Zusammenfassung
Vehicle-to-everything (V2X) interconnects participants in vehicular environments to exchange information. This enables a broad range of new opportunities. For instance, crowdsourced information from vehicles can be used as input for self-learning systems. In this paper, we propose iTLM-Q based on our previous work iTLM to optimize traffic light management in a privacy-friendly manner. We aim to ...
Zusammenfassung
Vehicle-to-everything (V2X) interconnects participants in vehicular environments to exchange information. This enables a broad range of new opportunities. For instance, crowdsourced information from vehicles can be used as input for self-learning systems. In this paper, we propose iTLM-Q based on our previous work iTLM to optimize traffic light management in a privacy-friendly manner. We aim to reduce the overall waiting time and contribute to a smoother traffic flow and travel experience. iTLM-Q uses Q-learning and is constraint-based in such a way that no manual traffic light cycles need to be defined in advance, hence, being able to always find an optimal solution. Our simulation-based on real-world data shows that it can quickly adapt to changing traffic situations and vastly decrease waiting time at traffic lights eventually reducing CO2 emissions. A privacy analysis shows that our approach provides a significant level of k-anonymity even in low traffic scenarios.
Altmetric