Accurate Cluster Head Selection Technique for Software Defined Network in 5G VANET

Abstract

The vehicular Ad Hoc Network is a very dynamic network with uncertain presence of vehicles on the road and data transmission between vehicles and road-side units. Centralizing and distribution management of this network is stimulating. The management of the vehicles’ network starts with the Cluster Head selection and regulating its stability to be Cluster Head, for the maximum time possible. The clusters thus formed are further updated by Software-Defined Network. In the update of the cluster head, the Software-Defined Network selects that vehicle as Cluster Head, which has a higher probability of selection as a Primary User for available spectrum usage. It makes the vehicle more trustworthy as no malfunctioned vehicle is selected as a Primary User. In this paper, a 2-level approach for Primary User detection in the network has been proposed. The energy of the received signal is sensed and Primary User is detected by fuzzy logic and this decision is further used as a threshold for final Primary User selection. Other parameters like network connectivity level, lane weight, average velocity, and average distance along with trustworthiness by cognitive radio sensing are used to elect the Cluster Head by the SDN. The proposed approach shows higher stability in cluster head selection as compared to state-of-the-art schemes. Primary Users are also detected with zero percent probability of false detection in the proposed method.