Trajectory Prediction of Target Light Source for Dynamic Visible Light Communication Systems with A Narrow Field of View

Weibin Jiang,Xianqing Jin,Yingwen Zhang,Meiyu Jin,Chen Gong,Zhengyuan Xu

DOI: https://doi.org/10.1109/iccworkshops49005.2020.9145246

2020-01-01

Abstract:For the vehicular visible light communication (VLC) system, one of key challenges is beam alignment between a light-emitting diode (LED)-based transmitter and a lens-based receiver with a narrow field of view (FoV) for high optical gain. Received optical intensity varies significantly at a mobile receiver due to the delay in beam detection and alignment. In order to maximize the received optical intensity for high signal-to-noise ratio, in this paper, a real-time light source tracking system with a trajectory prediction function based on the Kalman filter is experimentally designed and demonstrated for dynamic VLC applications. A two-axis gimbal is used to adjust the receiver attitude with the location of a wide FoV transmitter, which is detected by a high-speed camera. As the frame rate of practical cameras may not be equal to the frequency of the beacon signal, there is a variable delay in location prediction and beam alignment. An adaptive control based on the S-shaped speed planning model is proposed to improve the alignment performance. The effectiveness of the scheme is verified in both simulation and experiment. Experimental results show that with the proposed scheme the root mean square error (RMSE) in alignment deviation is improved by >50%.