DoA Estimation based on Deep Learning in Low SNR

A method for estimating the direction-of-arrival (DoA) in low SNR is proposed using a cascaded deep learning. By mapping the covariance matrix to the signal direction angle, this approach is implemented. There are primarily two steps in the processing process. Firstly, in order to categorize the direction-of-angle estimation into subregions, a convolutional neural network (CNN) is developed to classify signals utilizing multi-label classification model. Then, the direction of angle is estimated using a collection of simultaneous regression models powered by CNN. According to the simulation results, the proposed strategy greatly lowers angle estimation error when compared to the conventional approaches under low signal-to-noise ratio (SNR).