Traffic Management Systems become an important challenge for large cities due to the constant growth of vehicles. As the road mesh does not increase as well as the number of vehicles in the streets, technological solutions for the traffic congestion rise as alternative and easy-to-use applications. This work presents the ON-DEMAND: An adaptive and Distributed Traffic Management System using VANETS. The proposed solution is based on V2V communication and the local view of traffic congestion. During its displacement in a road, the vehicle monitors its traveled distance and the expected one considering a free-flow traffic condition. The difference between these measurements is used to classify a contention factor, i.e., the vehicle perception on the road traffic condition. Each vehicle uses the contention factor to classify the overall congestion level and this information is proactively disseminated to its vicinity considering an adaptive approach. In the case a vehicle does not have the necessary traffic information to estimate alternative routes, it executes a reactive traffic information knowledge discovery. The proposed solution is compared with three literature solutions, named DIVERT, PANDORA and s-NRR. Our results showed that ON-DEMAND presents better results regarding network and traffic congestion metrics.

由于车辆的持续增长，交通管理系统成为大城市的重要挑战。由于道路网的增加以及街道上车辆的增加，因此交通拥堵的技术解决方案作为替代且易于使用的应用而兴起。这项工作提出了ON-需求：使用VANETS的自适应分布式交通管理系统。所提出的解决方案基于V2V通信和交通拥塞的本地视图。在道路上移动时，车辆会根据自由流动的交通状况监控其行驶距离和预期行驶距离。这些测量值之间的差异用于对竞争因素进行分类，即对道路交通状况的车辆感知。每辆车都使用竞争因素对总体拥堵程度进行分类，并考虑采用自适应方法，将该信息主动传播到附近。如果车辆没有必要的交通信息来估计替代路线，则它会执行反应性交通信息知识发现。将提出的解决方案与三种文献解决方案（称为DIVERT，PANDORA和s-NRR）进行了比较。我们的结果表明ON的按需提出了关于网络和交通拥堵等指标较好的效果。