Measurement-based VLC channel characterization for I2V communications in a real urban scenario

TL;DR

This paper presents an experimental study on VLC channel modeling in urban traffic scenarios, demonstrating that data-driven models outperform traditional Lambertian models in accuracy while maintaining similar complexity.

Contribution

It introduces new measurement-based propagation models for VLC in urban environments, improving accuracy over conventional models.

Findings

Proposed models outperform Lambertian in accuracy.

Measurement campaign conducted with real traffic lights and photoreceivers.

Models maintain comparable complexity to traditional approaches.

Abstract

Visible light communication (VLC) is nowadays envisaged as a promising technology to enable new classes of services in intelligent transportation systems ranging, e.g., from assisted driving to autonomous vehicles. The assessment of the performance of VLC for automotive applications requires as a basic step a model of the transmission pattern and propagation of the VLC signal when real traffic-lights and road scenarios are involved. In this paper an experimental measurement campaign has been carried out by using a regular traffic-light as source (red light) and a photoreceiver positioned, statically, at different distances and heights along the road. A linear regression technique is used to come up with different propagation models. The proposed models have been compared, in terms of accuracy and complexity, to the conventional Lambertian model to describe the VLC channel in a real urban scenario. The proposed models provides a significant higher accuracy with comparable complexity.