Next-Generation Protection: Leveraging Federated Learning and Blockchain for Intrusion Detection in Smart Vehicle Network

Abstract

Smart Car's era ushers in new challenges, some of which are in the field of information security such as wise cyber-attack prevention. Traditional IDS systems would lose everything in such a dynamic environment with their centralized architectures, and this approach could create single points of failure and privacy issues.  Along with interconnectedness, cyber security becomes an inevitable problem as smart vehicles are incorporated into a daily life. Traditional security mechanisms usually lack scalability and privacy, which brings about the need to develop alter-nate or innovative methods. This research demonstrates a mixed security system that combines both federated learning and blockchain technologies to improve intrusion detection in smart vehicular networks. We evaluated the effectiveness of this framework using four machine learn-ing models as respectively; Support Vector Machine (SVM), Decision Tree, Neural Network, and Random Forest. Empirical results show that SVM had the highest accuracy of both 93.88% in training and 91.84% in validation, which is higher than Decision Tree, Neural Network, and Random Forest models. These findings evidently demonstrate that the federated learning and blockchain are a strong solution for the plausible security of smart vehicle networks; with SVM being employed mostly in complex security scenarios. DOI: https://doi.org/10.52783/pst.353