Akademik Veri Yönetim Sistemi
Cognitive Models and Artificial Intelligence Conference Proceedings, İstanbul, Türkiye, 25 - 26 Mayıs 2024, ss.13-19, (Tam Metin Bildiri)
– The Internet of Vehicles (IoV) is a branch of the Internet of Things that deals with vehicle-to-vehicle communication
and intelligent transport systems (ITS). But this connection is not without consequences, because the more a system exchanges
information, the more vulnerable it is to various attacks from malicious actors. (hackers). Vehicle Internet security is a great
challenge that security professionals face every day. Moreover, despite the deployment of diverse technologies by smart cities
to obtain varied, high-performance cloud services, security concerns continue to appear in communications entities that share
information. In this article, an intrusion detection system (IDS) based on machine learning is proposed to improve safety in
vehicle Internet systems (IoV). The IDS uses random forest (RF) algorithms, decision Tree, Adaboost, and gradient boost on an
IoV traffic data set. The hyperparameters of machine learning models are optimized using a meta-heuristic optimization
algorithm called the CDO (Chemotactic Differential Evolution) algorithm. The proposed IDS achieved high performance in
terms of up to 99.91% accuracy for the Adaboost algorithm in the binary case and 9.81% accuracy in the case of the decimal
dataset. High performance of precision, recall, and F1 score were also observed in this study. The optimization has significantly
improved the performance of the models by optimizing their hyperparameters. The study was conducted using data sets built by
CICIS (Canadian Institute of Cybersecurity) with a real vehicle to evaluate the proposed detection system. The experimental
results show that the proposed IDS has significantly higher detection performance.