Single LED visible light positioning scheme based on intelligent reflecting surfaces

Visible light positioning (VLP) systems have gained attention for their ability to operate without additional infrastructure. However, most positioning schemes rely on multiple LEDs, which presents a considerable challenge in scenarios where only a single-light-source is available. To address this issue, this work proposes using intelligent reflective surfaces (IRS) as virtual light sources for VLP. To our knowledge, a novel positioning scheme, referred to as the difference of received signal strength, is proposed to mitigate spatial interference from multiple IRS and accurately estimate their power, enabling positioning even with a blocked line-of-sight link. Unlike maximum likelihood estimators, this scheme eliminates the need for a precise noise model. Meanwhile, the Cramér-Rao lower bound is derived to analyze the impact of IRS parameters such as rotation angle, layout, and quantity, on positioning accuracy. Theoretical analysis indicates that IRS orientation and layout greatly influence positioning accuracy, with layout offering more significant improvement than IRS quantity. Simulation results demonstrate that the scheme achieves a positioning accuracy of 0.056 m with 16 IRS units, with further improvements through additional optimization. Moreover, the results align with theoretical analysis, with an optimized layout reducing the positioning error to 0.028 m, highlighting its pivotal role in enhancing IRS performance.