Falcon: Fused Application of Light Based Positioning Coupled With Onboard Network Localization

Abstract

Indoor localization based on visible light and visible light communication has become a viable alternative to radio frequency wireless-based techniques. Modern visible light position (VLP) systems have been able to attain sub-decimeter level accuracy within standard room environments. However, a major limitation is their reliance on line-of-sight visibility between the tracked object and the lighting infrastructure. This paper introduces fused application of light-based positioning coupled with onboard network localization (Falcon), a VLP system, which incorporates convolutional neural network-based wireless localization to remove this limitation. This system has been tested in real-life scenarios that cause traditional VLP systems to lose accuracy. In a hallway with luminaires along one axis, the Falcon managed to attain position estimates with a mean error of 0.09 m. In a large room where only a few luminaires were visible or the receiver was completely occluded, the mean error was 0.12 m. With the luminaires switched off, the Falcon managed to correctly classify the target 99.59% of the time to within a 0.9-m2 cell and with 100% accuracy within a 1.6-m2 cell in the room and hallway, respectively.

Keywords

• Indoor positioning systems (IPS)
• indoor localization
• visible light communication (VLC)
• visible light positioning (VLP)
• zigbee localization
• convolutional neural network (CNN)