高动态GNSS接收系统压制性干扰抑制技术研究

发布时间：2018-01-09 01:07

本文关键词：高动态GNSS接收系统压制性干扰抑制技术研究　出处：《武汉大学》2016年博士论文　论文类型：学位论文

【摘要】：GPS为代表的GNSS在取得广泛应用的同时,也暴露出严重的系统脆弱性问题：系统的抗干扰能力差。特别是高动态接收系统,其所处的电磁环境复杂,静态和低动态的干扰抑制技术不能直接套用,提高高动态接收机自身的干扰抑制能力及其可靠性是近年来的研究热点。本文围绕高动态环境下如何抑制压制性干扰展开理论研究。主要研究工作归纳如下：第一、建立多重信号分类及功率倒置联合抗干扰模型(MUSIC-PI)。传统的功率倒置算法(PI)无法区分干扰和信号,对期望信号有较大损失,MUSIC-PI算法将多重信号分类(MUSIC)的基本原理应用于实际导向矢量的估计,利用正交空间能量最小化原理辅助零陷形成。本模型可估计出最优权值,有效地抑制压制性干扰的同时减少对卫星信号的损伤。当输入信噪比降到为20dB以下时,最小均方误差算法(LMS)和PI算法失效,而MUSIC-PI算法能准确产生零陷,抑制深度达到-120dB,且稳健性提高明显。第二、研究高动态干扰时变问题,提出基于隐马尔可夫(HMM)和多星约束最小功率(Multi-PI)级联干扰抑制模型,即HMM-Multi-PI模型。HMM-Multi-PI模型的第一级是HMM算法；第二级是一种从功率倒置引导至多星约束最小的引导算法(PIMulti-PI),这是一种引导模型,其先用PI算法进行初级干扰抑制,再用Multi-PI算法来提升干扰抑制性能。捕星效果从好到差的顺序是：HMM-Multi-PI算法、Multi-PI算法、HMM算法和PI算法；实验结果表明：HMM-Multi-PI算法对抑制时变干扰、窄带宽带混合干扰是有优势的。第三、研究高动态干扰源波达方向(DOA)估计问题,提出一种基于DOA的凸优化约束模型(CCOB),能有效改善错配下的DOA性能,提升了系统的鲁棒性。针对卫星信号较弱,无法通过波束形成来确定DOA的特点,提出无主瓣零陷模型(NMLC),其只需估计压制干扰的DOA。并在此基础上,在DOA周边增加虚拟零陷,同时将剩余的空间作功率最大化处理,此算法称为无主瓣零陷展宽模型(NMLCB)。抑制深度从高到低的顺序是：NMLCB算法、导数约束算法(DN)和PI算法；零陷宽度从宽到窄的顺序是：NMLCB算法、DN算法和PI算法。实验结果表明：NMLCB算法对解决NSS接收机的高动态DOA失配问题是有效的。第四、研究高动态干扰源的欠采样问题,建立采样信号的凸优化重建模型(CCOB-CS)。本模型先对采样的信号进行CCOB处理,增强DOA的鲁棒性,再运用正交匹配追踪法(OMP)进行信号重建。实验结果表明：利用压缩感知理论,将稀疏重建原理应用于高动态抗干扰阵列处理领域可解决采样中遇到的少快拍问题。相对于传统抗干扰模型,信干噪比(SINR)提升明显。第五、研究欠采样DOA估计问题,建立l0范数的连续压缩感知的DOA估计框架(ZNM),这是一种无网格模型,突破了信号子空间模型的原有框架,适用于少快拍DOA估计问题。相对于经典的空间谱分析算法,如MUSIC算法、求根多重信号分类算法(root-MUSIC)、最小二乘旋转不变子空间算法(TLS-ESPRIT)、Toeplitz近似算法(TAM)。它们之间的误差关系从大到小的顺序是：TLS-ESPRIT算法、TAM算法、MUSIC算法、root-MUSIC算法和ZNM算法。实验结果表明：ZNM的误差最小,零延迟下码捕获实验验证了更高的DOA估计准确性,其有利于提升干扰抑制效果。第六、研制一套FPGA+DSP软硬一体的阵列GPS抗干扰系统,其主要模块有：干扰源、八路射频前端、基于FPGA+DSP的基带板。通过三个实验分别评测了干扰类型、采样频率和干扰源DOA三个参数变化对GPS接收机及干扰抑制效果的影响。综上所述,本文对高动态GPS接收机的干扰抑制问题开展了深入研究,提出了应对干扰时变性、干扰DOA随机性和欠采样性的相应理论解决方案,并开展了GPS抗干扰原型机的研制工作,验证了设备的可用性。本文提出的理论和实验结果对高动态干扰抑制领域的下一步研究工作奠定了基础。
[Abstract]:GPS as the representative of the GNSS in a broad application at the same time, also exposed the serious vulnerability problem: the anti-interference ability of the system is poor. Especially the high dynamic receiver system, the electromagnetic environment of the complex, static and dynamic low interference suppression technology can not be directly applied to the interference, improve its high dynamic receiver suppression ability and reliability is the research hotspot in recent years. This paper focuses on the high dynamic environment how to suppress the oppressive jamming theory research. The main work is as follows: first, the establishment of multiple signal classification and power inversion combined with anti interference model (MUSIC-PI). The traditional power inversion algorithm (PI) can not distinguish between interference and signal, have great loss to the desired signal, the MUSIC-PI algorithm to multiple signal classification (MUSIC) estimation principle is applied to the actual steering vector, using the principle of orthogonal spatial energy minimization Auxiliary nulling. This model can estimate the optimal weights, and effectively suppress the oppressive jamming of satellite signals to reduce damage. When the input signal-to-noise ratio is below 20dB to drop, the minimum mean square error algorithm (LMS) failure and PI algorithm, and MUSIC-PI algorithm can accurately generate null suppression depth to -120dB, and the robustness is improved significantly. Second. Research on variable high dynamic interference is proposed based on Hidden Markov model (HMM) and multiple constrained minimum power (Multi-PI) interference suppression cascade model, HMM-Multi-PI model,.HMM-Multi-PI model is the first stage of HMM algorithm; the second grade is a guide to the star from the power inversion constrained minimum to guide the algorithm (PIMulti-PI), which is a leading model, first using PI algorithm for primary interference suppression, and then Multi-PI algorithm to improve the performance of interference suppression. A star effect from good to bad order is: HMM-M Ulti-PI algorithm, Multi-PI algorithm, HMM algorithm and PI algorithm; the experimental results show that the HMM-Multi-PI algorithm of interference suppression, narrowband broadband hybrid interference is advantage. Third, research on high dynamic interference source direction of arrival (DOA) estimation problem, propose a convex optimization model based on DOA (CCOB), can effectively to improve the performance of DOA mismatch, enhance the robustness of the system. According to the satellite signal is weak, to determine the characteristics of DOA cannot be formed through the beam, the main lobe null model (NMLC), the only need to estimate the suppression of interference and DOA. on the basis of this, around DOA to increase the virtual null at the same time. The remaining space for maximum power processing, the algorithm called the mainlobe null broadening model (NMLCB). The depth of inhibition from high to low order is: NMLCB algorithm, derivative constraints (DN) algorithm and PI algorithm; null width from broad to narrow the order is: NMLC B algorithm, DN algorithm and PI algorithm. The experimental results show that the NMLCB algorithm for high dynamic DOA receiver NSS to solve the mismatch problem is effective. Fourth, the undersampling problem of high dynamic interference source, the convex optimization reconstruction model of sampling signal (CCOB-CS). The model of sampling signal processing CCOB and enhance the robustness of DOA, and then using the orthogonal matching pursuit method (OMP) for signal reconstruction. The experimental results show that the use of compressed sensing theory, the sparse reconstruction principle is applied to the high dynamic interference array processing can solve the problems encountered in small snapshot sampling. Compared with the traditional model of anti-jamming, SINR (SINR) improved significantly. Fifth, research on undersampling DOA estimation problem, establish the l0 Norm Continuous compressed sensing DOA estimation framework (ZNM), which is a meshless model, a framework of breakthrough original signal subspace model, suitable for small fast The problem took DOA estimation. Compared with the classical spatial spectrum analysis algorithms such as MUSIC algorithm, rootmusic algorithm (root-MUSIC), least squares rotation invariant subspace algorithm (TLS-ESPRIT), Toeplitz (TAM) approximation algorithm. The error of the relationship between them in order from big to small is: TLS-ESPRIT algorithm, TAM algorithm, MUSIC algorithm, root-MUSIC algorithm and ZNM algorithm. The experimental results show that the ZNM minimum error, zero delay capture experiment code verification in higher DOA estimation accuracy, it is conducive to improving the interference suppression effect. Sixth, the anti-jamming array GPS has developed a set of FPGA+DSP software and hardware integration, the main modules are: interference source, eight road RF, baseband board based on FPGA+DSP. Through three experimental types of interference evaluation respectively, effects of sampling frequency and interference source DOA change three parameters of inhibitory effects on GPS receiver and interference at all. In this paper the interference of high dynamic GPS receiver suppression problem to carry out in-depth research, put forward the corresponding interference interference DOA degeneration, randomness and less sampling of corresponding theoretical solutions, and developed the GPS anti-jamming prototype, verify the availability of the equipment. This paper presents theoretical and experimental results for the next step of the research work based interference suppression in the field of high dynamic.

【学位授予单位】：武汉大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：TN965.5

【相似文献】