A mobility metrics based dynamic clustering algorithm for VANETs

Clustering algorithm is a key technology in Vehicular Ad hoc network (VANET). However, due to the extremely high mobility and road-constrained features in VANETs, the existing clustering algorithms for MANET do not perform well in VANET. Based on the mobility metrics of the vehicles, a Dynamic Clustering Algorithm (DCA) for VANET is proposed in this paper in order to form more stable clusters, improve cluster lifetime and reduce the clustering reaffiliation times even in a highly dynamic environment. The cluster structure is determined by the spatial dependence, which is a description the mobility similarity relationship between different nodes. The simulation is performed with comparative studies using NS-2 and VanetMobiSim. Simulation results show that the performance of the DCA algorithm is superior to other widely used clustering algorithms, the Lowest-ID and Max-Degree clustering algorithm, in terms of cluster lifetime and reaffiliation times.