DOA Estimation of Underwater Object Based on Cross-Spectrum Deconvolution Algorithm

Direction of arrival (DOA) estimation of an underwater object is crucial in Sonar system, especially for the low signal-to-noise ratio (SNR) with loud interference situations. The representative algorithms like conventional beamforming (CBF) and minimum-variance distortionless response (MVDR) have either high-resolution or robustness in such condition. The deconvolved CBF (DCBF) algorithm, which develops from the CBF algorithm, has a low background level and narrow beamwidth while staying robust. However, the ability of the DCBF is weakened by noise amplification caused by the deconvolution processing. The decomposed DCBF algorithm can reduce the amplified noise, but it needs to estimate the target number. In this paper, a new algorithm named cross-spectrum DCBF (C-DCBF) is proposed to avoid the uncertain factor while keeping the advantages of the decomposed DCBF algorithm. Numerical results show that the proposed C-DCBF algorithm has lower background level than the DCBF algorithm (close to the decomposed DCBF algorithm) in low SNR environment, as well as an excellent anti-interference ability.