Content Aided Clustering and Cluster Head Selection Algorithms in Vehicular Networks

Abstract

Relying on clustering and the cluster head selection algorithms, vehicle-to-vehicle (V2V) and vehicle-to- infrastructure (V2I) based vehicular ad hoc networks (VANETs) play a critical role in intelligent transport system (ITS). However, the existing clustering and cluster head selection algorithms did not consider the influence of the vehicles' communication contents and their correlations. Specifically, the power-law characteristics of vehicle content demands are beneficial in terms both of achieving efficient clustering algorithm and selecting optimal cluster heads. In order to simulate the real vehicular communication scenarios, we commence with the mobility model design in this paper. Moreover, a novel clustering algorithm relying on content demands is proposed, which attracts vehicles to adopt V2V network through price advantage. Furthermore, based on the Fermi rule, i.e., one of the stochastic evolutionary strategies in complex networks, and evolution game, our cluster head selection algorithm is capable of representing more realistic vehicles' features, including selfishness, fairness and bounded rationality. Finally, the effectiveness and feasibility of our proposed algorithms are verified.