An Eigenvalue-Based Generalized MVDR Method for DOA Estimation of Underwater Targets Without Prior Knowledge of Source Number

The actual underwater acoustic field environment is subject to various interferences, resulting in high background noise, thereby complicating underwater target detection. Some traditional subspace direction-of-arrival (DOA) estimation algorithms often require prior knowledge of the number of sources, leading to performance degradation if the number of sources is estimated inaccurately. With advancements in sensor technology, scholars have developed acoustic vector sensors (AVS). However, some traditional algorithms applied directly to AVS result in degraded performance of the algorithm. To address these challenges, we propose a DOA estimation algorithm that is independent of the number of sources and is applicable to AVS. This is achieved by constructing the generalized minimum variance distortionless response (G-MVDR) model based on AVS and exploring the eigenvalue ordering principle under the G-MVDR framework. It provides high-resolution DOA results independent of the number of sources and can be flexibly applied to AVS arrays. The proposed algorithm is analyzed with respect to eigenvalue ranking, and the optimal threshold is determined. Finally, the accuracy and effectiveness of the proposed algorithm are validated through simulations and experiments.