Angle-of-Arrival Estimation via DAE-enhanced soft-weighted Clustering

Abstract

This paper proposes a preprocessing framework that combines a denoising autoencoder (DAE) with soft-weighted density-based spatial clustering of applications with noise (DB-SCAN) to enhance the robustness of convolutional neural network (CNN)-based angle-of-arrival (AoA) estimation in low-SNR environments. By creating a refined latent space and applying reliability-based weights, this approach improves the quality of input data. A comparative analysis is conducted by training CNN models with and without the proposed framework. Experimental results demonstrate that our method achieves more accurate AoA estimation across various SNR conditions.