西安电子科技大学学报 ›› 2022, Vol. 49 ›› Issue (1): 181-187.doi: 10.19665/j.issn1001-2400.2022.01.018

• 信息与通信工程 • 上一篇    下一篇

阵列误差下的近场源PCA-BP参数估计算法

王乐1(),赵佩瑶2(),王兰美1(),王桂宝3()   

  1. 1.西安电子科技大学 物理与光电工程学院,陕西 西安710071
    2.西安电子科技大学 通信工程学院,陕西 西安710071
    3.陕西理工大学 物理与电信工程学院,陕西 汉中 723001
  • 收稿日期:2020-12-14 出版日期:2022-02-20 发布日期:2022-04-27
  • 通讯作者: 王兰美
  • 作者简介:王 乐(1997—),女,西安电子科技大学硕士研究生,E-mail: lwang527@stu.xidian.edu.cn;|赵佩瑶(2000—),女,工学学士,E-mail: Monterey1001@icloud.com;|王桂宝(1977—),男,教授,博士,E-mail: gbwangxd@126.com
  • 基金资助:
    国家自然科学基金(61972239);国家自然科学基金(61772398);陕西省重点研发计划(2019SF-257);陕西省重点研发计划(2020GY-024)

Parameter estimation of the near-field source using the PCA-BPalgorithm with the array error

WANG Le1(),ZHAO Peiyao2(),WANG Lanmei1(),WANG Guibao3()   

  1. 1. School of Physics and Optoelectronic Engineering,Xidian University,Xi'an 710071,China
    2. School of Telecommunications Engineering,Xidian University,Xi'an 710071,China
    3. School of Physics and Telecommunication Engineering,Shaanxi University of Technology,Hanzhong 723001,China
  • Received:2020-12-14 Online:2022-02-20 Published:2022-04-27
  • Contact: Lanmei WANG

摘要:

当信号接收阵列存在误差时,阵列的导向矢量将会出现偏差,进而影响到参数估计算法的性能。为了减少阵列误差对参数估计结果的影响和降低计算复杂度,可以采用智能算法与主成分分析结合的方式。首先,利用后向传播神经网络方法将误差和其他因素包含在网络模型中,避开误差建模的繁琐过程;其次,由于后向传播神经网络训练近场源参数估计模型的时间过长,复杂度较高,为了缩短训练时间,减少计算量,在后向传播网络模型中引进主成分分析方法来降低信号特征矩阵维数,再把降维后的信号特征矩阵作为后向传播神经网络的输入特征,将近场源参数作为期望输出来进行训练,从而简化网络结构,减少训练过程中要估计的权值参数,缩短训练时间;最后,将包含待估计信号信息的接收数据输入到训练好的网络模型中,得到信号入射方向的估计值。该算法能够在接收阵列存在误差的情况下对近场源参数进行准确的估计,提高低信噪比下近场源信号参数的估计性能。仿真实验结果表明了该算法的有效性。

关键词: 主成分分析, 近场源, 后向传播神经网络, 到达角, 协方差矩阵

Abstract:

The steering vector of the array will be biased when there is an error in the signal receiving array,which will affect the performance of the parameter estimation algorithm.In order to reduce the influence of the array error on the parameter estimation results and reduce the computational complexity,a combination of intelligent algorithms and principal component analysis is used.First,in order to avoid the tedious process of error modeling,the back propagation neural network method is used to include errors and other factors in the network model.Second,it takes too long and is quite complicated for the back propagation neural network to train the near-field source parameter estimation model.In order to shorten the training time and reduce the amount of calculation,the principal component analysis method is introduced in the back propagation neural network model to reduce the dimension of the signal feature matrix.Then the reduced-dimensional signal feature matrix is used as the input feature of the back propagation neural network,and the near-field source parameters are used as the expected output for training,so as to simplify the network structure and shorten the training time.Finally,the received data containing signal information to be estimated is input into the trained network model to obtain the estimated value of the signal incident direction.This algorithm can accurately estimate the parameters of the near-field source in the presence of errors in the receiving array,and improve the estimation performance of the near-field source signal parameters under a low signal-to-noise ratio.Simulation experimental results show the effectiveness of the algorithm in this paper.

Key words: principal component analysis, near field source, the back propagation neural network, direction of arrival, covariance matrix

中图分类号: 

  • TN911.7