Self-stabilization
Multi-Agent Reinforcement Leraning for Traffic Light Control
ICML '00 Proceedings of the Seventeenth International Conference on Machine Learning
TrafficView: traffic data dissemination using car-to-car communication
ACM SIGMOBILE Mobile Computing and Communications Review
Progress and challenges in intelligent vehicle area networks
Communications of the ACM
| Hi-index | 0.00 |
Traffic jams remain a challenging open problem inside and in vicinity of big cities around the world. This leads to a loss of "precious" working time during rush hours, and to the waste of huge amount of fuel which is also undesirable in terms of carbon emission. In this context, traffic light constitutes the bottleneck of the traffic flow system. In addition, disregard of traffic light is the major cause of road accidents. In this paper, we propose a novel traffic light system in which the traffic light scheduling and location are broadcasted, using wireless technology, to each (automated) vehicle. Using a virtual or additional physical traffic light, in conjunction with the original physical traffic light, we develop a scheme which increases the capacity of the intersection. The increased performance of the proposed scheme is illustrated analytically and with simulations. We show in both cases an increase of 5 to 20% in the fraction of undelayed vehicles, depending on the scenario, and a decrease of up to 50% in the average waiting time. Moreover, our scheme makes the traffic light system much more safe by eliminating the immediate danger in crossing the "red" traffic light.