A Robust Clustering Scheme for Vehicular Communication Networks

TL;DR

This paper introduces a robust clustering scheme for UAV-assisted vehicular networks that improves stability and reliability by considering vehicle behavior and backup cluster heads, effectively managing high mobility and resource constraints.

Contribution

The paper proposes a novel clustering scheme with a backup list for UAV-assisted vehicular networks, enhancing stability and performance under high mobility conditions.

Findings

Outperforms benchmark schemes in clustering stability and reliability.

Performance remains stable with increasing vehicle numbers.

Effective in dense vehicular networks with resource constraints.

Abstract

Clustering, as a technique for grouping nodes in geographical proximity together, in vehicular communication networks, is a key technique to enhance network robustness and scalability despite challenges such as mobility and routing. This paper presents a robust clustering scheme based on cluster head backup list algorithm for unmanned aerial vehicles (UAVs)-assisted vehicular communication network, where multiple UAVs act as communication base stations for a vehicular network. To tackle the high mobility issues in vehicular communications, instead of allowing direct communication between all vehicles to the UAV, clustering methods will potentially be efficient in overcoming delay limitations, excessive power consumption and resource issues. Using the clustering technique, neighboring vehicles are grouped into clusters with a specific vehicle selected as the cluster head (CH) in each cluster. The selected CH connects directly to the UAV through an infrastructure-to-vehicle (I2V) link, subsequently establishing vehicle-to-vehicle (V2V) communications with vehicles in the same cluster. To increase cluster connectivity period, the proposed clustering scheme is developed based on considering the vehicle behavior for efficient selection of CHs and providing a CH backup list to maintain the stability of the cluster structure. Numerical evaluations show that the proposed system outperforms benchmark schemes in terms of clustering stability and reliability. It is also shown that the performance of the proposed scheme is not much affected by the increase in the number of vehicles. This indicates that the proposed scheme can be efficient in dense vehicular networks where resource constraints pose significant challenges.