Performance Analysis of Direction of Arrival Estimation Based on Deep Learning
Min Chen, Xingpeng Mao, Yi Gong · 2020
In this paper, a new efficient direction of arrival (DOA) estimation approach based on the deep neural networks (DNN) is proposed, in which a nonlinear mapping that relates the outputs of the receiving antennas with its associated DOA is learning by using the DNN-based network. The novel network architecture is divided into two stages, the detection phase and the DOA estimation phase. Additional detection network attached in our structure dramatically reduces the size of the training set. It has been shown that the proposed method not only can achieve reasonably high DOA estimation accuracy, but also can reduce the computational complexity required by traditional superresolution DOA estimation algorithms such as multiple signal classification (MUSIC). The computer simulation results are performed to investigate the generalization and effectiveness of the proposed approach in different scenarios.