【作者】 郭艺；

【导师】 沈荣骏； 张尔扬；

【作者基本信息】 国防科学技术大学 ， 信息与通信工程， 2007， 博士

【摘要】 GPS接收机空时抗干扰方法相对于单纯时域、频域和空域抗干扰方法有明显优势。空时处理可以抑制多种干扰,在不增加天线阵元的情况下,抗干扰处理的自由度大大增加,抗干扰能力有质的提高。论文重点研究了GPS接收机空时抗干扰理论与实现关键技术。首先针对GPS接收机空时自适应滤波抗干扰技术的理论基础展开讨论。给出了相应的数学模型;分析了空时处理的最优准则;推导了LCMV准则各种不同约束形式下的最优权值表达式;通过对不同干扰(宽带干扰、部分频带干扰、窄带干扰)情况下,不同维数空时权矢量空频响应的数字仿真,得出空时自适应滤波抗干扰的一些有益结论,总结了各种干扰与空频自由度之间的对应关系;研究了空时抗干扰的信号失真问题,通过理论分析和仿真实验,指出在合适的最优准则和约束条件下,空时处理可以可靠地应用于GPS抗干扰,不会造成严重的信号失真。上述内容对GPS接收机空时抗干扰理论研究和实践有一定参考价值。空时处理在提高抗干扰能力的同时计算量显著增加,以降低计算量为主要目的的降维简化算法是GPS接收机空时抗干扰研究的重要问题。本文深入研究了多级维纳滤波(MWF)和辅助向量滤波(AVF)两种典型多步迭代降维方法,在一定加强条件下将这两种方法统一到同一降维框架内。在此基础上,推导分析了一种计算量小、性能优良的简化多步迭代降维(MIRR)方法,将其应用于GPS空时抗干扰中,取得良好效果。该方法采用并行结构,在数据域进行处理,适合硬件实现。针对空时抗欺骗干扰情况,提出部分解扩结合多步迭代降维的简化方法,克服了权矢量长度过长、计算量大、收敛速度慢,易受多普勒频偏影响等缺点。根据GPS空时抗强干扰接收信号中有用信号、噪声和干扰的不同分布特点,将子空间投影方法与空时处理相结合,推导出空时权值最优解的子空间投影表示。提出一种利用相关相减结构多级维纳滤波器(CSA-MWF)的分析滤波多级分解特性构建空时干扰子空间的方法;考虑到GPS空时抗干扰中干扰非常强的实际情况,针对原自适应PASTd子空间投影算法提出一种改进方法,增加了特征向量正交化和特征值对角加载步骤,用来估计干扰子空间,使得估计精度和收敛速度大大提高。针对非平稳的LFM干扰,提出用STFT-Hough变换方法提取干扰参数,精确重构干扰瞬时频率,结合干扰的空域特征构造空时干扰子空间。仿真结果表明,本文提出的几种方法能够简单、有效构建接收数据空时干扰子空间,将接收数据投影到空时干扰子空间的正交子空间中,从而达到抑制干扰的目的。深入分析了GPS空时抗干扰数字接收机设计中的AD量化问题。研究了无干扰情况下不同量化位数对接收机输出性能的影响;指出在一定量化条件下,强干扰下的接收机AD量化性能分析可转化为无干扰时的AD量化性能分析;建立一种简化的线性等效量化分析模型来研究接收机中频数字处理模块中有限字长运算对系统性能带来的影响。给出了一种GPS接收机空时抗干扰模块设计实现方案。详细讨论了其系统构成,给出天线阵列、射频、中频抗干扰、DSP控制等四个子模块的结构和设计指标;重点分析了中频抗干扰子模块的FPGA实现过程。更多还原

【Abstract】 Space-time interference suppression technology shows great predominance compared with only time, frequency or space field techniques for GPS receiver. Many kinds of interferences can be counteracted by space-time adaptive processing (STAP) technology. It can increase the degrees of freedom so remarkably that the anti-jamming capability can be enhanced essentially without increasing the number of antennas. This dissertation researched the theory and key realization technologies of space-time interference suppression for GPS receiver.Firstly, the dissertation discussed the theoretical foundation of space-time anti-jamming adaptive filtering for GPS receiver. The corresponding mathematical model was proposed; The optimal criterion of STAP was analyzed; The optimal weight expression in different constraints of LCMV criterion was deduced; The space-frequency response of STAP weight was simulated in different jammers (wideband, partialband and narrowband jammers) circumstances, the relationship was summarized between the jammer and degrees of space-frequency freedom and some available conclusions of space-time anti-jamming processing were drawed; The distortion of GPS signals induced by STAP was discussed. Through theoretical analysis and simulation experiments, it was pointed out that STAP could apply to GPS anti-jamming reliably in appropriate optimal criterions and constraints, and would not result in serious signal distortion. The discussion above has some referential value for the anti-jamming theory research and design of GPS space-time receiver.Along with the anti-jamming capability improvement through STAP, the computational complexity increases greatly. The reduced-rank simplification method aim at computation reduction is one of the key techniques for STAP. Two typical multistage iterative reduced-rank (MIRR) methods, multistage nested wiener filtering (MWF) algorithm and auxiliary vector filtering (AVF) algorithm were deeply analyzed and unified to the same reduced-rank frame on a certain enhanced condition. A simplified MIRR algorithm was deduced and analyzed to be used in space-time interference suppression for GPS receiver, which has low computational complexity and good performance. The algorithm has a parallel structure, processes receive signal in a data domain and suits hardware realization. For space-time spoofing interference suppression, a new approach combing PD (partial despreading) with MIRR was proposed, which overcomes the disadvantages like very long weight vector, large computation, slow convergence and Doppler shift influence.A new kind of approach was proposed for interference suppression of GPS receiver. It combines subspace projection with STAP based on the characteristics of GPS signals, noise and jammers. The optimal filter weight was deduced using subspace projection form. It was proposed that space-time interference subspace is constructed simply and effectively using the multistage decomposition character of the MWF based on the correlation subtractive architecture. Considering the strong interference in GPS receiver, an improved adaptive PASTd subspace tracking technique is proposed. It adds the steps of eigenvector orthogonalization and eigenvalue diagonal loading. The interference subspace is estimated more precisely and quickly. For LFM nonstationary interferences, it was proposed to detecte interferences and estimate parameters using STFT-Hough transformation, reconstructed interference instantaneous frequencies precisely. Combining the space characteristic, Space-time interference subspace is constructed. The numerical simulation results demonstrated that these algorithms proposed can construct space-time interference subspace simply and effectively. Interference suppression is achieved through projecting the received signal to the orthogonal subspace of space-time interference subspace.The quantization problem was deeply analyzed for designing a GPS space-time receiver. The loss of the receiver output SNR is researched for different length of AD quantization without interference. It was pointed out that the performance analysis of AD quantization with strong interferences can change to the quantization analysis without interferences on a certain condition. It was established a simple linearly equivalent model for analyzing the influence of the finite word length operation in IF-digitalized processor.A design scheme of space-time anti-jamming module was proposed for GPS receiver. The system construction was discussed. The structure and design target about the four submodules was given including antenna array submodule, RF submodule, IF anti-jamming submodule and DSP control submodule. The FPGA realization of IF anti-jamming submodule is analyzed emphatically.更多还原