Detection and Isolation of Malicious Nodes in Vehicular Networks Using Trust Aggregators

Vehicular networks have grown significantly in popularity in recent years as a result of their ability to increase traffic efficiency, increase road safety, and serve as a platform for numerous cutting-edge applications. They present themselves as vulnerable targets for malicious users in the network. Malicious nodes are entities in vehicular networks that are intentionally or unintentionally ‘misbehaving’, and they can cause various problems, such as disrupting communication, stealing sensitive data, or even disabling critical safety systems. This work builds upon the TruMan model to detect and isolate nodes in vehicular networks. An aggregator function is proposed in this work to summarize the behavior of nodes in a network segment, and to identify anomalies that shall indicate the presence of a malicious node. With the integration of the aggregator function to the original TruMan model, the proposed model shows an improvement in the overall rate of network discoverability.