空间分集MIMO雷达的DOA估计新方法

多输入多输出(MIMO)雷达分布式布阵获得空间分集增益,但DOA估计是一个难题。该文分析了DOA估计的旁瓣一致现象,从空间分集MIMO雷达信号模型出发,推导了发射正交PCM信号,在接收端匹配滤波信号分选构造虚拟子阵的过程,提出了一种谱值优选算法进行空间分集MIMO雷达DOA估计。该算法得到目标在各子阵的最优估计,在发射端获得空间分集增益抑制目标闪烁的同时,在接收端进行稳健的多目标波达方向超分辨估计。最后仿真实验与相控阵进行了比较,表明了该算法的有效性,以及抑制目标闪烁改善DOA估计的能力。      

MIMO雷达的近场模型DOA估计

MIMO(Multiple input multiple output)雷达为实现目标角度分集,系统的天线阵一般分布式布阵,有效孔径比较大,目标一般处于天线的近场.该文从近场DOA估计的原理出发,利用MIMO的有源特性将距离参数作为已知,建立了多输入多输出系统的近场DOA估计模型,从原理上克服了传统时延定位方法存在的多目标时延对应难题.与一般近场DOA估计需要进行距离和角度的二维搜索不同,文中方位向估计只需要角度维的一维搜索,提高了运算速度.该文给出了系统可分辨最大目标数,并在理论上分析了方位向估计的CRB.     

MIMO雷达DOA估计方法分析

在介绍多输入多输出（MIMO:Multiple-Input Multiple-Output）雷达的信号模型的基础上,给出了一种目标波达方向（DOA:Direction of Arrival）估计方法--多信号分类（MUSIC:Multiple Signal Classification）方法在MIMO雷达系统DOA估计中的应用。与传统相控阵雷达相比,在MIMO雷达系统中该方法可以分辨的信号源数目有较大增加。通过仿真实验,详细分析了在MIMO雷达系统DOA估计中,多种因素对MUSIC方法测角精度的影响。     

基于L型MIMO雷达二维DOA估计新算法

将MIMO思想应用于L型阵雷达,建立了L型MIMO雷达信号模型,并在此基础上,根据DOA矩阵法,利用MIMO体制雷达产生的虚拟子阵列来构造一个二阶统计量,从而得到了目标方位角和俯仰角的闭式解.该算法不涉及多维谱峰搜索,因此具有较小计算量;此外,该算法估计出的目标的方位角和俯仰角可自动配对,不存在参数的配对问题.计算机仿真结果证明了本文算法的正确性和可行性.     

基于协方差匹配SLO算法的MIMO雷达DOA估计

利用加权平滑l0范数(Smoothed l0, SL0)算法估计MIMO雷达目标DOA时,需要把协方差矩阵进行矢量化来获得相应的稀疏重构模型,并利用信号和噪声子空间的正交性来构造加权向量。然而当存在相干信源时,MIMO雷达协方差矩阵的秩将退化,这会使得稀疏重构模型的误差较大以及无法正确区分信号和噪声子空间,导致加权SL0算法的DOA估计性能恶化。针对上述问题提出了一种基于协方差匹配SL0算法的MIMO雷达DOA估计方法。该方法利用协方差匹配准则重构出一个满秩的协方差矩阵,恢复MIMO雷达协方差矩阵的Toeplitz特性,并利用协方差逆矩阵的高阶幂来近似噪声子空间从而计算加权向量。仿真分析表明,该方法能够在无需预知信源数目的情况下有效地完成对相干信号的DOA估计。     

基于迭代近端投影的MIMO雷达多快拍DOA估计

传统基于压缩感知的MIMO雷达DOA估计算法将非凸非平滑稀疏表示问题近似成凸或平滑函数问题进行求解,稀疏表示模型误差的存在导致DOA估计性能不理想。为此,提出了一种基于迭代近端投影的MIMO雷达多快拍DOA估计方法。该方法首先将高维回波数据转换至低维空间以降低空域维度,并对降维后的数据进行奇异值分解(Singular Value Decomposition,SVD),提取信号子空间以降低时域维度,利用近端函数优化模型来表示MIMO雷达多快拍DOA估计中的非凸非平滑稀疏表示问题,然后采用SCAD(Smoothly Clipped Absolute Deviation Penalty)函数获得近端算子以求解该模型。仿真结果表明,该方法在低快拍和低信噪比下相干信源的DOA估计性能优于现有算法。     

阵元失效下基于矩阵重构的MIMO雷达DOA估计

针对阵元失效下MIMO雷达目标DOA估计性能下降问题,提出一种基于虚拟阵列采样数据矩阵重构的MIMO雷达DOA估计方法。MIMO雷达的阵元失效分为冗余虚拟阵元失效和非冗余虚拟阵元失效两种情况。当冗余虚拟阵元失效时,通过合并空间上位置相同的正常冗余虚拟阵元输出数据以实现信号降维与失效阵元数据填充。当非冗余虚拟阵元失效时,经降维填充后的数据矩阵中仍存在整行缺失数据,根据降维数据矩阵的低秩和稀疏先验,建立带低秩和稀疏约束的矩阵填充模型,并利用ALM-ADMM算法求解来恢复完整的降维数据矩阵。最后利用root-MUSIC算法估计目标DOA。仿真结果表明,本文方法能够有效提高MIMO雷达在阵元失效时的DOA估计精度。     

天地波MIMO雷达多目标角度估计仿真

天地波混合组网雷达系统将天地波一体化雷达和地波超视距雷达组网技术的特点和优势结合起来。多输入多输出雷达体制综合采用多通道、等效发射波束形成技术,为天波超视距雷达面临的多径、多模效应和杂波多普勒扩展等问题提供了一条有效解决途径。将MIMO(Multiple Input Multiple Output)技术与天地波雷达系统相结合,研究了天地波MIMO雷达的信号模型,以L阵配置下的天地波MIMO雷达为例,利用MUSIC(Multiple Signal Classification)算法实现多目标角度估计,实现了对天地波MIMO雷达的多目标角度估计的仿真试验,天地波MIMO雷达的仿真结果验证了该模型的有效性。     

基于协方差矩阵重构的阵元失效MIMO雷达DOA估计

在实际应用中由于恶劣环境或人为干扰等因素而导致多输入多输出(Multiple-Input Multiple-Output,MIMO)雷达部分阵元失效,使得其接收数据缺失及其协方差矩阵秩亏,从而导致子空间类算法的波达方向(Direction of Arrival,DOA)估计性能恶化甚至完全失效。针对上述问题,提出了一种接收阵元失效下基于协方差矩阵重构的MIMO雷达DOA估计方法。该方法根据MIMO雷达协方差矩阵中以接收阵元数划分的子方块矩阵具有Toeplitz特性,利用正常工作接收阵元的协方差矩阵元素来恢复相应的缺失元素,从而重构出完整的数据协方差矩阵,提高阵元失效MIMO雷达的DOA估计性能。仿真结果验证了所提方法的有效性。     

DOA estimation for monostatic MIMO radar based on unitary root-MUSIC

Direction of arrival (DOA) estimation is an important issue for monostatic MIMO radar. A DOA estimation method for monostatic MIMO radar based on unitary root-MUSIC is presented in this article. In the presented method, a reduced-dimension matrix is first utilised to transform the high dimension of received signal data into low dimension one. Then, a low-dimension real-value covariance matrix is obtained by forward–backward (FB) averaging and unitary transformation. The DOA of targets can be achieved by unitary root-MUSIC. Due to the FB averaging of received signal data and the eigendecomposition of the real-valued matrix covariance, the proposed method owns better angle estimation performance and lower computational complexity. The simulation results of the proposed method are presented and the performances are investigated and discussed.     

An ESPRIT-like algorithm for coherent DOA estimation based on data matrix decomposition in MIMO radar

A direction-of-arrival (DOA) estimation method for coherent sources is presented for MIMO radar. It uses symmetrical array mode for both the transmit and receive arrays and reconstructs a special data matrix from the range-compressed receive data. In the reconstructed matrix, the signal term is a Toeplitz matrix with the rank only related to the DOAs of the signals and independent with their coherency. Taking the noise term into account, the average method of multiple pulses is utilized to obtain the signal and noise subspaces. And then the DOA can be resolved via the SVD-based ESPRIT algorithm. Furthermore, the presented method is also useful in spatial colored noise scenario for MIMO radar. Theoretical and numerical simulations show the effectiveness of the proposed algorithm.     

基于MIMO雷达虚拟阵列的二维波达方向估计算法

基于L型配置MIMO雷达系统,提出了一种利用MIMO雷达单次回波的二维波达方向估计新算法。该算法通过单次回波信号形成的虚拟阵列数据来构造两个具有特定关系的矩阵,再用这两个矩阵构造一个波达方向矩阵,根据该矩阵的特征值和特征向量之间的关系,导出了目标方位角和高低角的闭式解。该算法不涉及二维谱峰搜索,计算量小,且不存在参数的配对问题;此外,当发射阵元数大于2时,其估计的最大目标数可大于或等于接收阵元数。计算机仿真结果证明了该算法的正确性和可行性。     

A RD-ESPRIT algorithm for coherent DOA estimation in monostatic MIMO radar using a single pulse

This paper discusses the problem of coherent direction of arrival (DOA) estimation in a monostatic multi-input multi-output (MIMO) radar using a single pulse, and proposes a reduced dimension (RD)-estimation of signal parameters via rotational invariance techniques (ESPRIT) algorithm. We reconstruct the received data and then utilise it to construct a set of Toeplitz matrices. After that, we use RD-ESPRIT to obtain the DOAs of the sources. The proposed algorithm is effective for coherent angle estimation based on a single pulse, and it has much better angle estimation performance than the forward backward spatial smoothing (FBSS)-ESPRIT algorithm and the ESPRIT-like of Li, as well as very close angle estimation performance to the ESPRIT-like of Han. For complexity comparison, our algorithm has very close complexity to the FBSS-ESPRIT algorithm, and lower complexity than the ESPRIT-like of Han and the ESPRIT-like of Li. Simulation results present the effectiveness and improvement of our approach.     

一种基于信道估计的MIMO雷达目标角度估计方法

MIMO雷达是一种新体制雷达。它的信号传输矩阵包含了我们需要的目标信息,如角度信息。本文首先建立了包含目标角度信息的MIMO雷达接收信号模型。然后根据MIMO雷达接收信号的统计特性,提出了新的MIMO雷达传输信道的估计方法,然后从传输矩阵中估计出目标角度信息。最后,我们通过仿真实验验证了本文方法的有效性。     

Joint parameter estimation of DOD/DOA/polarization for bistatic MIMO radar

In the paper,polarization-sensitive array is exploited at the receiver of multiple input multiple output (MIMO) radar system,a novel method is proposed for joint estimation of direction of departure (DOD),direction of arrival (DOA) and polarization parameters for bistatic MIMO radars. A signal model of polarimetric MIMO radar is developed,and the multi-parameter estimation algorithm for target localization is described by exploiting polarization array processing and the invariance property in both transmitter array and receiver array. By making use of polarization diversity techniques,the proposed method has advantages over traditional localization algorithms for bistatic MIMO radar. Simulations show that the performance of DOD and DOA estimation is greatly enhanced when different states of polarization of echoes is fully utilized. Especially,when two targets are closely spaced and cannot be well separated in spatial domain,the estimation resolution of traditional algorithms will be greatly degraded. While the proposed algorithm can work well and achieve high-resolution identification and accurate localization of multiple targets.     

阵元失效下基于核范数和SCAD惩罚的MIMO雷达DOA估计

阵元失效下多输入多输出(Multiple-Input Multiple-Output, MIMO)雷达虚拟阵列协方差矩阵出现大批整行整列元素缺失,破坏原有内在完整结构,导致波达方向(Direction of Arrival, DOA)估计性能下降。为此,提出一种联合核范数和SCAD(Smoothly Clipped Absolute Deviation)惩罚的完整协方差矩阵重构方法,以利于阵元失效下MIMO雷达DOA的有效估计。首先对待恢复的协方差矩阵建立核范数和SCAD惩罚双先验约束模型,并利用等正弦空间稀疏化方式划分粗网格空间,在可容忍的模型误差内能大大降低运算复杂度;然后利用ALM-ADMM(Augmented Lagrange Multipliers-Alternating Direction Method of Multipliers)算法对双先验约束模型进行求解,从而恢复协方差矩阵中大量整行整列的缺失数据;最后通过RD-ESPRIT(Reduced Dimensional ESPRIT)算法进行目标DOA估计。仿真结果验证该方法能快速恢复虚拟协方差矩阵中的缺失数据,从而有效...     

统计MIMO雷达的多目标定位和速度估计方法

在多目标情况下,统计多输入多输出(MIMO)雷达需要对不同观测通道中估计出的目标参数进行配对,才能完成多个目标定位及其速度估计。针对这一问题,对统计MIMO雷达提出了一种映射搜索配对法来实现在不同观测通道中的目标距离延时和多普勒频率估计值的配对,从而获得目标位置及其速度估计值。仿真结果表明,该方法在目标雷达散射截面(RCS)闪烁情况下能对多个目标进行有效定位和速度估计。     

Low complexity 3D-OMP algorithms for DOD DOA and Doppler frequency estimation in bistatic MIMO radar

The biggest challenge of the traditional 3D orthogonal matching pursuit (OMP) method for direction-of-departure (DOD), direction-of-arrival (DOA) and Doppler frequency estimation in bistatic multiple-input multiple-output (MIMO) radar is the heavy computational burden due to a large number of atoms in the overcomplete dictionary. In this paper, low complexity 3D-OMP algorithms are investigated. First, the traditional 3D-OMP algorithm is given. Then, two-dimensionality reduced OMP-based algorithms are proposed exploiting the property of Khatri-Rao product and proper sparse representation. Also, the multiple measurement vectors (MMV) model is introduced to our OMP algorithms to guarantee the estimation performance. The simulation results show that the DOD, DOA and Doppler frequency can be effectively estimated with a small number of pulses and low computation cost. With similar accuracy compared with the traditional 3D-OMP method, much lower computational burden can be achieved by using the proposed methods.     

双基地分布式阵列MIMO雷达的DOA和DOD估计方法

结合分布式阵列和双基地多输入多输出(Multiple-Input Multiple-Output, MIMO)雷达的特点, 提出了一种新的双基地分布式阵列MIMO雷达的接收角(Direction of Arrival, DOA)和发射角(Direction of Departure, DOD)估计方法.根据发射阵列和接收阵列的空域旋转不变特性, 利用旋转不变估计技术(Estimation of Signal Parameters via Rotational Invariance Techniques, ESPRIT)获取无模糊DOA粗估计和高精度周期性模糊的DOA、DOD精估计; 再利用无模糊DOA粗估计、目标的双基地距离信息以及双基地MIMO雷达的几何特点, 解除DOA、DOD精估计的周期性模糊, 得到高精度且无模糊的DOA和DOD估计.最后, 根据ESPRIT算法原理和估计误差的概率统计特性进行算法的性能分析, 给出算法基线模糊门限的近似计算方法.该算法有效地放宽了发射阵列孔径扩展程度的限制, 从而提高了阵列在大孔径下的角度估计精度, 且能够实现DOA和DOD估计的自动配对.仿真结果验证了所提算法和性能分析方法的有效性.