Title:Hybrid Near-Field and Far-Field Localization with Multiple Holographic MIMO Surfaces

Abstract:Localization using multiple base stations (BSs) has gained much attention for its advantage in localization accuracy. However, the performance of the multi-BS system suffers from its limited number of antennas. To solve the above issue, we propose to use reconfigurable intelligent surfaces (RIS) serving as antennas. Existing localization methods enabled by multiple RISs mainly focus on the far-field (FF) region of each RIS. As the scale of RIS increases, the near-field (NF) region of each RIS expands, where FF methods struggle to achieve high localization accuracy. In this letter, a hybrid NF and FF localization method aided by multiple RISs is proposed. In such scenarios, achieving user localization and RIS optimization becomes challenging due to the high complexity caused by the exhaustive search through all candidate locations to match the signals. Moreover, the interference from multiple RISs degrades the localization accuracy. To address this challenge, we propose a two-phase localization method that first estimates the relative locations of the user to each RIS and fuses the results to obtain the estimation. This approach reduces the complexity by decreasing the number of candidate locations considered in each step. Also, we introduce a constraint in the RIS optimization problem that limits the sidelobe levels directed towards other RISs, effectively minimizing inter-RIS interference. The effectiveness of the proposed method is verified through simulations.