The Effect of Selfish Nodes on the Performance of Event-Based Trust Management In VANETS

With the increasing number of vehicles in traffic on a global scale and the production of smart vehicles, VANET systems have taken their place in smart city applications as an important sub-branch of intelligent smart transportation. Event-based VANET systems, which are modeled by the communication between vehicles and the infrastructure for events occurring in traffic, provide both faster and earlier action in the network, thus achieving high performance in traffic management and safety applications. However, the nature of VANET, which is sensitive to attacks based on its high dynamism, has led to special security requirements for these systems in the literature. While traditional or smart attacks are generally modeled in trust systems developed for VANET, selfish behavior, where vehicles refuse to provide information to the network, has mostly been ignored. Although selfish behavior does not directly harm the network like other attacks, it can cause erroneous decisions due to incomplete information collection. Therefore, in this study, an event-based VANET system is created; two different types of selfish behavior are modeled, including refusing to send warning message about the event and to send the local trust report to the center. The responses of recent central trust mechanisms in the literature to these selfish behaviors and the damage caused by selfishness to the network are modeled using simulation method and analyzed using fault tolerance and attack detection parameters. The results show that current approaches are robust against selfishness of not broadcasting, but are vulnerable to selfishness of not reporting. As a result, it is concluded that new trust mechanisms with different types of selfishness should be developed in event-based VANET systems.