI system. In Chapter6, we introduce a conceptcalled error equivalent baseline (EEB) to describe the influence of these factors anddevelop a novel method of moving targets location for the scan-ground moving targetindication (SCAN-GMTI) system. The proposed method involves measuring the cluttersuppression sharpness ratio (CSSR) corresponding to the potential EEBs, determiningthe optimal EEB, and locating the moving targets. The key procedure of estimating theoptimal EEB that maximizes the CSSR virtually provides a good fit for the errorcharacteristic of the measured data. Therefore, the location performance of the obtainedoptimal EEB is superior to that of the nominal baseline. Processing results of measureddata validate the effectiveness of the proposed method.6. Chapter7deals with the issue of moving target relocation for wide-area groundmoving target indication (GMTI) using dual-channel radar systems. Due to channelmismatch, along-track baseline error, the existence of across-track baseline, and etc, it isdifficult to accurately relocate the detected targets. In this letter, we propose a newknowledge-based (KB) method for target relocation. The key idea of the proposedmethod is to use such a fact that a moving target and its neighboring clutter which issituated at the same azimuth angular position and the same range cell as this target inthe observed scene, have the same interferometric phase. This new KB method, as anindirect one, does not employ the conventional relocation formula, and therefore is notinfluenced by most of (if not all) interferometric-phase errors. Experimental results forreal radar data demonstrate a fairly high degree of target-relocation accuracy. |
PDF Full Text Request