【作者】 戴凌燕；

【导师】 王永良；

【作者基本信息】 国防科学技术大学 ， 信息与通信工程， 2009， 博士

【摘要】 自适应波束形成技术从理论知识走向工程应用仍面临许多实际问题。其中,自适应波束形成算法在误差情况下的稳健性能直接影响到实际的应用效果。同时,实际环境中存在的一些特殊形式的干扰也给自适应波束形成技术带来了严重的影响,如多径相干干扰、闪烁干扰、灵巧干扰等。本文主要围绕算法稳健性和特殊干扰抑制两大问题展开研究,主要内容如下:研究了几种典型的基于不确定集约束的稳健自适应波束形成算法,明确了算法彼此间的关系。并对基于球形不确定集约束的稳健Capon法性能进行了深入分析,揭示了各种因素对自适应权矢量最优加载因子以及算法性能的影响关系。研究了基于球形不确定集约束的稳健最小方差无失真响应(MVDR: Minimum Variance Distortionless Response)波束形成算法,提出了一种球心位于期望信号导向矢量估计值的新的球形不确定集,不仅克服了输出信干噪比性能依赖于期望信号导向矢量误差模值上限的不足,而且在导向矢量严重失配时,仍具有良好的稳健性能。针对稳健MVDR波束形成的自适应方向图旁瓣电平较高问题,提出一种自适应副瓣控制方法。该方法可在波束指向误差情况下,对自适应方向图的任意副瓣区域进行副瓣电平控制,并获得理想的控制效果,因而实现了稳健自适应波束形成和低副瓣自适应方向图的有效结合。针对波束指向误差情况下相干干扰抑制问题,提出基于特征空间的迭代Toeplitz算法。该算法首先采用迭代Toeplitz化方法对相干干扰信号进行去相关处理,然后利用基于特征空间的自适应波束形成克服指向误差的影响。因而算法稳健性较强,尤其在小快拍时仍具有良好的干扰抑制性能。研究了严重非平稳环境下的干扰抑制问题,提出了利用单次快拍的直接数据域最小二乘优化算法,增加了系统自由度,保证了对期望信号的有效接收和对干扰信号的有效抑制,并降低了自适应方向图的旁瓣电平,从而显著改善了算法性能。研究了“灵巧干扰”的对抗方法,基于干扰样本识别与采集思想,分别提出了独立对抗“灵巧干扰”时空级联自适应旁瓣对消方法,以及同时对抗压制性噪声干扰和“灵巧干扰”的“对消-识别-对消”的干扰抑制方法,均取得了较好的干扰抑制效果。更多还原

【Abstract】 From theories to applications, many practical issues also exist for adaptive beamforming technique. For example, the robustness performance of adaptive beamforming algorithm would be an immediate cause for the effectiveness in practice. Moreover, some jammings with special styles, like multipath coherent jamming, blinking jamming, smart jamming and so forth, also bring serious impacts on traditional adaptive beamforming. Therefore, the investigations of this dissertation are focused on the two issues of the robustness of algorithm and peculiar jamming suppression. The main research contents are summarized as follows.Several representive robust adaptive beamforming algorithms based on uncertainty set constraints are investigated. And the relationships between them are clarified. The performance of robust Capon beamforming based on spherical uncertainty set constraint is deeply analyzed, which show clear that several parameters how to influence the optimum diagonal loading factor of the adaptive weight vector and the performance of the algorithm.A modified robust minimum variance distortionless response(MVDR) beamforming algorithm is presented, which is on the basis of a novel spherical uncertainty set constraint whose center locates the estimated steering vector of the desired signal. The method not only overcomes the shortcoming of the output performance dependent on the upper limit of of the norm of steering vector error, but also can hold good robustness under serious mismatches of steering vector of the desired signal.For the sidelobe level of adaptive pattern of robust MVDR beamforming algorithm is relative high, a sidelobe control approach is presented. In the case of beam pointing error, the approach is able to control the arbitrary sidelobe areas of the adaptive pattern and attains perfect control effects, which realizes the effective combination of robust adaptive beamforming and low sidelobe level of the adaptive pattern.An approach of iterative Toeplitz based on the eigenspace-based beamforming is presented for eliminating the coherent jammings under beam pointing error. The iterative Toeplitz technic is firstly applied to conduct decorrelation for the coherent jammings. And then the adaptive beamforming based on eigenspace is used to mitigate the impact caused by the pointing error. Therefore, the approach has better robustness, especially has better capability for jamming suppression in the presence of small sample size.The jamming suppression issue under highly nonstationary circumstance is investigated. An optimized direct data domain least squares approach using single snapshot is presented, which increases the adaptive freedoms, ensures adaptive reception for the desired signal and suppression for the jamming signals, decreases the sidelobe level of adaptive pattern, and therefore obviously improves the output capability.The approaches against smart jamming are investigated. Combined with the idea of jamming samples recognition and collection, the time-space cascade adaptive sidelobe canceling method only confronting the smart jamming and the“canceling-recognizing-canceling”method simultaneously confronting the noise jamming and the smart jamming is put forward respectively. The two methods both take on favorable effects for jamming suppression.更多还原