Adaptive K-Harmonic Means Clustering Algorithm for VANETs

In recent years, vehicular ad-hoc networks (VANETs) have received considerable attentions. As a promising approach to future Intelligent Transportation System (ITS), VANET is capable of providing safety related applications, Internet accessing and various user applications for drivers and passengers. To support efficient data interaction among vehicles, clustering based topology can be applied which groups vehicle nodes in geographical vicinity together, supports direct interaction inside one cluster and inter-cluster data interaction through cluster heads (CHs). Both K-means and K-Harmonic Means (KHM) algorithms are commonly-used clustering algorithms for wireless sensor networks, however, these algorithms cannot be applied to VANETs directly due to the specific characteristics of VANETs. In this paper, we propose an improved KHM algorithms, called Adaptive K-Harmonic Means (AKHM) clustering algorithm for VANETs, which jointly considers the available bandwidth of candidate CHs, and relative distance and velocity between cluster members (CMs) and CHs. To perform the proposed algorithm, the initial values of the number of clusters and the positions of each centroids are chosen and the weighted distance between vehicles and centroids is defined, based on which the objective function can be formulated, and the optimal CHs and the association between CMs and CHs can then be determined. The simulation results demonstrate the efficiency of AKHM algorithm.