A Cluster-Based Beaconing Approach in VANETs: Near Optimal Topology Via Proximity Information

A key component for safety applications in Vehicular ad-hoc network (VANET) is the use of periodic beacon messages which provide vehicles with a real-time vehicle proximity map of their surroundings. Based on this map, safety applications can be used for accident prevention by informing drivers about evolving hazardous situations. In order to allow synchronized and cooperative reactions, the target of this work is to design a beacon dissemination process that provides a real-time, broad and coordinated map under the challenging VANET conditions. In order to obtain the desired map, we consider an aggregation-dissemination based scheme for a beacon dissemination process that based on top of a cluster-based topology. To this end, we propose the Distributed Construct Underlying Topology (D-CUT) algorithm tailed specifically to provide an optimized topology for such beacon dissemination process. To deal with the heavy load of beacon messages required for an accurate and broad map, we propose a topology that allows the execution of extensive but reliable spatial bandwidth reuse. Our D-CUT algorithm exploits the real-time and coordinated map for constructing an adaptive and robust topology to deal with the dynamic nature of the VANET environment. We present theoretically provable bounds demonstrating the ability of the algorithm to deal with the dynamic nature of the VANET environment supported by simulation results.