【作者】 岳晋；

【导师】 冯文江；

【作者基本信息】 重庆大学 ， 电路与系统， 2005， 硕士

【摘要】 自适应阵列天线系统能根据信号的来波方向调整方向图,跟踪期望信号,减少或抵消干扰信号,提高信干噪比。在移动通信系统中,采用自适应阵列天线可以提高频谱利用率、增加系统容量、扩大基站覆盖范围、减小电磁污染,明显改善系统的通信质量。自适应阵列天线已经成为第三代移动通信研究的热点之一。在给定各阵元的取样数据后,自适应阵列天线信号处理的问题一般可以归结到三个方面:信号源数目的确定、波达方向(DOA: Direction of Arrival)的多用户检测和数字波束形成(DBF: Digital Beamforming)。这几个方面不是相互独立的,而是相互关联的。论文在查阅大量国内外相关科技文献资料的基础上,围绕自适应阵列天线技术中的波达方向估计和数字波束形成技术做了以下几个方面的研究: 1. 总结了传统的DOA 估计算法,在相关和非相关信号等条件下,对基于子空间的MUSIC 方法的DOA 估计性能进行了分析和仿真研究,并将其与传统的DOA估计方法进行了对比研究。2. MUSIC 算法是谱估计算法中的经典算法,它具有测向分辨率高,对信号个数和DOA 角度可以进行渐进无偏估计等优点,但是在分辨相关信号时需要进行平滑处理,增加了计算量。针对多用户多径情况下的DOA 估计问题,提出了一种空时平滑算法。该算法能有效识别出各来波的DOA,与传统的空间平滑技术相比,该算法分辨高,计算量小。3. 提出了基于最大特征值的空时平均处理算法。该方法结合基于最大特征值的DOA估计算法和空时平均处理方法而提出,在保持最大特征值法优点的基础上,引入空时平均处理技术,在低信噪比和小样本的情况下,能有效的提高DOA 估计的分辨率和准确度。4. 在深入研究基于特征空间(ESB: Eigenspace-based)波束形成算法性能的基础上,提出了基于最大特征值的ESB 波束形成算法。算法摒弃了权向量在干扰用户与多径信号空间中的分量,权向量范数更小,输出噪声功率较小,在提高信干噪比的同时能够获得比原ESB 算法更快的收敛速度,并对指向误差不敏感。此外,算法受相关干扰和多径干扰的影响小,算法的稳定性和鲁棒性较好。更多还原

【Abstract】 Adaptive Array Antennas (AAA) can adjust its beam patterns based on the arrival direction of signal, tracking the anticipant signal, decreasing or counteracting the interferences and improving the SIR (Signal to Interference Ratio). The introduction of adaptive array antennas in the mobile communication system will develop the using of frequency spectrum, enlarge the capability of the system, expand the radiation of the base station, reduce the electromagnetic pollution and improve the communication quality of the system obviously. Adaptive array antennas have become one of the spotlights of the Third Generation mobile communication. Given the sampled data of each element, the signal processing of AAA consists of three parts: the determination of source number, the DOA (Direction of Arrival) multi-user detection and DBF (Digital Beamforming). They are not independent, but associated with one another. Having read a lot of transactions in the area of interest, the author aims at the subject on DOA and DBF in AAA technologies and makes efforts to study the problems below: 1. The conventional DOA methods are summarized. The DOA estimation performance of MUSIC algorithm in the scenario of correlated signals and uncorrelated signals are studied. Then the performances of MUSIC and that of the conventional methods are compared with the simulation results given. 2. MUSIC algorithm is the classical algorithm to Spectral analysis, and it has many virtues expect that Spatial Smoothing, which the algorithm needs to distinguish correlated signals, result in excess work. The spatial-temporal smoothing algorithm is proposed to distinguish each DOA of different path of different user effectively. Compared with the traditional Spatial Smoothing method, the spatial-temporal smoothing method can achieve higher resolution. 3. An algorithm based on the largest eigenvalue using temporal-spatial averaging is proposed. The algorithm integrates the DOA algorithm based on the largest eigenvalue and temporal-spatial averaging method. It keeps virtues of the former algorithm and through temporal-spatial averaging processing the performance of DOA estimation is improved greatly, especially on the condition that SNR (Signal-to-Noise Ratio) is low and the number of snapshot is not enough. 4. ESB (Eigenspace-based) beamforming algorithm is introduced and an improved algorithm based on the largest eigenvalue is proposed on the base of it. The improved algorithm eliminates the weight of interference users and multi-path signals, decreases the norm of the weight vector and the output noise power, improves SINR and converges more rapidly than the original ESB algorithm. It is not sensitive to the direction error, correlated inference and multi-path inference and has quite good stability and robustness.更多还原