Evaluation of Intersection Control Strategies Supported by Communication Infrastructure in Presence of Selfish Vehicles

Nowadays, the need of mobility is increasing and new solutions are required to smooth traffic flow in transportation networks. In the context of Intersection Control (IC), traffic can be improved by adjusting the traffic signal controller duration. This adjustment can be made with the support of data achieved via sensors installed in the infrastructure, using images captured by cameras, or with communication support using a Vehicular Ad-Hoc Network (VANET). In the latter case, vehicles implement a Virtual Traffic Light (VTL) by exchanging messages and agreeing in a strategy to organize a crossing order. However, the use of communication gives opportunity for attacks. A selfish vehicle may lie in order to be able to go faster through the intersection. Thus, in this article, we evaluate different strategies to deal with control of traffic intersection, using VANET support. Moreover, we evaluate the behavior of different IC strategies in the presence of selfish vehicles, using as metrics waiting time, throughput and fairness. Experiments were conducted using a traffic simulator. Regarding the simulation results, we found that, despite selfish nodes, platoon-based IC strategies are still the best solution for increasing throughput, though strategies based on individual vehicle behaviour achieved short waiting time and depicted more fairness.