| 题名 | 稳健自适应波束形成理论及其实时处理体系研究 |
| 作者 | 陈模江 |
| 学位类别 | 博士 |
| 答辩日期 | 2009-05-27 |
| 授予单位 | 中国科学院声学研究所 |
| 授予地点 | 声学研究所 |
| 关键词 | 窄带波束形成 稳健自适应波束形成 盲波束形成 波束域预处理 实时系统 |
| 其他题名 | On robust adaptive beamforming algorithms and the dsp system design of their implements |
| 学位专业 | 信号与信息处理 |
| 中文摘要 | 阵列信号处理技术在声呐、雷达、无线通信等众多领域具有广泛的应用，是 近几十年来得以迅速发展的一个领域。波束形成是阵列信号处理的重要内容之 一，其处理过程是对传感器阵列所采集的数据进行线性组合处理得到一个标量波 束输出，目的是接收期望方向信号的同时对其他方向的干扰和噪声进行尽可能地 衰减。波束形成的加权系数可以是确定后不改变的固定波束形成器，也可以是随 应用环境变换不断进行自动调节的自适应波束形成器。Capon波束形成器是一种 在期望信号方向向量和噪声协方差矩阵精确已知的情况下，具有很好的分辨能力 与干扰抑制能力的自适应波束形成器。但是，当导向向量存在误差时，Capon波 束形成器把期望信号误认为是干扰进行抑制，造成阵列输出的信干噪比急剧下 降，甚至比常规波束形成器还差。因此，本文对稳健自适应波束形成算法展开了 讨论。另外，由于波束域处理可以获得比阵元域中进行处理更好的性能，本文也 对波束域预处理技术进行了探讨。最后，结合国防基础科研基金项目，深入研究 了稳健波束形成算法的工程实现相关问题。本文的主要研究内容及创新点包括： 1. 在阵列流形未知时，如果信号与干扰噪声场具有一些特定性质时，基于 高阶累积量和 CAB 类盲波束形成算法可以估计出期望信号方向向量，由于估计 的方向向量与真实方向向量有一定的误差，从而影响后续的波束形成器的性能。 本文借鉴稳健波束形成器的思想，将基于最差性能最佳化的稳健方法引入这两种 盲波束形成算法中，从而改进其稳健性。 2. 稳健 Capon 波束形成算法（RCB）的主要运算量集中在特征分解和牛顿 迭代法求解方程。通过深入研究RCB算法的求解过程， 本文提出了一种简化RCB 算法求解的方法，新的求解方法不仅在性能与 RCB 算法接近，而且更直观地揭 示了影响对角加载因子的物理量。 3. 波束域预处理技术的关键是波束转换矩阵设计，本文提出了基于 MVDR 的最佳波束转换矩阵设计算法，可以通过二阶锥规划方法对其求解。仿真实验表 明，该方法的波束转换矩阵增益能够保证通带无失真，并在干扰位置有较深的零 陷，而且阻带的衰减幅度是可以根据用户参数进行控制。 4. 本文详细分析了 RCB算法的运算量、存储器及吞吐量的需求，设计出了 基于 TMS320C6455为核心的多 DSP处理系统。利用湖试数据进行测试表明，该 系统能够满足声呐信号处理的数据精度及系统实时性要求。 |
| 英文摘要 | Array signal processing has wide applications in sonar, radar, wireless communications, and other fields. Beamforming is one of the most important tasks in array processing. The basic theory of beamforming is that spatial samples of propagating wave fields collected by sensor array are processed by a linear operator. Its objective is to receive the signal from a desired direction and attenuate those from other locations. Beamformers can be classified as either fixed or adaptive beamformers. The weights of a fixed beamformer do not depend on the array data, while those of an adaptive beamformer are chosen depending on the array data. The standard Capon beamformer, a classical adaptive beamformer, is known to have better resolution and much better interference rejection capability than the fixed beamformer (e.g. Delay-and-sum beamformer) when the array steering vector is accurately known. However, when the array steering vector is not imprecise, the Capon beamformer may suppress the signal of interest (SOI) as an interference, which results in drastically reduced array output signal-to-interference-plus-noise ratio (SINR). The performance of a Capon beamformer may become worse than that of a fixed beamformer. In this thesis, a complete analysis of the robustness of the adaptive beamformer is provided. Beam-space processing may improve the performance of an array over the element-space processing. This paper also discusses the beam-space preprocessing. Supported by the Commission of Science Technology and Industry for National Defense, this dissertation studies the practical application of robust adaptive beamformers with emphasis on sonar systems. The main contributions are as follows: 1. The estimated steering vector from cumulant-based and cyclic adaptive beamformers may be different from the true steering vector and such mismatch can degrade the performance of the sequent beamformer. By combining the worst-case performance optimization based robust beamformer and the cumulant-based and cyclic adaptive algorithms, we provide two new blind beamformers with improved robustness. 2. The computational load of the robust Capon beamformer (RCB) by Li et al focuses on the eigenvalue decomposition and solving equation based on Newton’s method. After deeply studies the RCB algorithm, an improved RCB algorithm was provided. This new method can reduce the computational load and directly give the parameter that affected the diagonal loading. The performance of the new method approximates that of the RCB. 3. The key step of the beam-space preprocessing is the design of the weight matrix. A new weight matrix design method is given in this dissertation, which can be solved using second-order cone programming (SOCP). Simulation results show that this algorithm can keep the passband (in-band) distortionless and form nulls in the directions of interferences. 4. According to the analysis of the computational load, memory, and communication of the RCB, a multiple-DSP system was designed. We test the system with the real data from lake-experiment. The testing results show that the system can be used in the real-time application. |
| 语种 | 中文 |
| 公开日期 | 2011-05-07 |
| 页码 | 131 |
| 内容类型 | 学位论文 |
| 源URL | [http://159.226.59.140/handle/311008/466]  |
| 专题 | 声学研究所_声学所博硕士学位论文_1981-2009博硕士学位论文 |
| 推荐引用方式 GB/T 7714 | 陈模江. 稳健自适应波束形成理论及其实时处理体系研究[D]. 声学研究所. 中国科学院声学研究所. 2009. |
| 个性服务 |
| 查看访问统计 |
| 相关权益政策 |
| 暂无数据 |
| 收藏/分享 |
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。
修改评论