A singular value decomposition-based positioning algorithm for indoor visible light positioning system

In this work, a visible light positioning (VLP) system using an image sensor (a smartphone camera) as the receiver is considered. We focus our research on the system acquisition time and success rate, both of which are largely determined by the employed positioning algorithm. To speed up the positioning process and enhance its robustness, we derive a close-form expression to determine the receiver's position and orientation using the singular value decomposition (SVD) technique, and a new SVD-based positioning algorithm is proposed accordingly. Simulation results prove that the proposed SVD-based positioning algorithm is 50-80 times faster than the conventional iterative Levenberg-Marquardt (LM) based algorithm and avoids possible failures caused by bad initial guesses.