Performance Evaluation of Intrusion Detection System Using Gradient Boost

Vehicular Ad Hoc Networks (VANETs) are pivotal in modern intelligent transportation systems, enabling real-time vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communication. These networks facilitate various applications like traffic management, collision avoidance, and infotainment services. However, the open and dynamic nature of VANETs exposes them to distinct security challenges, necessitating the deployment of Intrusion Detection Systems (IDS) to ensure the security and privacy of vehicular communications. This study introduces an innovative IDS for VANETs, focusing on addressing the unique security issues prevalent in this domain, particularly through advanced feature selection techniques, handling class imbalance with Synthetic Minority Oversampling Technique (SMOTE), and leveraging the Gradient Boost algorithm for classification. The efficacy of the proposed IDS is evaluated on the NSL-KDD dataset, demonstrating exceptional performance compared to existing models the Random Forest algorithm, renowned for its robustness, with an accuracy rate of 100 and 99% for precision, recall, and f1 score, as well as a precision-recall curve with an AP score of 1.0.