Single Snapshot DOA Estimation Based on Spatial Smoothing MUSIC and CNN

Deep learning (DL) based direction of arrival (DOA) estimation methods have low computational complexity and broad application prospects. However, existing DL-based methods require multiple snapshots measurement data to accurately estimate DOA, which are not suitable for the case of single snapshot. To solve this problem, this paper proposes a single snapshot DOA estimation method based on spatial smoothing MUSIC algorithm and convolutional neural network (CNN). In the proposed method, the single snapshot measurement data is processed by a spatial smoothing algorithm and then input to the CNN for feature extraction. The CNN directly estimates the array spatial spectrum to achieve the rapid DOA estimation. Simulation results show the feasibility and superiority of the proposed method.