Using road IDs to enhance clustering in vehicular ad hoc networks

Vehicular ad hoc networks (VANETs) where vehicles act as mobile nodes is an instance of Mobile Ad hoc NET-works (MANETs), which are essentially developed for intelligent transportation systems. A challenging problem when designing communication protocols in VANETs is coping with high vehicle mobility, which causes frequent changes in the network topology and leads to frequent breaks in communication. The clustering technique is being developed to reduce the impact of mobility between neighboring vehicles. In this paper, we propose an Adaptive Weighted Cluster Protocol for VANETs, which is a road map dependent and uses road IDs and movement direction in order to make the clusters structure as stable as possible. The experimental results reveal that AWCP outperforms four other most commonly used clustering protocols in terms of control packet overhead, the packet delivery ratio, and the average cluster lifetime, which are the most usual metrics used for comparing performance.

I. Introduction

Research on Vehicular Ad hoc Networks (VANETs) has attracted increasing interest over recent years due to its capability to improve road safety by using Vehicle To Vehicle (V2V) and/or Vehicle To Infrastructure (V2I) communications [1]. It can also be used to improve traffic management conditions and to provide on-board infotainment such as Internet access, video streaming, etc. VANETs are characterized by the self-organization of the nodes, where their nodes can be vehicles, roadside units or sensors. Due to the varying vehicular densities caused by high vehicle mobility, supporting network connection requires a high communication overhead for exchanging and updating topology information. For instance, in a fully distributed VANET, each vehicle is required to maintain its own connectivity to its one hop neighboring vehicles. Without using expensive components such as central points, establishing a hierarchical clustering structure within the network can reduce the relative mobility between neighboring vehicles, thereby reducing communication overhead [2]. Clustering allows the formation of organized groups and is used to coordinate channel access, simplify routing, and security. In VANETs, each vehicle is equipped with a digital road map and a positioning system, e.g. Garmin Nuvi 50 GPS that allows it to obtain the time, its speed and position [3] and the ID of the road on which it is traveling. Therefore a clustering protocol for VANET network can be designed using a MAP digital and GPS based approach. However, the main challenge for clustering protocols in VANETs is ensuring topology stability by taking into account many mobility parameters. This should be performed with a minimum overhead. In this paper, we propose a multi-metrics based Adaptive Weighted Clustering Protocol (AWCP) that takes advantage of the geographic information of vehicles. The main contributions of the paper are listed below: •

We identify certain essential features that the clustering protocols must satisfy to build stable clusters in VANETs.

•

We propose a Map and GPS based clustering protocol using the WCA algorithm [4] for VANET in which a vehicle only considers neighbors moving on the same road and in the same direction. To the best of our knowledge, this is the first attempt to design a clustering protocol by taking advantage ID of the road on which the vehicles are traveling.

•

We evaluate the proposed protocol on realistic VANET mobility scenarios taken from the metropolitan area of San Jose (California) and we validate its performance by comparing it with other clustering protocols in the literature.