Work in progress paper: Utility based clustering algorithm for VANET

Clustering is expected to be one of the most efficient and promising solutions for data transmission in Vehicular Ad hoc NETworks (VANET). The particular characteristics of VANET, including highly dynamic network topology, unstable transmission channel, and various service requirements, pose challenges and difficulties on clustering algorithm design. In this paper, a clustering algorithm based on utility function is proposed for VANET. The utility function of vehicles is formulated which jointly considers current state of vehicles characterized by the vehicle degree, relative velocity and position between adjacent vehicles, and long-term transmission performance of vehicles, defined as credit history function, which is characterized by the average available bandwidth, queue length and the duration for being a cluster head (CH) in a past period. The process of cluster forming is presented, and the vehicles with large utility function are selected as CHs successively. Numerical simulations demonstrate that comparing to previously proposed algorithms, the proposed algorithm offers better performance in terms of packet delivery rate, average transmission delay and total throughput.