Traffic congestion manager, a cost-effective approach

Abstract

Traffic congestion is a worldwide problem which challenges both scientific community and governments. This paper proposes a novel cost-effective approach which aims to predict congestion levels of road segments and thus helps preventing high traffic congestions. Unlike other approaches based on cooperative vehicular communication, this work adopts a cooperation-without-communication approach inspired by social insects. In fact, vehicles cooperate and share their experiences through RSU (Road Side Unit) controllers placed in the traffic lights of the studied area. Each vehicle is equipped with a navigation device which notifies the driver about congestion levels of road segments. Hence, vehicles’ distribution over the road network tends towards homogeneity. An evolutionary algorithm which optimizes traffic reports exchange between vehicles and RSU controllers is described.