【作者】 戴奉周；

【导师】 吴顺君；

【作者基本信息】 西安电子科技大学 ， 信号与信息处理， 2010， 博士

【摘要】 与窄带雷达相比,宽带雷达在许多方面都具有优势。例如,可以获得更多的目标和环境信息,以实现目标成像或地形测绘,也有利于实现可靠的目标识别和分类;可以提高雷达对杂波中目标的检测能力;可以更加精确地测量目标的位置和运动参数;具有更好的电磁兼容性能,以对抗各种有源和无源干扰;具有更好的低截获性能;可以更加灵活地使用发射波形与目标和环境相匹配,以改善雷达的整体性能。因此,宽带雷达具有应对日益复杂的任务需求和战场环境的潜力,成为雷达发展的趋势。研究宽带雷达信号处理的基本理论和方法,具有重要的理论意义和工程价值。本文主要对宽带雷达信号处理中的目标检测、杂波抑制和目标回波的相参积累以及基于在线波形优化的宽带认知跟踪等三方面的基本理论和技术进行了深入研究。具体工作可概括为如下的五个部分:1.比较分析了宽带和窄带雷达在噪声中检测目标的性能。首先讨论了宽带与窄带雷达的雷达方程和目标散射截面积(RCS);随后,介绍了三种根据真实目标散射中心分布情况而抽象出的宽带雷达回波统计模型:宽带无起伏目标,宽带瑞利起伏目标及宽带莱斯起伏目标;然后推导了宽带和窄带雷达对符合这三种模型的目标的检测概率,指出宽带雷达的检测性能与窄带雷达相比,是由目标RCS起伏减小和检测器积累损失增加这一对矛盾因素共同决定的;最后得出结论:虽然理论上在高检测概率时宽带雷达的检测性能比窄带雷达高,但二者检测概率曲线的交点随包含目标回波的距离单元的增加而越来越高,使宽带雷达的检测优势失去意义。2.研究了在高斯杂波中基于单次回波的宽带雷达目标检测问题。首先分析了宽带雷达对杂波中的目标进行检测的优势和问题所在;接下来回顾了几种已有的基于单次回波的宽带雷达目标检测算法,并分析了它们各自的优点和存在的问题,以及实际中提高宽带雷达目标检测性能的难点在于无法获得待检测目标散射中心分布的先验信息;随后,从解决这一难点入手,提出了一种基于顺序统计量的宽带雷达目标序贯检测算法,从理论、仿真实验和实测数据验证三方面证明了该算法在完全不需要散射中心分布先验信息的条件下,对于散射中心稀疏分布的目标与以往的宽带雷达检测算法相比具有更好更稳健的检测性能。3.研究了非高斯时域相关杂波中基于多个相参脉冲回波的宽带雷达目标检测问题。首先介绍了宽带雷达目标多次相参脉冲回波的准确模型,并根据实际应用中的雷达和目标参数得到了其简化模型;然后介绍了球不变随机向量(SIRV)描述的非高斯宽带雷达杂波模型及其仿真方法;随后设计了基于广义似然比的宽带雷达目标广义自适应子空间检测器,对其性能进行了完整的理论分析,并提出了相应的高效实现算法;此外还考虑到实际目标可能还存在多普勒扩展的问题,设计了可对在距离和多普勒方向上都存在扩展的目标进行检测的宽带雷达检测算法;最后通过仿真实验验证了本章设计的检测器可以实现对宽带雷达目标回波的相参积累,与各种已有算法相比,显著提高了宽带雷达的检测性能。4.研究了基于子带处理的宽带雷达杂波抑制与目标回波相参积累的方法。首先根据宽带雷达杂波和目标回波的特点,指出了使用过采样滤波器组对宽带雷达回波进行子带分解的必要性,并介绍了相应的实现子带分解和综合的滤波器组的设计方法;随后指出宽带雷达实现杂波抑制和相参积累的目的并不仅仅像常规的窄带雷达那样是为了实现目标检测,而更重要的目的在于增强回波中的目标冲激响应,提高目标识别的可靠性,并根据这一要求设计了通过子带加权处理,从包含杂波的回波中重构目标冲激响应的最优和次优处理方法,最后通过仿真和实测数据验证了算法的有效性。5.研究了宽带雷达的认知跟踪问题。首先根据目标冲激响应的物理概念和特点,提出了采用宽平稳—不相关散射模型来描述雷达目标冲激响应及其动态模型;接下来根据这一模型首先研究了在时域设计最优跟踪波形的问题,并分析了它的不可实现性;随后提出了一种在频域设计次优跟踪波形的方法,并给出了它的实现算法;从实际雷达要求发射波形具有恒包络这一需求出发,还分别给出了在给定发射信号功率谱的情况下用相位驻点法和迭代加权最小二乘法设计恒包络信号的方法;最后,通过仿真实验说明,在一定条件下该方法可以减小宽带雷达跟踪时对目标参数的估计误差。更多还原

【Abstract】 Wideband radars have many advantages over the narrowband radars, for instance, more information of the targets and environment, which can be used in imaging and mapping, additional reliable classification and recognition for the targets; stronger capability of target detection in clutter; more accurate measurements of the location and movement parameters of the targets; better electromagnetic compatibility and immunity to active and passive interference; low probability of interception and better secrecy of illumination; more flexible waveforms to match the targets and environments. So wideband radars have the potential to cope with the increasingly complex task and the battlefield environment, and become the trend of radar development. Research on the wideband radar signal processing is significant in both the theory and application.This dissertation is addressed on the wideband radar signal processing, and its scope covers the target detection, clutter suppression and target return coherent and cognitive tracking based on waveform agile on fly. The work of this dissertation can be surmmarized as the following five aspects:1. The target detection performances in noise of wideband and narrowband radar are analyzed and compared. Firstly, the radar equation and target radar cross section (RCS) of the wideband and narrowband radar are discussed; then three statistical models induced from the scattering of real targets are introduced: wideband nonfluctuating target, wideband Rayleigh target and wideband Ricean target; the detection probabilities of these three target models for wideband and narrowband radars are deduced, and we point out that the detection performance in noise of the wideband radar is determined by the tradeoff between the two opposite factors, that are the target RCS decrease of the target RCS and the increase of the integration loss; finally, we conclude that though the performance of the target detection in noise of wideband radars outperforms the narrowband radar under high signal-to-noise ratio (SNR) in theory, the cross-point of the two detection probability curves is too high, and the result is that the detection advantage of the wideband radar is meaningless for real targets.2. The problem of wideband radar target detection in Gaussian clutter by one pulse is researched. Firstly, the advantage and problem of detecting targets in clutter by wideband radar are discussed; several present algorithms are reviewed, additionally, their advantages and disadvantages are analyzed; then we point that the difficult of wideband radar detection is the absence of the knowledge about the spatial distribution of the target scattering centers in real applications; to overcome this difficulty we design a novel algorithm for wideband radar detection—order statistics based modified sequential testing (OSBMST); by theory analysis, simulations and actual target returns validations, we prove the OSBMST can achieve better and more robust performance than the present algorithms for sparse scattering target when the prior knowledge of the targets’scattering centers distribution is absent.3. The problem of wideband radar target detection in non-Gaussian clutter by the returns of multiple coherent pulses is researched. The precise and approximate models of wideband target return are introduced firstly; then the spherical invariant random vector (SIRV) model of non-Gaussian wideband clutter and the method of generation are presented; based on the target and clutter models we design the wideband radar generalized adaptive subspace detector based on the generalized likelihood ratio test (GLRT). The statistical properties of the detector are discussed completely and rigorously, and a high efficiency algorithm for the detector is designed. Additionally, the detector is expended to fit the wideband radar target with Doppler spread. Finally, the performances of the detector are verified by simulations. The results show that the detector we designed can integrate the target returns from multiple pulses coherently, and has much better performance than the present analogous detectors.4. The method of clutter suppression and targets return coherent integration based on subband processing is developed. Firstly, the necessity of analyzing the wideband returns into subbands by the oversampled filterbank is discussed based on the characters of the wideband target return and clutter, and then the method of filterbanks design is introduced; additionally, we show that the purpose of clutter suppression and coherent integration for wideband is to enhance the target impulse response in the returns to increase the reliability of the target recognition, rather than achieve higher signal-clutter-ratio as the narrowband radar; in the sequel, the optimal and suboptimal weights for clutter suppression and target enhancement in subband are designed. Finally, the validity of this method is proved by the results of simulations and actual returns processing.5. The problem of cognitive tracking by wideband radar is investigated. Firstly, the concept and properties of the target impulse response (TIR) are discussed and the wide sense stationary - uncorrelated scattering (WSS-US) model is introduced to describe the TIR and its dynamical model; the problem of optimal waveform design for tracking in time domain is researched, and its infeasibility is analyzed; then a suboptimal waveform designed method in frequency domain and its realization algorithm are proposed; additionally, according to the constant envelope requirement of the radar transmitter power amplifiers, two methods of constant-envelope waveform design in terms of given spectrum, namely, the stationary phase method and the iterative weight least square method are given; finally, we show that the method we proposed can decrease the estimation error for wideband radar tracking in some condition.更多还原