Combined noise reduction and DOA estimation algorithm for MEMS vector hydrophone based on variational mode decomposition

Abstract

Purpose This paper aims to solve the problem that strong noise interference seriously affects the direction of arrival (DOA) estimation in complex underwater acoustic environment. In this paper, a combined noise reduction algorithm and micro-electro-mechanical system (MEMS) vector hydrophone DOA estimation algorithm based on singular value decomposition (SVD), variational mode decomposition (VMD) and wavelet threshold denoising (WTD) is proposed. Design/methodology/approach Firstly, the parameters of VMD are determined by SVD, and the VMD method can decompose the signal into multiple intrinsic mode functions (IMFs). Secondly, the effective IMF component is determined according to the correlation coefficient criterion and the IMF less than the threshold is processed by WTD. Then, reconstruction is carried out to achieve the purpose of denoising and calibration baseline drift. Finally, DOA estimation is achieved by the combined directional algorithm of preprocessed signal. Findings Simulation and field experiments results show that the algorithm has good noise reduction and baseline drift correction effects for nonstationary underwater signals, and high-precision azimuth estimation is realized. Originality/value This research provides the basis for MEMS hydrophone detection and positioning and has great engineering significance in underwater detection system.