VRDD: Vehicular Relevance based Data Dissemination in Vehicular Ad-hoc Network

Safety message (SM) dissemination in vehicular ad hoc network needs to be fast as it allows vehicles to become alert and avoid adverse situations. An SM usually gets disseminated within a topology by multi-hop broadcast technique. Existing protocols have proposed various controlled flooding mechanisms to address the issues of broadcast storm and network partition. However, no work has been done to take care the direction of message dissemination. In a real scenario, all vehicles within an area of interest (AOI) are not equally affected by an emergency incident. The vehicles of the road where the incident happened, or the vehicles of roads which are connected to incident location by nearby junctions, are most affected and hence are more relevant to receive the SM as fast as possible. Hence, this study proposes a novel method to predict the relevance of vehicles using a distributed algorithm and heuristically tries to disseminate SM toward the relevant vehicles. The authors have compared their work with three state of-the-art algorithms and found that vehicular relevance-based data dissemination is compatible with the existing algorithms on the metrics of packet loss and dissemination delay, but clearly leading in the metric of relevant dissemination.