Optical IRS Assisted-Visible Light Positioning in Indoor Non-LOS IoV Scenarios

The demand for high-precision localization services has surged significantly due to the rise of intelligent transportation and autonomous driving in large-scale indoor factories. To mitigate the challenge of reduced positioning accuracy resulting from line-of-sight (LOS) occlusion in indoor visible light positioning (VLP) technology, this paper proposes a positioning strategy that leverages optical intelligent reflecting surfaces (IRSs) within indoor Internet of Vehicles (IoVs) environments. This scheme utilizes the principle of range positioning based on the time difference of arrival (TDOA). A weighted least squares (WLS) method is initially derived as a benchmark positioning approach based on TDOA. Additionally, two high-accuracy positioning methods, namely, the Chan method and Taylor series expansion method, are proposed for different scenarios. The Chan method provides a closed-form solution suitable for low-computation cases, while the Taylor series expansion method can be iteratively exploited in complex situations. The detailed procedures of these three positioning approaches are also presented. Furthermore, theoretical analyses of the computational complexity of the WLS method, Chan method, and Taylor series expansion method are provided. Additionally, the positioning performance of the Cramér-Rao lower bound (CRLB) is analyzed and derived for the considered system model. The simulation results illustrate that the system model and positioning methods outlined can asymptotically achieve the derived CRLB, thereby validating the efficacy of the proposed positioning scheme and methods. These advancements hold significant potential for industrial automation, logistics operations, and safety-critical autonomous guided vehicles (AGVs) in smart factories, where robust centimeter-level positioning is essential for collision avoidance and task coordination under dynamic occlusion conditions.