Development Of Visible Light Communication Platform For Vehicular Networking And Research On Vehicle LED Tracking Technology

With the development of the automotive industry and wireless communication technology,vehicular networking is important for improving road safety,optimizing transportation efficiency,and building novel smart city transportation system.Communication technology is the cornerstone of vehicular networking.Visible light communication(VLC)for vehicular networking transmits data through visible light whose intensity changes at a high speed,which achieves data transmission and illumination at the same time.VLC is license free and could avoid the interference in the traditional wireless communication,making the data transmission more reliable and the delay shorter.VLC can also reduce costs.Therefore,VLC is a promising technology for vehicular networking.The research work of this thesis is supported by Huawei Innovation Research Project(No.20150101)and Beijing University of Posts and Telecommunications Science and Technology Achievements Transformation Seed Project.At present,there are some problems in VLC for vehicular networking,such as lacking anti-noise robustness,limited communication distance,insufficient support for mobility,lacking experimental system.Therefore,a VLC platform for vehicular networking is developed to verify and analyze the performance through field experiments.The adaptive vehicle light emitting diode(LED)area tracking algorithm based on back propagation(BP)neural network is proposed,which effectively improves the support of VLC for mobility.The main research contents of this thesis are as follows:1.VLC platform for vehicular networking is designed and implemented.The hardware selection and circuit implementation are completed.At the transmitter,vehicle data is generated and modulated to LED after being encoded.In order to achieve multi-vehicle data reception and overcome the issue incurred by unstable frame rate of the image sensor,receiver modules including image sensor control,image region segmentation and data recovery are designed and implemented.The platform achieves real-time reliable data communication between multiple vehicles,the Bit Error Rate(BER)is less than 10-3 when vehicles are relatively stationary within 125 meters.The mobility support problem of VLC for vehicular networking is analyzed through experimental results.2.The adaptive vehicle LED area tracking algorithm based on BP neural network is designed and implemented and its effectiveness is verified by field experiments.The improved frame difference method is designed to detect pixels where LEDs are located.The inter-area distance matrix is designed to achieve tracking of the vehicle LED areas in the received image.The adaptive frame difference threshold based on neural network is designed and implemented,which effectively improves the reliability of the tracking algorithm under strong interference lights and unstable image exposure.The experimental results show that the algorithm can effectively improve the mobility support of VLC for vehicular networking,the communication BER is lower than 10-3 when the relative velocity between vehicles is less than 67.5 km/h.