Ph.D. Forum: Toward Adaptive Visible Light Positioning

Visible light positioning (VLP) has shown great promise for high-precision indoor localization. This research aims to enhance VLP accuracy by leveraging spectral information, overcoming dynamic lighting challenges, and reducing the effort of fingerprint collection. Initially, Iris improves localization precision using spectral information and addresses dynamic lighting effects through a background subtraction module. Building upon this, LiDARSpectra employs synthetic mapping with low-cost LiDARs to automate fingerprint collection. Lastly, ongoing research with Large Language Model seeks to further improve accuracy and generalization across diverse indoor environments. Our work provides an adaptable framework for dynamic light-based IoT applications.