Joint Localization of UEs, Scatterers, and Obstacles with Spatial Non-Stationary ELAA

This paper addresses the joint localization of user equipments (UEs), scatterers, and obstacles (JLUSO) in near-field spatial non-stationary (SnS) environments using extremely large aperture arrays. After modeling the near-field SnS scenario, we introduce a three-stage JLUSO framework to solve the sensing problem using low-complexity methods. First, we define the channel time-reversal spatial spectrum (CTRSS) and use the CTRSS-based grid search (CTRSS-GS) method to locate the UEs and scatterers. Second, we develop the CTRSS-based contribution decision on the sliding window (CTRSS-CDSW) method to adaptively refine visible region (VR) identification. Finally, we propose a ray-tracing-based geometric localization (RTGL) method with closed-form solutions for obstacle localization. Simulations validate the effectiveness of the JLUSO framework. CTRSS-GS achieves decimeter-level localization accuracy within a region of tens of meters, while CTRSS-CDSW reduces the VR identification error to less than 1%. The RTGL method achieves obstacle localization errors comparable to obstacle sizes. These contributions offer a comprehensive solution for sensing in nearfield SnS scenarios, advancing 6G sensing capabilities.