Abstract: The visible light communication of the Internet of vehicles is not only affected by atmospheric turbulence, but also disturbed by background lights such as traffic lights, street lights, front and rear lights of vehicles and high brightness billboards. The conventional light intensity flicker model can not accurately describe the channel characteristics of the visible light communication system of the Internet of vehicles, a night background light noise model of vehicle network visible light communication system expressed by double Gaussian function superposition is proposed. Through the measured data in different regions and different weather, the model is numerically fitted to verify the effectiveness of the model. The results show that there is no obvious difference between the probability density distribution in rainy days and that in sunny days, and the intensity fluctuation of road light is less than that of other light sources. The model has certain research significance for the establishment of visible light channel model and noise removal of vehicle networking.