An end-to-end DOA estimation method based on deep learning for underwater acoustic array

The Direction of arrival (DOA) estimation of signals is a fundamental problem in array signal processing and one of the tasks in many fields such as radar arrays and sonar arrays. In this paper, we propose an end-to-end deep learning-based network structure for DOA estimation task of underwater acou...

Full description

Saved in:
Bibliographic Details
Published inOCEANS 2022, Hampton Roads pp. 1 - 6
Main Authors Peng, Junyu, Nie, Weihang, Li, Ta, Xu, Ji
Format Conference Proceeding
LanguageEnglish
Published IEEE 17.10.2022
Subjects
Acoustic arrays
Array signal processing
Convolutional neural network
Deep learning
Direction-of-arrival estimation
DOA estimation
Estimation
Parallel processing
Signal processing algorithms
Underwater acoustic array signal processing
Online AccessGet full text

Cover

More Information
Summary:The Direction of arrival (DOA) estimation of signals is a fundamental problem in array signal processing and one of the tasks in many fields such as radar arrays and sonar arrays. In this paper, we propose an end-to-end deep learning-based network structure for DOA estimation task of underwater acoustic arrays, and train the model by extracting array features. We compared the proposed model with traditional methods such as CBF and MVDR, and found that the method proposed in this paper can effectively identify the incoming wave direction of unknown signals in water after training. Compared with various traditional algorithms, the method proposed in this paper significantly improves the effective measurement range of DOA and reduces the angular resolution.
DOI:10.1109/OCEANS47191.2022.9977187