An Opportunistic Traffic Management System for Vehicular Networks

Road congestion is a serious problem both in terms of costs and time wasted. State of the art traffic management techniques such as traffic light synchronization, dynamic message signs and traffic management decisions are based on vehicle counts at streets and intersections. Vehicular networks could radically transform traffic systems. A peer to peer vehicular network can enable cars to prevent accidents and manage congestion. The use of an opportunistic vehicular network is both efficient and cost effective: cars are in-traffic sensors that provide real-time delay estimates on a fine grained scale of both space and time. A mobility management system should be distributed, act locally, be capable of sharing traffic information fairly among cars and take advantage of typical urban mobility patterns. In this paper we propose an opportunistic vehicular network system in which vehicles receive local information about traffic towards their destination and dynamically update their routes given this information. We show the performance evaluation results in terms of efficiency and effectiveness on a realistic simulator that models both mobility and wireless network connectivity. We also indicate some results of a real testbed evaluation.