Two-Dimension DOA Estimation Base on HOSVD under the Fully-Polarized RIS Architecture

This study addresses the problem of two-dimensional (2D) direction of arrival (DOA) estimation in Non-Line-of-Sight (NLOS) propagation scenarios. It specifically focuses on the reconfigurable intelligent surface (RIS) architecture, which is equipped with arbitrarily placed Electromagnetic Vector Sensors (EMVS). A novel method is introduced to utilize the multidimensional characterization of received signals. This is achieved by applying High-Order Singular Value Decomposition (HOSVD). It shows the outstanding superiority especially in low signal-to-noise ratio (SNR) and limited samples.