Parallel swarm optimization based cluster for emergency message dissemination in VANETs

In intelligent transportation system, emergency message has to be disseminated rapidly and extensively in order to inform as many vehicles in limited time. Hence, the broadcast storm problem is prone to occur in vehicular ad hoc networks. In the past, several approaches have been proposed to solve the problem. They are counter-, distance-, location-, cluster-, and probabilistic-based schemes. In this paper, we analyze existing cluster based protocols and present a novel scheme that is designed to form stable clusters, reduce the overhead of cluster maintenance during emergency message dissemination. The proposed clustering algorithm uses particle swarm optimization to assigns weight values for different factors during the cluster head election and cluster formation. It ensures the dissemination of the emergency message onto every part of the road, and when there is no relay vehicle, the emergency message will be reactivated. We use the GrooveNet simulator to demonstrate that the proposed scheme has a low collision probability, low overhead, more stable cluster, and a short end-to-end latency.