A novel beaconing approach based on sender-centric movement predictions for collision avoidance in VANETS

The detection of potential collisions between vehicles is one of the main functionalities of Vehicular Ad-Hoc Networks (VANETs). Vehicles periodically exchange beacon messages by using wireless communication. From the obtained information the receiver calculates the collision probability by predicting the potential movement of every neighbor. Due to the highly dynamic movement of vehicles the accuracy of a prediction is mainly related to the detailed knowledge about the sender. However, as wireless communication suffers from interference due to a saturated communication channel, the context information within a beacon has to be limited. To accomplish this issue while achieving highly accurate predictions, we propose a scalable sender-centric prediction approach based on the available sender's context information.