基于互相关函数相角特征的RBF神经网络来波方位估计

张旻; 李鹏飞

doi:10.3724/SP.J.1146.2008.01677

基于互相关函数相角特征的RBF神经网络来波方位估计

doi: 10.3724/SP.J.1146.2008.01677

张旻,
李鹏飞

基金项目:

国家自然科学基金(60672161)资助课题

计量
- 文章访问数: 3624
- HTML全文浏览量: 67
- PDF下载量: 723
- 被引次数: 0
出版历程
- 收稿日期: 2008-12-10
- 修回日期: 2009-07-16
- 刊出日期: 2009-12-19

Direction of Arrival Estimation Approach Based on Phase Angle Feature of Correlation Function Using RBF Neural Networks

摘要

摘要: 有效的方位特征获取对构建智能来波方位估计模型具有十分重要的意义。该文在分析阵列接收信号相关函数的基础上，首次提出利用相邻阵元信号互相关函数的相角作为来波方位特征。与常用的协方差矩阵上三角特征相比，剔除了与来波方位无关的幅度信息和冗余的方位特征信息，在不损失有效方位信息的基础上使特征维数得到极大地降低。实验结果表明，利用相角特征构建的RBF神经网络的结构更简洁，泛化性能更好，来波方位估计精度高，实时性好，具有广阔的工程应用价值。
- 信号处理; DOA估计; 协方差矩阵; 相角特征; 径向基神经网络
Abstract: Effective feature extraction is very important when building the smart DOA estimation model. Based on analyzing the correlation function of the array signal, this paper firstly presents using the angles of contiguous array signals correlation function for DOA estimation purpose instead of common used upper triangular half of the covariance matrix, which eliminates the irrelevant magnitude information and redundant direction characteristic. Therefore the feature dimension is largely reduced without losing any DOA information. Experimental results show that the performance of RBF neural network using proposed Phase-feature is superior to the common used upper triangular half of the covariance matrix in terms of neural network size, generalization, estimation precision and real-time performance, so it has a broad application value.

HTML全文

参考文献(1)

安冬, 王守觉. 基于仿生模式识别的DOA 估计方法[J].电子与信息学报.2004, 26(9):1468-1472浏览[2]Guo Wo, Qiu T S, and Tang H, et al.. Performance of RBFneural networks for array processing in impulsive noiseenvironment [J].Digital SignalL Processing.2008, 18(2):168-178[3]Wang M, Yang S, and Wu S, et al.. A RBFNN approach forDOA estimation of ultra wideband antenna array[J].Neurocomputing. 2008, 71(4-6): 631-640.[4]Vigneshwaran S, Sundararajan Narasimhan, andSaratchandran P. Direction of arrival (DOA) estimationunder array sensor failures using a minimal resourseallocation neural network[J].IEEE Transactions on Antennasand Propagation.2007, 55(2):334-343[5]Dourado O D, Doria A D, and Da Mata W. Determination ofmultiple direction of arrival in antennas arrays with radialbasis functions[J].Neurocomputing.2006, 70(1-3):55-61[6]Kuwahra and Matsumoto. Experiments of direction finder byRBF neural network with post processing. IEEE ElectronicLetters, 2005, 41(10): 24-25.[7]严颂华, 吴世才, 吴雄斌. 基于神经网络的高频地波雷达目标到达角估计[J].电子与信息学报.2008, 30(2):339-342浏览[8]于斌, 尹成友, 黄冶. 阵列误差影响下的神经网络波达方向估计[J]. 微波学报, 2007, 23(6): 21-25.Yu Bin, Yin Cheng-you, and Huang Ye. Direction ofarrival(DOA) estimation for an array with errors using RBFneural network [J]. Journal of Microwaves, 2007, 23(6): 21-25.[9]安冬, 王守觉. 基于仿生模式识别和PCA/ICA 的DOA 估计方法[J]. 电子学报, 2004, 32(9): 1448-1451.An Dong and Wang Shou-jue. A DOA estimation methodbased on biomimetic pattern recognition and PCA/ICA [J].Acta Electronica Sinica, 2004, 32(9): 1448-1451.[10]Shieh Ching-Sung and Lin Chin-Teng. Direction of arrivalestimation based on phase differences using neural fuzzynetwork[J].IEEE Transactions on Antennas and Propagation.2000, 48(7):1115-1123[11]陈荆花, 易辉跃, 周希朗. 基于特征矢量输入的神经网络测向方法[J]. 上海交通大学学报, 2003, 37(3): 373-379.Chen Jing-hua, Yi Hui-yue, and Zhou Xi-lang. Direction ofarrival estimation method with eigenvector based radial basisfunction neural network[J]. Journal of Shanghai Jiao TongUniversity, 2003, 37(3): 373-379.

施引文献

资源附件(0)

访问统计