Among the key goals of urban policies, the reduction of traffic congestion and pollutant emissions surely tops the list. Although solutions leveraging vehicular communication, such as GLOSA, have been proposed to smooth traffic at regulated intersections, cities normally have a large number of unregulated intersections where queues can build up, worsening emissions caused by the stop-and-go motion of vehicles. The problem is further compounded by the future presence of self-driving cars, where the need to coordinate with other autonomous cars at intersections will be even greater. In this paper, we propose (VTL)-T-3, a Vehicle-to-Vehicle (V2V) Virtual Traffic Light system for infrastructure-less unregulated crossroads. We aim at providing a low-complexity, yet effective, algorithm and communication protocol to let vehicles at an unregulated intersection decide if a virtual traffic light is needed and, in that case, self-organize to establish one. Our results highlight significant improvements compared to both unregulated and traffic light-based intersections. We tested the performance of V3TL in different, realistic scenarios (isolated or consecutive three-and four-way junctions, in single-and multi-lane configuration) and different vehicle generation rates, obtaining improvements in terms of number of passing vehicles per minute, number of stop-and-go maneuvers and scheduling fairness, when compared to unregulated or traffic-light regulated junctions.
展开▼
