DOA estimation method for moving target echoes under multiple external radiation sources

Abstract

The invention belongs to the technical field of communication technology and signal processing, and discloses a DOA estimation method for moving target echoes under multiple external radiation sources, and the method comprises the steps: carrying out the preprocessing of a mixed echo signal received by an antenna array, solving a covariance matrix of the signal, extracting a real part and an imaginary part of an upper triangular element, and constructing a one-dimensional matrix as the input of a sparse auto-encoder; classifying the signals from different regions by using a sparse auto-encoder; p results output by the sparse auto-encoder forming a one-dimensional matrix, then converting the one-dimensional matrix into a covariance matrix form, and dividing the matrix into a real part matrix and an imaginary part matrix to serve as dual-channel input to be sent to P convolutional neural networks; realizing DOA estimation of different subarea signals by using the convolutional neural network, and output layer neurons of the P convolutional neural networks representing the angles of the P sub-regions in the horizontal direction; and when the signal-to-noise ratio is greater than 0dB,the normalized mean square error of signal-to-noise ratio estimation being less than 1.

Description

technical field [0001] The invention belongs to the technical fields of communication technology and signal processing, and in particular relates to a method for estimating DOA of echoes of moving targets under multiple external radiation sources. Background technique [0002] Currently, the closest state-of-the-art: Direction of Arrival estimation method based on Radial Basis Neural Network. Extract the real part and imaginary part of the triangular elements of the covariance matrix of the received signal to form a one-dimensional matrix as input, and use the radial basis neural network to realize the direction of arrival estimation of the incoming wave signal; compared with the traditional direction of arrival estimation method, the array defect The fitness is high; but because the relationship between the real part and the imaginary part of the covariance matrix elements is not fully utilized, the estimation accuracy is not very high. [0003] Direction-of-arrival (DOA) ...

AI Technical Summary

Problems solved by technology

[0006] To sum up, the problems existing in the existing technology are: the existing traditional DOA estimation method is not highly adaptable to different arrays, and the DOA estimation method based on deep learning has a large burden of network generalization when performing DOA estimation, and The relationship between input data is often not fully utilized, resulting in low accuracy of estimation results

[0007] The difficulty of solving the above technical problems: the existing deep learning-based DOA estimation methods often process the covariance matrix of the received signal as the extracted feature input neural network, the relationship between the real part and the imaginary part of the covariance matrix elements underutilized

Images