Semantic interpretation of visible light communication traffic signals for autonomous driving

Abstract Visible light communication (VLC) has emerged as a promising technology to enable high-speed, secure, and interference-free communication in intelligent transportation systems. This paper presents a novel framework for the semantic interpretation of VLC-based traffic signals to enhance the perception and decision-making capabilities of autonomous vehicles. By embedding semantic information – such as traffic states, warnings, and control commands – into modulated light signals emitted by traffic infrastructure, the proposed system allows autonomous vehicles to receive and interpret critical road information in real time. The framework integrates signal decoding, temporal synchronization, and semantic parsing to extract actionable insights from received optical signals. A deep learning-based classifier is used to map decoded messages to high-level driving decisions under varying ambient lighting and environmental conditions. Extensive simulations demonstrate the system’s robustness, achieving over 95 % accuracy in semantic message recognition and maintaining reliable communication at distances up to 30 m in outdoor conditions. This work highlights the potential of VLC as a dual-purpose medium for both illumination and intelligent traffic signalling, providing a scalable solution for enhancing situational awareness in autonomous driving.