The failure detector is one of the fundamental components for maintaining high availability of Vehicular Ad-hoc Networks (VANETs). However, the dynamic nature of VANETs caused by the high mobility of vehicles and communication link failures has a serious impact on the performance of failure detection. Therefore, it is very meaningful to design a suitable failure detector that can deal with the dynamic nature of VANETs well. In this paper, we propose a hierarchical failure detector based on the architecture of VANETs. This failure detector can adapt to the dynamic network conditions and meet the different Quality of Service (QoS) requirements of multiple applications in VANETs. Different from existing failure detectors, we propose a failure detector that employs a detection-result sharing mechanism and groups the nodes according to the architecture of VANETs. We evaluate our proposed failure detector by using NS2 and GT-ITM to simulate the work environment of VANETs. The experimental result shows that our proposed failure detector can improve the detection time by at most 45% and the detection accuracy by at most 25% under similar detection overhead.

Introduction

Vehicular Ad-hoc Networks (VANETs) are dynamic, nonstructured, self-organizing networks with asynchronous and distributed characteristics, and the nodes (vehicles) move at high speeds compared to other mobile ad-hoc networks (MANETs) [1]–[۳]. The main purpose of VANETs is to provide a medium for intervehicular communication allowing for intervehicle (V2V) and vehicle to roadside infrastructure (V2R) data exchange, with multiple applications for Intelligent Transportation Systems (ITS). One of the important features is the high mobility of vehicles in VANETs; vehicles can suddenly quit or enter the network, and communication links among vehicles may suffer from signal degradation due to obstacles, changes in vehicle densities, etc. [4]–[۶]. For the ITS applications running on VANETs, VANETs must be fault-tolerant to mitigate communication problems among network nodes so that decisions can be made safely and with confidence [7]. The failure detector plays a key role in a fault-tolerant system [8]. By periodically detecting the status of nodes in systems, it can provide information to achieve routing recovery, application deployment, real-time communication, etc. Thus, the high availability of VANETs can be guaranteed. An effective failure detector can provide information regarding suspected nodes in a timely and accurate manner so that correction actions can be performed as soon as possible. At present, there are numerous studies related to failure detectors in distributed systems [9]–[۱۳].