Traffic Prediction Based on a Local Exchange of Information
Bruno Barbieri, Guillaume Sartoretti, Jean-Luc Falcone, Bastien Chopard and Martin J. Gander
We propose a decentralized method for traffic monitoring, fully distributed over the vehicles. An algorithm is provided, specifying which information should be tracked to reconstruct an instantaneous map of traffic flow. We test the accuracy of our method in a simple cellular automata traffic simulation model, for which the traffic condition can be controlled and analyzed theoretically. We show how local communication parameters affect the method accuracy.