Distributed road traffic congestion quantification using cooperative VANETs

Abstract

The well-known traffic congestion problem in urban environments has negative impact on many areas including economy, environment, health and lifestyle. Recently, a number of solutions based on vehicle-to-vehicle communications were proposed for traffic congestion detection and management. In this paper we present an algorithm designed to enable each vehicle in the network to detect and quantify the level of traffic congestion in completely distributed way, independent of any supporting infrastructure and additional information such as traffic data from local authorities. Based on observations of traffic congestion by every vehicle, and by adapting the broadcast interval, it enables dissemination of the traffic information to other vehicles. The algorithm also makes every vehicle aware about the congestion level on the streets that are spatially separated from their current location by several streets. Its robustness keeps the vehicle's overall knowledge about congestion consistent, despite the short-term changes in vehicle's motion. Since the quantification of congestion is based on per-vehicle basis, the algorithm is able to operate even when only 10% of vehicles in the network are VANET enabled. Data aggregation and adaptive broadcasting are used to ensure that vehicles do not send redundant information about the traffic congestion. The simulations are conducted in Veins framework based on OMNeT++ network simulator and SUMO vehicular mobility simulator.