非均匀环境下的STAP与MIMO雷达参数估计研究

发布时间：2018-04-16 03:38

本文选题：空时自适应处理 + 运动目标检测　；参考：《西安电子科技大学》2014年博士论文

【摘要】：机载雷达下视工作时容易受到强地杂波的影响。传统的一维时域滤波和空域滤波难以有效滤除这些在多普勒域和空域上都扩散的杂波。空时自适应处理(Space-Time Adaptive Processing,STAP)技术联合了多个空间通道和相干脉冲进行二维自适应滤波,能够在抑制杂波的同时保证目标获得一定的相干处理增益,是机载雷达杂波抑制的一项关键技术。为了获得更优的STAP处理性能,理论上用于估计杂波统计特性的独立同分布样本数不应少于系统自由度的两倍。但实际杂波环境的非均匀性会导致STAP处理器无法获取如此之多的独立同分布样本,从而导致STAP的杂波抑制性能严重下降。如何提高STAP在非均匀环境下的杂波抑制性能是十分重要的问题。多输入多输出(Multiple-Input Multiple-Output,MIMO)雷达是近年来出现的一种新体制雷达,它的每个天线阵元具有发射任意波形的能力。与常规相控阵雷达相比,MIMO技术能够显著增加系统的自由度,扩展阵列虚拟孔径,提高目标参数估计精度。随着系统自由度的增加,常规参数估计方法的运算复杂度会相应地增加,同时低信噪比、低快拍数等非理因素会使参数估计精度严重下降。如何在低信噪比、低快拍数的非理想环境下提高MIMO雷达参数估计精度,同时降低运算量是值得研究的问题。本论文围绕这上述问题开展研究,具体包括以下几个方面。一、研究了机载非正侧面阵雷达距离非平稳杂波抑制的问题。当机载雷达的天线非正侧视安装时,其杂波谱会随着距离的变化而变化,从而导致杂波出现距离非平稳的现象。杂波的距离非平稳性会导致传统STAP无法获取独立同分布的训练样本。在高重频工作模式下情况将变得更加复杂,近程杂波和远程杂波由于距离模糊耦合在一起,杂波抑制变得更加困难。提出了一种基于正交波形的俯仰自适应波束形成算法用来抑制距离非平稳的近程杂波,而剩余的距离平稳的远程杂波由方位-多普勒STAP来抑制。所提俯仰波束形成算法能够有效地抑制近程杂波,同时可以保护远程目标。在滤除了距离非平稳的近程杂波后,杂波样本的距离平稳性得到明显的增强,能够为后续方位-多普勒STAP处理器提供独立同分布的训练样本,使其杂波抑制性能得到明显的改善。二、研究了非均匀环境下采用直接数据域STAP方法进行杂波抑制的问题。杂波功率上的非均匀会导致杂波抑制不够充分或者凹口过深,而离散杂波则容易导致虚警,密集动目标环境则容易引起STAP自适应方向图畸变和目标自相消。理论上直接数据域方法能够解决上述杂波非均匀的问题。但传统的直接数据域方法存在空时孔径损失和受阵列结构限制等问题,提出了一种基于迭代自适应谱估计算法的直接数据域STAP方法。该方法能有效避开传统统计类STAP算法在非均匀环境下所面临的样本数不足和样本污染的难题,且无空时孔径损失,适用于任意形状的阵列。三、研究了机载雷达在密集欺骗式干扰环境下的杂波抑制问题(密集欺骗式干扰环境可看作是非均匀环境的一种特殊情况)。机载雷达欺骗式旁瓣干扰容易引起大量虚警,使雷达跟踪数据处理器饱和。而欺骗式主瓣干扰不仅会引起大量虚警,还会污染空时自适应处理器的训练样本,引起自适应方向图畸变,导致期望信号相消和杂波抑制性能下降。利用目标信号与欺骗式干扰信号极化特性或者空间达到角的差异,提出了一种极化-空域联合自适应波束形成算法来抑制欺骗式主瓣或旁瓣干扰。在增加了极化辅助通道的基础上,首先采用重叠滑窗的子阵合成方式进行极化-空域联合自适应波束形成,然后通过STAP抑制剩余的杂波。该方法可以有效地滤除密集欺骗式干扰,减少由其引起的虚警,减轻后续雷达数据处理器的负担,改善机载STAP雷达的杂波抑制性能。四、研究了机载前视阵雷达检测近距离地面运动目标时的杂波抑制问题。机载前视阵雷达的近程杂波具有很强的距离非平稳性,易使STAP的杂波抑制性能下降。在采用中脉冲重复频率探测近距离运动目标时,模糊的远程杂波会带来额外的盲速。基于谱补偿的STAP算法可以用来对付距离非平稳的近程杂波,但它们忽略了模糊的远程杂波的影响。提出了一种高效的基于波束空间的俯仰预滤波算法来滤除远程杂波,从而消除远程杂波对目标检测的影响。所提俯仰波束形成算法不仅具有很好的远程杂波抑制性能,还具有保护近程目标信号不受损失的优点。该算法不仅能够有效消除由远程杂波带来的额外盲速,还能够改善后续STAP的杂波抑制性能。另外,所提基于波束空间的降维自适应波束形成算法不仅能够加快收敛速度,还能够降低运算量,有利于工程实时实现。五、研究了单基地MIMO雷达目标波达方向(Direction of Arrival,DOA)估计的问题。在低信噪比、低快拍数的情况下,常规DOA估计算法的性能会严重下降。针对此问题,提出了一种新的DOA估计算法:降维酉ESPRIT算法。该算法首先通过降维变换将MIMO雷达数据变换至低维信号空间,然后在该低维信号空间构造实值旋转不变性方程估计目标的DOA。该方法能够在低信噪比、低快拍数的恶劣环境下获得较常规ESPRIT方法更高的DOA估计精度,同时具有更低的运算量。
[Abstract]:Airborne radar as the work is easily affected by the strong ground clutter. Time domain filtering and spatial filtering to effectively filter the traditional diffusion in Doppler domain and spatial clutter. Space time adaptive processing (Space-Time Adaptive, Processing, STAP) technology combined with a number of spatial channels and coherent pulse two-dimensional adaptive filtering at the same time, can restrain the clutter target to ensure coherent processing gain of airborne radar clutter suppression is a key technology. In order to obtain STAP better performance, the theory is used to estimate the statistical characteristics of clutter i.i.d. samples should not be less than two times the number of degrees of freedom but the actual system. Clutter heterogeneity will lead to the STAP processor can not get so many i.i.d. samples, resulting in STAP clutter suppression performance severely. How to improve the STAP in Non uniform clutter suppression performance of environment is a very important problem. Multiple input multiple output (Multiple-Input Multiple-Output MIMO) radar is a new radar system in recent years, it has the capability to launch each antenna element of arbitrary waveform. Compared with the conventional phased array radar, MIMO technology can significantly increase the system freedom, virtual aperture extension of the array, the goal of improving the precision of parameter estimation. With the increase of the degree of freedom of the system, conventional parameter estimation methods, the computational complexity will be increased at the same time, low SNR, low number of snapshots and other non physical factors would cause a serious decline in the parameter estimation accuracy. In low SNR, improve MIMO radar parameter estimation accuracy of non ideal low number of snapshots, and reduce the amount of calculation is a problem worthy of study. This thesis focuses on the above problems, including the following, Study on airborne sidelooking radar non distance nonstationary clutter suppression for airborne radar antenna. When the non sidelooking installation, the clutter spectrum will change with the change in distance, which leads to the clutter from the non-stationary clutter phenomenon. The distance of non stationarity will lead to the traditional STAP cannot get independent with the distribution of training samples. In high repetition rate mode, the situation will become more complex, short-range clutter and remote clutter due to the distance of fuzzy coupling together, clutter suppression becomes more difficult. This paper proposes a pitch adaptive beamforming algorithm based on orthogonal waveform distance used to suppress the non-stationary short-range clutter, and remote the remaining distance stationary clutter range by Doppler STAP to suppress. The pitch beamforming algorithm can effectively suppress the short-range clutter, and can protect the remote target. In the distance filter Non stationary short-range clutter, clutter sample distance stability was increased, for subsequent range - Doppler STAP processor provides independent and identically distributed training samples, the clutter suppression performance can be improved significantly. Two, studied the non-uniform environment using direct data domain method for problem STAP clutter suppression. The clutter power will lead to the non-uniform clutter suppression is not sufficient or notch too deep, while the discrete clutter can easily lead to false alarm, dense moving target environment is likely to cause STAP adaptive pattern distortion and target self cancellation. The direct data domain method to solve the homogeneous clutter the problem. But the direct data domain method of traditional space-time aperture loss and array structure limitation, puts forward an iterative adaptive spectral estimation algorithm of the direct data domain STAP method based on this method. The problem can effectively avoid the traditional statistical STAP algorithm in the face of non-uniform environment samples and sample the lack of pollution, and no loss of aperture of space-time array for arbitrary shape. Three, clutter suppression of airborne radar in dense deception jamming environment (dense deception jamming environment as a special case of non-uniform environment). Deception sidelobe interference of airborne radar can cause a large number of false alarm, the radar tracking data processor and saturation. The deceptive mainlobe jamming not only causes a large amount of false alarms, but also pollute the space-time adaptive processor of the training sample, caused by adaptive pattern distortion, leading to the desired signal cancellation and clutter suppression performance decreased. The difference reached angle by using the target signal and deception interference signal polarization or space, presents a polarization space adaptive beamforming The algorithm for suppressing the deceptive mainlobe and sidelobe interference. In increasing the basic polarization of the auxiliary channel, first using the overlapping sliding window of sub array synthesis of polarization space adaptive beamforming, and then through the STAP to suppress the residual clutter. This method can effectively suppress the dense deceptive jamming, reduce false alarm by the radar data processor, reduce the follow-up burden, improve the performance of clutter suppression for airborne STAP radar. Four, the research of airborne phased array radar detection of near ground moving target when the clutter suppression for airborne phased array radar. The short-range clutter has strong distance nonstationarity, easy to STAP the performance of clutter suppression is decreased. In the pulse repetition frequency detection near the moving object, fuzzy remote clutter will bring additional blind speed. STAP spectrum compensation algorithm can be used to deal with the non distance based on Smooth the short-range clutter, but they ignore the fuzzy remote clutter. This paper presents an efficient pitch filtering algorithm based on beam space distance to filter out clutter, thus eliminating the remote effect of clutter target detection. The proposed beamforming algorithm not only has the pitch remote good clutter suppression the performance, also has the advantages of protecting the short-range target signal without loss. The algorithm not only can effectively eliminate clutter by remote additional blind speed wave, can also improve the performance of clutter suppression following STAP. In addition, the proposed adaptive beamforming algorithm can not only accelerate the convergence speed of dimension reduction based on beam space, but also can reduce the amount of computation to realize in engineering. In five, the monostatic MIMO radar target DOA (Direction of Arrival, DOA) estimation problem. In low SNR, low sampling number under the condition of conventional DOA The performance of the estimation algorithm will decrease seriously. To solve this problem, we propose a new DOA estimation algorithm: dimensionality reduction unitary ESPRIT algorithm. The algorithm firstly reduced dimensional transformation of MIMO radar data is transformed into a low dimensional signal space, then this method can estimate the target DOA. in the low SNR in the low dimensional signal the space structure of real valued rotation invariance equation, compared with the conventional ESPRIT method DOA higher estimation accuracy for low number of snapshots in the harsh environment, and has lower computation.

【学位授予单位】：西安电子科技大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TN957.51

【相似文献】