一种新的宽带目标识别雷达杂波抑制方法

宽带雷达相邻回波间会出现越距离单元走动现象,传统杂波抑制方法不适用．针对这一情况,提出一种在频率域-多普勒域宽带雷达杂波抑制方法．在目标速度未知时,采用改进的Hough变换将目标信息提取出来,等同于完成对杂波的抑制过程; 如果有目标速度的先验信息,则可以先进行运动补偿,然后提取目标信息,上述两种提取目标信息的途径避免了虑除杂波过程中对目标信号能量的损失．计算机仿真结果表明,该方法在输入信杂比为-10dB(速度未知)或-15dB(速度已知)的情况下仍可保持较高的识别性能,并且在输入信噪比为-5dB时,算法依然保持稳定的性能,对噪声有较好的稳健性．     

Joint adaptive waveform and baseline range design for bistatic radar

The problems of joint adaptive waveform design and baseline range design for bistatic radar to maximize the practical radar resolution were considered. Distinguishing from the conventional ambiguity function （AF）-based resolution which is only related with the transmitted waveform and bistatic geometry and could be regarded as the potential resolution of a bistatic radar system, the practical resolution involves the effect of waveform, signal-to-noise ratio （SNR） as well as the measurement model. Thus, it is more practical and will have further significant application in target detection and tracking. The constraint optimization procedure of joint adaptive waveform design and baseline range design for maximizing the practical resolution of bistatic radar system under dynamic target scenario was devised. Simulation results show that the range and velocity resolution are enhanced according to the adaptive waveform and bistatic radar configuration.     

采用OFDM-LFM的MIMO雷达高速目标波形设计

单基地多输入多输出雷达由于其宽发射波束的特性而能有效解决目标高速运动跨越发射波束带来的能量损失问题;但是现有波形对速度容忍性较差,高速目标回波匹配滤波能量损失较大．因此,有必要研究适用于高速目标背景下的多输入多输出雷达波形． 文中提出了一种基于正交频分复用线性调频信号的多输入多输出波形． 该波形不同发射通道分时复用一组不同频段的线性调频信号,从而得到带宽相同的可分离多路多输入多输出雷达波形． 仿真结果表明,文中设计的多输入多输出雷达波形能有效降低高速目标回波匹配滤波的能量损失,适用于高速目标背景．     

Optimal radar waveform design method in colored and non-Gaussian noise

This paper considers the design of optimal waveforms for detection of point targets in colored non-Gaussian noise. We present the likelihood ratio processor and derive an expression for its output signal-to-noise ratio (SNR), and then use this as the basis for a cost function in a waveform design optimization which is solved by using a one dimensional alternating maximization algorithm. Finally simulations are performed to verify the effectiveness of the new waveforms.     

基于核函数的多极化HRRP识别

针对多极化雷达高分辨距离像(HRRP)识别中数据量大、分布复杂和识别算法复杂的问题,提出了基于核函数的识别方法.该方法首先定义了两种基于多极化HRRP的核函数,然后将其分别应用到核主分量分析(KPCA)中降维和提取特征,最后采用最近邻(1NN)分类器和支持矢量机(SVM)分类器对目标进行分类.该方法可以在不丢失极化信息的情况下,将多极化HRRP作为一个整体进行识别,降低了识别算法的复杂度.多极化HRRP数据的仿真实验结果显示,该方法的识别率比单极化HRRP提高7%～10%;与其他多极化HRRP识别方法相比,该方法不仅降低了提取特征的维数,而且还提高了识别性能.     

一种基于互信息的波形优化设计方法

将注水法改进成一种适用于杂波背景下的通用注水法,通过最大化目标回波与扩展目标特性之间的互信息降低目标特性的不确定性．然后把这种思想应用于目标识别波形优化问题,从而增加各类目标的可分性．仿真结果证明,相对于线性调频信号和注水信号,该方法优化波形可以得到更大的互信息,并能将目标识别率提高3％左右,在信杂噪比较低的时候提高更加明显．     

在雷达HRRP识别中多特征融合多类分类器设计

在传统线性关联向量机的基础上,设计了一种多特征融合的多类分类器.该分类器基于多类Probit回归模型将传统的两类线性关联向量机推广为多类关联向量机,利用线性关联向量机的特征选择功能,对融合的高维特征向量进行降维和合理的幂次扩展,使线性关联向量机具有构造非线性分类界面的能力,以保证对非线性多类分类问题稳健的融合识别性能.针对雷达高分辨距离像目标识别问题,提取3种平移不变特征,使用提出的多特征融合的多类分类器在基于实测数据的识别实验中得到了稳健的融合识别结果.     

基于目标子空间的雷达自动识别系统

分析了HRRP空间RATR方法的冗余噪声匹配问题．探讨基于目标子空间的RATR方法．与HRRP空间相比,目标子空间理论上不存在冗余噪声,能够更加准确地反映目标的结构特征．对外场ISAR实测数据的处理表明,此目标子空间算法的识别结果比传统的HRRP空间算法的效率更高,比常规方法具有明显的优越性．     

宽带雷达动目标的一种互相关系数累加检测器

针对宽带雷达动目标检测问题,提出了一种在高斯噪声中的互相关系数累加检测器．宽带雷达动目标信号在距离频率域进行傅里叶变换取绝对值消除目标运动带来的相位项后,求互相关系数．利用相邻宽带雷达信号具有较大相关性这一特点,将多次相邻回波的互相关系数累加作为检测统计量．分析了高斯噪声互相关系数的概率分布,给出了检测门限设置方法．理论分析表明,互相关系数累加检测器具有恒虚警检测性能．计算机仿真试验表明,互相关系数累加检测器相比多普勒导向矢量已知的不依赖散射点分布密度的广义似然比检测器,在检测率为90％时有2dB的改善．     

信噪比加权空间分集雷达目标检测算法

对于空间分集雷达,各雷达站与目标的距离不同会引起信噪比的不同,从而降低传统的非相干积累检测器的检测性能.基于尼曼-皮尔逊准则,提出一种对各站接收信号进行基于信噪比加权的信号融合检测器.在已知各站回波信噪比的条件下,它可以达到检测性能的上界.数值试验表明,该检测器的检测性能较传统的非相干积累检测器有明显改善,特别在信噪比较低时.     

MIMO雷达发射方向图匹配和波形优化方法

为了匹配MIMO雷达系统给定的发射方向图,采用半正定规划(SDP)来优化发射信号的协方差矩阵．基于该协方差矩阵,利用循环算法(CA)和极小极大法,通过优化波形相位设计出了具有低空域和时域副瓣的恒模发射信号．此外,又提出了用凸规划设计空域失配滤波器和时域失配滤波器的方法,进一步降低了空域和时域副瓣．     

稳健的MIMO雷达发射波形和接收滤波器联合优化设计

针对杂波条件下目标反射信号方向不确定的情况,提出一种稳健的MIMO雷达发射波形和接收滤波器联合优化设计方法.在考虑各阵元发射功率相等的约束条件下,利用发射加权矩阵构造不确定集范围内关于输出信干噪比(SINR)的Maxmin优化模型;在此基础上,利用序列优化、半正定松弛和Charnes-Cooper转换,将非凸的联合优化问题转化为关于发射波形和接收滤波器的凸优化问题进行迭代求解;最后通过随机向量合成方法计算最终发射波形和接收滤波器,并对算法计算复杂度和收敛性给出分析和证明.所提方法在满足阵元发射功率一致的约束条件下,提高了算法稳健性并降低了计算复杂度.仿真实验证明了算法的有效性.     

MIMO radar waveform design in the presence of clutter and colored noise

The existing MIMO radar waveform design methods only deal with the optimization of radar waveform with the presence of clutter or colored noise, and unfortunately can not get the optimal waveform with the scenario when the clutter and colored noise coexist, which is more realistic in practical applications. To address this problem, a novel MIMO radar waveform design method based on mutual information in the presence of clutter and colored noise is proposed. The proposed scheme considers the influence of the target's response, clutter and colored noise on the optimization of the radar waveform. The mathematical expression for the transmit waveform matrix is achieved by theoretical derivation. What is more important, an optimal pairing of the eigenvectors of the target, clutter and colored noise is given and the power allocation for each antenna is optimized by the water-filling method. Simulation results show that when the total transmit power is 10dB, the proposed method's MI is 8bit and 4bit more than that of the general water-filling method and the method which only considers the pairing of the target's response and noise, respectively.     

雷达波形优化的特征互信息方法

针对扩展目标识别中距离敏感性的波形优化问题,在通用注水法的基础上提出了基于特征提取的波形优化设计方法,简称为特征互信息方法．该方法通过最大化识别特征与目标特性之间的互信息以达到增加各类目标可分性的目的．仿真结果表明,相对于线性调频信号和注水法优化信号,特征互信息方法优化信号可以得到更大的互信息,并能较好地提高目标的识别性能．     

一种针对目标跟踪的自适应波形选择方法简

Within an adaptive interacting multiple model (IMM) framework, an approach to selecting the transmitted waveform of radar is proposed. The adaptive IMM algorithm is attained when the constant-velocity (CV) model and the current statistical model are applied to the IMM filter. Firstly, the error ellipse orthogonal method is derived from information theory, and an equivalent form is given in terms of control theory. According to the equivalent form, the range and velocity tracking error can not be adjusted effectively using the error ellipse orthogonal method. To solve this problem, an adaptive transmitted waveform selection approach is proposed based on the minimum mean-squared error, combining with the adaptive IMM algorithm. Simulations show that the proposed algorithm is feasible and effective.     

MIMO radar waveform design method via low sidelobe beampattern modification

In the design of the multiple-input multiple-output(MIMO) radar waveform, decreasing transmit beampattern sidelobes could reduce the energy of clutters and false targets from sidelobes and increase the received signal to noise ratio (SNR) so as to improve the angle estimation property. Motivated by this idea, a transmit waveform design method based on low sidelobe beampattern modification is proposed for MIMO radar. First, the transmit waveform cross-correlation matrix can be obtained by the semidefinite relaxation (SDR) technique. Then the essence of our method is to establish an optimization modification model to reduce the beampattern sidelobe by changing the non-diagonal elements of the waveform cross-correlation matrix. Finally, the corresponding transmit beamforming weight matrix is obtained by this modified matrix, and the ESPRIT algorithm is used for the direction of arrival (DOA) estimation. Meanwhile, the feasibility of our method is demonstrated. Simulation results show the superiorities of our method in sidelobe suppression, DOA estimation accuracy and angle resolution.     

Transmit beampattern synthesis and waveform design for 2-D wideband MIMO radar

For matching the desired transmit beampattern of the two dimensional(2-D) wideband MIMO radar system, a new transmit beampattern synthesis and waveform design algorithm is proposed based on the frquency invariant beamforming. In this algorithm, to eliminate any dependency on frequency of the emiting waveform, the 2-D fourier relationship between uniform rectangular array(URA) transmit beampattern and spectrum of emiting waveform is used. Then the cost function of waveform matching frequency response is established, which imposes a constraint on constant modulus. Finally, the alternating matrix fitting method is used to design unimodular sequences. The proposed method effectively mitigates frequency dispersion of the 2-D wideband MIMO radar transmit beampattern. Numerical simulations have proved the validity of this algorithm.     

转发式干扰环境中的机载雷达运动目标检测

针对机载雷达运动目标检测中转发式干扰会引起虚警、抬高检测门限并降低雷达系统的检测概率的问题,提出了一种基于最大似然估计的机载雷达运动目标检测算法．首先估计转发式干扰的空域导向矢量,然后采用最大似然方法估计出运动目标的幅度,并以此来进行检测．仿真实验表明,该算法在完成杂波抑制的同时,可有效抑制旁瓣转发式干扰对运动目标检测的影响,并能在一定程度上降低主瓣转发式干扰对运动目标检测的影响．     

机动目标跟踪波形和检测门限自适应算法

由于现有雷达目标跟踪波形和检测门限自适应算法大多围绕以距离、距离率作为量测的一维运动目标展开,忽略了角度对目标跟踪的影响,从而无法对目标进行跟踪定位。为此,提出一种杂波背景下针对二维机动目标跟踪的雷达波形和检测门限联合自适应算法。首先,对传统基于时延-多普勒分辨单元的理论作进一步扩展,设计出一种包含距离-距离率和方位为量测的具有“棱柱”结构的分辨单元;然后,给出了包含波形参数和检测门限的量测误差协方差的联合近似表达式;最后,利用认知雷达思想,以滤波误差协方差的迹最小为代价函数自适应选择下一时刻的波形参数和检测门限,以提高系统的跟踪性能。仿真结果表明,波形和检测门限联合自适应算法的跟踪性能要明显优于传统的固定参数跟踪算法。     

基于逆向云模型的雷达目标识别

提出了一种基于逆向云模型的雷达目标识别方法。首先,基于模-1距离准则建立了云滴确定度,对高分辨距离像(HRRP)距离单元进行正态逆向云模型建模;然后,定义逆云隶属度表征待测样本属于某个训练类别的概率。对目标模板库中5种飞机高分辨距离像数据的仿真结果表明,该方法具有识别率高、对训练样本量的要求较为宽松和方位角划分不敏感等优点,是一种有效的雷达目标识别方法。