Direction-of-arrival estimation method for coherent signals with a small number of acoustic vector arrays based on knowledge distillation

Abstract The estimation of multi-target direction of arrival (DOA) is an important research topic in array signal processing. Traditional algorithms typically rely on the correct representation of the signal and noise subspaces, resulting in relatively low computational accuracy and efficiency. The correlation of the signals also reduces their applicability. Therefore, this paper proposes a DOA estimation method for coherent signals based on a vector hydrophone array. This method uses the knowledge distillation approach and can estimate the DOA of coherent signals by learning only a small amount of data. Simulation results show that compared with traditional methods, the accuracy and root mean square error are improved. The research findings provide a new solution for estimating the DOA of coherent signals.