多通道雷达信号自适应检测技术研究
发布时间：2019-06-21 08:30
【摘要】：信号检测是信号处理的主要问题之一。早期的信号检测理论主要针对单通道数据。随着科技的发展,信号检测所要处理的数据通常以多通道形式呈现,例如相控阵天线的应用使雷达接收数据为多通道形式,且以复值向量或矩阵形式存在。此外,由于杂波和干扰的影响,雷达面临的噪声环境通常为色噪声。本文针对未知色噪声环境中的多通道雷达信号检测问题,研究有效的自适应检测方法。主要贡献如下:第二章给出了复值多通道信号检测对应的Rao准则和Wald准则,以及用于衡量复值参数估计性能的克拉美-罗界(Cramér-Rao Bound,CRB)。与现有方法相比,所提方法不需要将复数据按实部和虚部级联处理,而是把数据直接在复数域处理,从而降低了设计难度及计算复杂度。第三章研究了信号失配时的点目标检测问题。通过增加虚拟随机或确定干扰,基于Wald准则和GLRT准则,提出了失配敏感(即对失配信号抑制能力强)检测器。在此基础上,提出了单参数和双参数可调自适应检测器。通过参数调节,可调检测器可实现对失配信号的有效检测。此外,对上述自适应检测器进行了统计性能分析,得到了检测器的统计分布及检测概率和虚警概率的数学解析式。第四章研究了存在干扰时的点目标检测问题,分析了三种典型干扰类型下的检测技术。(1)针对已知干扰,根据Rao准则和Wald准则,提出了相应的自适应检测器。此外,提出了干扰已知时新的检测策略,即先干扰抑制后检测(Interference Cancellation Before Detection,ICBD)。与常规检测器设计方法相比,该方法更加简便,并且可工作在训练样本数少于待检测数据维数的环境中。(2)针对部分已知干扰,基于广义似然比检测器(Generalized Likelihood Ratio Test,GLRT)准则、Rao准则和Wald准则,提出了三种有效检测器,并分析了三者的统计性能。(3)针对完全未知干扰,根据GLRT准则和Wald准则,提出了有效的检测器,与现有检测器相比,新检测器具有更高的检测概率。第五章研究了存在信号失配时的扩展目标检测问题。通过人为增加虚拟确定干扰,提出了两种失配敏感检测器。然后在此基础上提出了一种参数可调检测器。根据设计需要,该检测器可实现对失配扩展目标的有效检测或者有效抑制。第六章研究了双子空间信号(Double-Subspace Signal,DOSS)的检测问题。DOSS信号指的是矩阵信号的行和列均位于已知子空间中。DOSS模型推广了现有分布目标检测问题中的信号模型。根据GLRT准则、Rao准则和Wald准则,以及谱范数检测器(Spectral Norm Test,SNT)的思想,提出了多种检测器。此外,分析了各检测器对应的多种特例。最后,通过仿真实验比较了各检测器在不同参数设置下的性能差异。第七章研究了三种目标方向信息不确定时的检测问题。(1)针对目标方向信息完全未知时的检测问题,基于Rao准则和Wald准则,提出了两种有效的检测器。(2)当目标回波均来自同一个方向,且相应的信号导向矢量位于一个已知子空间时,相应的检测问题称为方向检测。基于Wald准则,提出了一种有效的方向检测器,并与现有检测器进行了比较。(3)当目标的空域导向矢量和时域导向矢量均位于已知子空间时,根据GLRT准则和Wald准则及二者的两步实现,提出了四种有效的广义方向检测器,并给出了每种广义方向检测器的特例。第八章研究了小样本条件下的机载雷达空时自适应检测(Space-Time Adaptive Detection,STAD)问题。基于对角加载、主分量分析和Krylov子空间技术,提出了训练样本数不足时的多种有效检测器,分析了各检测器的渐近统计分布,并推导了渐近检测概率和虚警概率的解析式。第九章研究了天线共置多输入多输出(Multiple-Input Multiple-Output,MIMO)雷达的自适应检测问题。基于Rao准则和Wald准则,提出了相应的自适应检测器,分析了其统计性能,并与对应的相控阵雷达自适应检测器进行了比较。
[Abstract]:Signal detection is one of the main problems in signal processing. The early signal detection theory is mainly for single-channel data. With the development of science and technology, the data to be processed in the signal detection is usually presented in a multi-channel form, for example, the application of the phased array antenna enables the radar to receive data as a multi-channel form and is present in the form of a complex value vector or a matrix. In addition, due to the influence of clutter and interference, the noise environment facing the radar is usually color noise. In this paper, the problem of multi-channel radar signal detection in an unknown color noise environment is studied, and the effective self-adaptive detection method is studied. The main contribution is as follows: In the second chapter, the Rao criterion and the Wald criterion corresponding to the multi-channel signal detection of complex value are given, and the Cram-Rao Boround (CRB) is used to measure the parameter estimation performance of complex values. Compared with the prior art, the method does not need to process the complex data according to the real part and the virtual department, but the data is directly processed in a plurality of domains, so that the design difficulty and the calculation complexity are reduced. In the third chapter, the problem of point target detection in signal mismatch is studied. By increasing the virtual random or determined interference, a mismatch-sensitive (i.e., a strong-to-mismatch signal suppression capability) detector is proposed based on the Wald criterion and the GLRT criterion. On this basis, a single-parameter and two-parameter adjustable self-adaptive detector is proposed. By means of the parameter adjustment, the adjustable detector can realize the effective detection of the mismatch signal. In addition, the statistical performance analysis of the adaptive detector is carried out, and the statistical distribution of the detector and the mathematical analysis of the detection probability and the false alarm probability are obtained. In the fourth chapter, the problem of the detection of the point object in the presence of the interference is studied, and the detection technology under the three typical interference types is analyzed. (1) For known interference, a corresponding adaptive detector is proposed according to the Rao criterion and the Wald criterion. In addition, a new detection strategy is proposed, that is, interference cancellation before detection (ICBD). Compared with the conventional detector design method, the method is more convenient and can work in an environment where the number of training samples is less than the dimension of the data to be detected. (2) For some known interference, based on the generalized likelihood ratio detector (GLRT) criterion, the Rao criterion and the Wald criterion, three effective detectors are proposed, and the statistical properties of the three are analyzed. (3) For completely unknown interference, a valid detector is proposed according to the GLRT criterion and the Wald criterion. Compared with the existing detector, the new detector has higher detection probability. The fifth chapter studies the problem of extended target detection in the presence of signal mismatch. Two kinds of mismatch sensitive detectors are proposed by artificially increasing the virtual determination interference. Then, a parameter-adjustable detector is proposed. According to the design requirements, the detector can realize the effective detection or the effective suppression of the mismatch expansion target. In chapter 6, the detection of Double-Subspace Signal (DOSS) is studied. The dss signal refers to the rows and columns of the matrix signal being in the known sub-space. The DOSS model extends the signal model in the existing distribution target detection problem. According to the GLRT criterion, the Rao criterion and the Wald criterion, and the idea of the spectral norm detector (SNT), a variety of detectors are proposed. In addition, various special cases corresponding to each detector are analyzed. Finally, the performance difference of each detector under different parameter settings is compared through the simulation experiment. Chapter 7 studies the detection problem when the three target direction information is uncertain. (1) Two effective detectors are proposed based on the Rao criterion and the Wald criterion when the target direction information is completely unknown. (2) When the target echoes come from the same direction and the corresponding signal guide vector is located in a known subspace, the corresponding detection problem is called direction detection. Based on the Wald criterion, an effective direction detector is proposed and compared with the existing detector. (3) Four effective generalized directional detectors are proposed according to the GLRT criterion and the Wald criterion and the two-step implementation of both the GLRT criterion and the Wald criterion, and a special case of each generalized direction detector is given. In chapter 8, the space-time adaptive detection (STAD) problem of airborne radar under the condition of small samples is studied. Based on the diagonal loading, the principal component analysis and the Krylov subspace technique, a variety of effective detectors with insufficient training samples are proposed, the asymptotic statistical distribution of each detector is analyzed, and the analytical expressions of the asymptotic detection probability and the false alarm probability are derived. Chapter 9 studies the self-adaptive detection of multiple-input multiple-output (MIMO) radar. Based on the Rao criterion and the Wald criterion, a corresponding adaptive detector is proposed, and its statistical performance is analyzed and compared with the corresponding phased-array radar adaptive detector.
【学位授予单位】：国防科学技术大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TN957.51
，
本文编号：2503933
[Abstract]:Signal detection is one of the main problems in signal processing. The early signal detection theory is mainly for single-channel data. With the development of science and technology, the data to be processed in the signal detection is usually presented in a multi-channel form, for example, the application of the phased array antenna enables the radar to receive data as a multi-channel form and is present in the form of a complex value vector or a matrix. In addition, due to the influence of clutter and interference, the noise environment facing the radar is usually color noise. In this paper, the problem of multi-channel radar signal detection in an unknown color noise environment is studied, and the effective self-adaptive detection method is studied. The main contribution is as follows: In the second chapter, the Rao criterion and the Wald criterion corresponding to the multi-channel signal detection of complex value are given, and the Cram-Rao Boround (CRB) is used to measure the parameter estimation performance of complex values. Compared with the prior art, the method does not need to process the complex data according to the real part and the virtual department, but the data is directly processed in a plurality of domains, so that the design difficulty and the calculation complexity are reduced. In the third chapter, the problem of point target detection in signal mismatch is studied. By increasing the virtual random or determined interference, a mismatch-sensitive (i.e., a strong-to-mismatch signal suppression capability) detector is proposed based on the Wald criterion and the GLRT criterion. On this basis, a single-parameter and two-parameter adjustable self-adaptive detector is proposed. By means of the parameter adjustment, the adjustable detector can realize the effective detection of the mismatch signal. In addition, the statistical performance analysis of the adaptive detector is carried out, and the statistical distribution of the detector and the mathematical analysis of the detection probability and the false alarm probability are obtained. In the fourth chapter, the problem of the detection of the point object in the presence of the interference is studied, and the detection technology under the three typical interference types is analyzed. (1) For known interference, a corresponding adaptive detector is proposed according to the Rao criterion and the Wald criterion. In addition, a new detection strategy is proposed, that is, interference cancellation before detection (ICBD). Compared with the conventional detector design method, the method is more convenient and can work in an environment where the number of training samples is less than the dimension of the data to be detected. (2) For some known interference, based on the generalized likelihood ratio detector (GLRT) criterion, the Rao criterion and the Wald criterion, three effective detectors are proposed, and the statistical properties of the three are analyzed. (3) For completely unknown interference, a valid detector is proposed according to the GLRT criterion and the Wald criterion. Compared with the existing detector, the new detector has higher detection probability. The fifth chapter studies the problem of extended target detection in the presence of signal mismatch. Two kinds of mismatch sensitive detectors are proposed by artificially increasing the virtual determination interference. Then, a parameter-adjustable detector is proposed. According to the design requirements, the detector can realize the effective detection or the effective suppression of the mismatch expansion target. In chapter 6, the detection of Double-Subspace Signal (DOSS) is studied. The dss signal refers to the rows and columns of the matrix signal being in the known sub-space. The DOSS model extends the signal model in the existing distribution target detection problem. According to the GLRT criterion, the Rao criterion and the Wald criterion, and the idea of the spectral norm detector (SNT), a variety of detectors are proposed. In addition, various special cases corresponding to each detector are analyzed. Finally, the performance difference of each detector under different parameter settings is compared through the simulation experiment. Chapter 7 studies the detection problem when the three target direction information is uncertain. (1) Two effective detectors are proposed based on the Rao criterion and the Wald criterion when the target direction information is completely unknown. (2) When the target echoes come from the same direction and the corresponding signal guide vector is located in a known subspace, the corresponding detection problem is called direction detection. Based on the Wald criterion, an effective direction detector is proposed and compared with the existing detector. (3) Four effective generalized directional detectors are proposed according to the GLRT criterion and the Wald criterion and the two-step implementation of both the GLRT criterion and the Wald criterion, and a special case of each generalized direction detector is given. In chapter 8, the space-time adaptive detection (STAD) problem of airborne radar under the condition of small samples is studied. Based on the diagonal loading, the principal component analysis and the Krylov subspace technique, a variety of effective detectors with insufficient training samples are proposed, the asymptotic statistical distribution of each detector is analyzed, and the analytical expressions of the asymptotic detection probability and the false alarm probability are derived. Chapter 9 studies the self-adaptive detection of multiple-input multiple-output (MIMO) radar. Based on the Rao criterion and the Wald criterion, a corresponding adaptive detector is proposed, and its statistical performance is analyzed and compared with the corresponding phased-array radar adaptive detector.
【学位授予单位】：国防科学技术大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TN957.51
，
本文编号：2503933
本文链接：https://www.wllwen.com/kejilunwen/wltx/2503933.html
