Toward Strongley Connected Clustering Structure in Vehicular Ad Hoc Networks

Abstract

Clustering in Vehicular Ad-hoc Networks (VANET) is one of the control schemes used to organize media access and make VANET global topology less dynamic. Most VANET clustering algorithms are derived from Mobile Ad-hoc Networks (MANET). But, these algorithms model clustering schemes considering only position and direction of vehicles located in a geographic proximity regardless of the degree of the speed difference among them. Since the existence of VANET nodes in the same geographic location doesn't mean that they exhibit similar mobility pattern, we believe that VANET clustering models need to be redefined such that, their characterization is identified by considering the full status elements; speed, location, and direction instead of considering location and direction only. In this paper, we propose a new approach for grouping vehicles, showing similar mobility patterns, in a cluster and at the same time try to minimize the total number of created clusters. A new multi-metric Cluster Head (CH) election technique has also been developed. This technique can be used by vehicles to determine their suitability on the fly to lead the cluster. The simulation is conducted to evaluate the proposed technique and compare it with the most common methods. The simulation results show that our technique increases the cluster stability by about 50% compared to that of the most commonly used techniques.