MIMO雷达波形优化与参数估计方法研究

发布时间：2018-04-10 00:16

  本文选题：MIMO雷达　切入点：波形设计　出处：《电子科技大学》2014年博士论文

【摘要】：多输入多输出(multiple-input multiple-output, MIMO)雷达近年来受到雷达界和相关学术界科研人员的高度关注和研究。相比传统的相控阵雷达,MIMO雷达利用波形分集和空间分集特性能够极大地提高角度分辨率、改进目标参数的识别能力、提高低速目标和弱目标的检测以及改善目标参数估计等方面性能,并能够突破传统体制雷达的性能限制。围绕MIM O雷达的波形分集与目标分辨、及处理复杂等带来的新问题,本论文提出MIMO雷达波形设计以及参数估计的新方法。本论文的主要工作及创新概况如下：(1)窄带MIMO雷达系统中的方向图合成和波形优化设计。首先,针对MIMO雷达中全方位(包括主瓣和旁瓣区域)同时逼近期望方向图的设计准则和求解方法,构造了一种MIM O方向图匹配设计协方差矩阵的统一优化框架,已有的一些方法可视作本框架下的特例。其次,实际中希望设计的方向图在主瓣区域逼近,而旁瓣区域能够控制到一定值,为此,提出一类将主瓣、旁瓣区域分开控制的MIMO雷达方向图合成,其包括三种具体的设计方法,其目的是保证期望区域信号无失真传输,且维持旁瓣低于预定的门限值。(2)单向和宽角度的宽带MIMO雷达发射波形设计。首先,考虑设计的波形具有单模或低的峰均比(PAPR)特性,提出一种基于单向发射方向图合成的波形设计方法。其次,围绕宽角度、宽带的MIMO雷达方向图合成,提出一种混合范数的宽角度的发射方向图合成方法,与最新的WBFIT方法相比,提出的方法设计方向图的主瓣在宽带频率段基本上保持恒定,且旁瓣控制在期望值-40dB以下。(3)静止目标的DOA及其反射信号的幅度估计。针对Capon和APES不能同时获得较高的分辨率和准确的幅度估计问题,提出一种基于稀疏表示的自适应参数估计方法。相比LS, Capon和APES,在目标θ=-25°和θ3=-21°很接近的情况,提出的方法也能提供精确的目标参数估计(4)运动目标的距离-方位-多普勒三维参数估计。针对运动目标的距离-方位-多普勒三维估计问题,提出一种迭代重复加权lq最小化方法。相比重复加权L1范数,提出的方法不需借助凸优化软件求解,只需通过多次迭代求解便可提供精确的目标三维估计；与DAS、最新的IAA和IAA-R相比,提出的方法不仅能在距离-方位二维截面的真实位置处获得精确地估计,而且也能在距离-多普勒单元处获得精确地估计。以上提出的MIMO雷达的波形设计与参数估计方法,都已利用MIMO雷达仿真实验数据进行了仿真,验证了新方法的有效性。
[Abstract]:Multi-input multiple-output multiple-input multiple-output (MIMOR) radar has attracted much attention and research in recent years.Compared with traditional phased array radar, MIMO radar can greatly improve the angle resolution and the recognition ability of target parameters by using waveform diversity and spatial diversity.The performance of low speed target and weak target detection and target parameter estimation can be improved, and the performance limitation of traditional system radar can be broken.In this paper, a new method of waveform design and parameter estimation for MIM O radar is proposed, which brings about some new problems such as waveform diversity and target resolution, and complex processing.The main work and innovation of this thesis are as follows: 1) pattern synthesis and waveform optimization in narrowband MIMO radar system.Firstly, a unified optimization framework of MIM O pattern matching design covariance matrix is constructed for the design criterion and solution method of omnidirectional (including main lobe and sidelobe region) approaching expected pattern in MIMO radar simultaneously.Some existing methods can be regarded as special cases under this framework.Secondly, the desired pattern is approximated in the main lobe region, and the sidelobe region can be controlled to a certain value. For this reason, a class of MIMO radar pattern synthesis is proposed, which can control the main lobe and sidelobe region separately.It includes three specific design methods, the purpose of which is to ensure that the desired region signal is transmitted without distortion, and to keep the sidelobe below the predetermined threshold value. (2) Wideband MIMO radar waveform design with unidirectional and wide angle.Firstly, considering the single mode or low peak to average ratio (PAPR) characteristics of the designed waveform, a waveform design method based on unidirectional emission pattern synthesis is proposed.Secondly, around the wide angle, wideband MIMO radar pattern synthesis, a hybrid norm wide angle emission pattern synthesis method is proposed, which is compared with the latest WBFIT method.The proposed method is used to design the DOA of the stationary target and the amplitude estimation of the reflected signal of the main lobe in the wideband frequency band, and the sidelobe is controlled below the expected value of -40dB.An adaptive parameter estimation method based on sparse representation is proposed to solve the problem that Capon and APES can not obtain high resolution and accurate amplitude estimation at the same time.Compared with LS, Capon and APESs, the proposed method can also provide accurate range azimuth Doppler 3D parameter estimation of moving targets in the case of 胃 -25 掳and 胃 3 ~ (-21 掳).An iterative repeated weighted LQ minimization method is proposed for the range azimuth Doppler 3D estimation of moving targets.Compared with repeated weighted L1 norm, the proposed method does not need to be solved by convex optimization software, but only through multiple iterations to provide accurate three-dimensional estimation of the target, compared with Das, the latest IAA and IAA-R.The proposed method can obtain accurate estimation not only at the real position of the range-azimuth two-dimensional cross section, but also at the range-Doppler unit.The waveform design and parameter estimation of MIMO radar presented above have been simulated using MIMO radar simulation data, and the validity of the new method has been verified.
【学位授予单位】：电子科技大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TN957.51
，

本文编号：1728809