OIRS-assisted NLoS visible light positioning: An improved GWO dual-feature fusion approach for SISO systems.

Abstract

This study addresses the challenge of achieving high-precision indoor positioning in non-line-of-sight (NLoS) environments through the development of an innovative visible light positioning (VLP) system that utilizes optical intelligent reflecting surfaces (OIRS). Unlike current hybrid methodologies that combine both line-of-sight (LoS) and NLoS techniques tailored for Internet of Things (IoT) environments, our novel single-LED architecture relies solely on signals reflected by an OIRS to facilitate accurate positioning in intricate indoor settings where direct light paths are often obstructed. This system employs a two-stage maximum likelihood estimation framework that effectively integrates received signal strength (RSS) and timeof-arrival (ToA) characteristics, thereby addressing the shortcomings of traditional single-feature methods and ensuring reliable performance in densely populated IoT scenarios. To tackle the non-convex optimization problem, we propose an improved grey wolf optimization (IGWO) algorithm, which exhibits superior positioning accuracy and convergence properties when compared to particle swarm optimization and genetic algorithms. Simulation results substantiate the framework’s efficacy, demonstrating improved positioning accuracy. The proposed system presents a cost-effective solution for complex indoor environments where direct light paths are frequently obstructed, thereby advancing the practical application of VLP technologies.