Abstract
Due to the increasing usage of Internet-of-things (IoT) devices, their positioning is an essential need in many real-world applications. Along this line, visible light communication (VLC) is a principal approach due to its advantages of high accuracy, low computation cost, ubiquitous lighting infrastructure, and being free of RF interference. Still, there is a growing need for improving the positioning accuracy and speed. This paper proposes and prototypes a VLC-based positioning solution using retroreflectors mounted on the IoT devices of interest. The proposed method exploits the retro-reflected power received at several PDs for rendering the coordinates of reflectors in the real world. Specifically, the relative relationships between the magnitudes of reflected light ray power are used to figure out the x-y coordinates such that the centroid of power magnitudes is regarded as the x-y coordinate. Besides, the variance of those magnitudes is used to infer the height of the retroflector (the z-coordinate). The proposed solution excels in its simplicity and fast computation and alleviates the need for sensory devices and active operation. This is validated over executive experiments conducted on a prototype in a real-world environment.