Addressing the issues of low DOA estimation accuracy of array signals under conditions of small snapshots, multiple sources, and coherent sources in traditional subspace and existing sparse reconstruction algorithms, a grid-less DOA estimation method is proposed based on deep learning and atomic norm minimization. Firstly, under the premise of the Toeplitz matrix Vandermonde decomposition theorem, the Toeplitz matrix is constructed by the correlation between pressure and velocity from the vector hydrophone. Then, the Toeplitz matrix is reconstructed using the deep learning-based alternating direction method of the multipliers algorithm. Finally, the Root-music algorithm is used to decompose the Toeplitz matrix and obtain the DOA estimates. Simulation experiments show that this algorithm has a lower root mean square error compared to other traditional algorithms and some sparse reconstruction algorithms under the above conditions, demonstrating a clear advantage in DOA estimation performance.
ZHONGS, XUL, YANGH, et al. DOA estimation of underwater object based on cross-spectrum deconvolution algorithm[C]//Proceedings of Chinese Intelligent Automation Conference. Singapore: Springer, 2022: 627-636.
[2]
PENGS, CHENB, YANGM. Joint sparse recovery for direction of arrival based on the generalized MUSIC criterion[J]. Digital Signal Processing, 2022, 122: 103382.
[3]
BENSALEMM, BARKATO. DOA estimation of linear dipole array with known mutual coupling based on ESPRIT and MUSIC[J]. Radio Science, 2022, 57(2): 1-15.
[4]
WANGP, HUANGJ, HEW, et al. Maximum likelihood DOA estimation based on improved invasive weed optimization algorithm and application of MEMS vector hydrophone array[J]. AIMS Mathematics, 2022, 7(7): 12342-12363.
HEWeijia, HUANGJiajun, ZHANGJingqi, et al. Joint angle-frequency estimation with maximum likelihood based on tuna swarm optimization algorithm[J]. Journal of Test and Measurement Technology, 2023, 37(1): 60-65. (in Chinese)
[7]
DONOHOD L. Compressed sensing[J]. IEEE Transactions on Information Theory, 2006, 52(4): 1289-1306.
[8]
CANDESE J, ROMBERGJ, TAOT. Robust uncertainty principles: exact signal reconstruction from highly incomplete frequency information[J]. IEEE Transactions on Information Theory, 2006, 52(2): 489-509.
[9]
ZHUS, CHENS, PENGX, et al. A signal reconstruction method of wireless sensor network based on compressed sensing[J]. EURASIP Journal on Wireless Communications and Networking, 2020(1): 106.
DIAOHongyang, HUZhouyong, YUYongzhi. An improved generalized orthogonal matching pursuit method used in the direction of arrival estimation[J]. Applied Science and Technology, 2020, 47(4): 54-58. (in Chinese)
[12]
KUKREJAS L, LÖFBERGJ, BRENNERM J. A least absolute shrinkage and selection operator (LASSO) for nonlinear system identification[J]. IFAC Proceedings Volumes, 2006, 39(1): 814-819.
LIUYonghao, XUMing. DOA estimation scheme of vector hydrophone by alternating direction multiplier method[J]. Ship Science and Technology, 2024, 46(12): 140-143. (in Chinese)
ZHENGWenkang, WEIZhiqing, BAIYanping, et al. DOA estimation based on separable surrogate function algorithm[J]. Journal of Shaanxi University of Science & Technology, 2024, 42(1): 197-205. (in Chinese)
SUNGoucang, HUIJunying, GUOLongxiang. Combined frequency estimation of sound pressure and velocity[J]. Computer Engineering and Applications, 2008(10): 154-156. (in Chinese)
YAOZhixiang, JIANGKeyu, GUORui, et al. An improved bearing estimation algorithm using acoustic vector sensor array based on rotational invariance technique[J]. Transactions of Beijing Institute of Technology, 2012, 32(5): 513-516. (in Chinese)
ZHANGXu, SHIShengguo, ZHUXiaochun, et al. Direction of arrival estimation by joint processing of vector array sound pressure and vibration velocity using covariance matrix decomposition[J]. Acta Acustica, 2023, 48(5): 937-949. (in Chinese)
[25]
WAGNERM, PARKY, GERSTOFTP. Gridless DOA estimation and root-MUSIC for non-uniform linear arrays[J]. IEEE Transactions on Signal Processing, 2021, 69: 2144-2157.
[26]
YANGZ, LIJ, STOICAP, et al. Sparse methods for direction-of-arrival estimation[M]. Amsterdam: Elsevier, 2018: 509-581.
[27]
BHASKARB N, RECHTB. Atomic norm denoising with applications to line spectral estimation[C]//2011 49th Annual Allerton Conference on Communication, Control, and Computing (Allerton), 2011: 261-268.
[28]
XIAOP, LIAOB, DELIGIANNISN. DeepFPC: a deep unfolded network for sparse signal recovery from 1-Bit measurements with application to DOA estimation[J]. Signal Processing, 2020, 176: 107699.
[29]
LIY, CHIY. Off-the-grid line spectrum denoising and estimation with multiple measurement vectors[J]. IEEE Transactions on Signal Processing, 2016, 64(5): 1257-1269.
[30]
TOHK C, TODDM J, TÜTÜNCÜR H. On the implementation and usage of SDPT3–A Matlab software package for semidefinite-quadratic-linear programming, version 4.0[M]. New York: Springer, 2012: 715-754.
[31]
BOYDS. Distributed optimization and statistical learning via the alternating direction method of multipliers [J]. Foundations and Trends in Machine Learning, 2010, 3(1): 1-122.
[32]
LIUS, MAOZ, ZHANGY D, et al. Rank minimization-based toeplitz reconstruction for DoA estimation using coprime array [J]. IEEE Communications Letters, 2021, 25(7): 2265-2269.
[33]
BARABELLA. Improving the resolution performance of eigenstructure-based direction-finding algorithms[C]//ICASSP IEEE International Conference on Acoustics, Speech, and Signal Processing, 1983: 336-339.
[34]
RAOB D, HARIK V S. Performance analysis of root-music[J]. IEEE Transactions on Acoustics, Speech, and Signal Processing, 1989, 37(12): 1939-1949.
[35]
ZHUH, FENGC, FENGW, et al. A deep learning approach for sparse single snapshot DOA estimation[J]. Journal of Signal Processing, 2022, 38: 2114-2123.
[36]
RUMELHARTD E, HINTONG E, WILLIAMSR J. Learning representations by back-propagating errors[J]. Nature, 1986, 323: 533-536.