Abstract:Nowadays, VANETs are becoming a very interesting research topic for researchers as the benefits are very high in terms of ensuring driver comfort, enhancing road effectiveness and minimizing the risk of accidents. VANET is a wireless network directly linked to the Internet that links multiple vehicles through the use of OBUs (onboard units) to contact and communicate with the other units and RSUs (roadside units). This can be both an advantage and a risk for VANETs as the number of communications offered by this type of network including both vehicle-to-infrastructure (V2I) and vehicle-to-vehicle (V2V) forms of communication continues to grow, making VANETs increasingly susceptible to many types of cyber security attacks, such as denial of service attacks DOS, false and alternative messages, drive-by downloads, and false alarms. By identifying misbehaving vehicles, intrusion detection systems (IDS) significantly contribute to the protection of vehicle networks. In this research, we present an ensemble learning method, AdaBoost, as the basis for an IDS. To address the attack class imbalance problem, we employed the synthetic minority oversampling approach, or SMOTE, and for feature selection, we used the Chi-Squared technique. By creating new synthetic examples close to the other objects, the SMOTE technique helps to improve the minority classes while preventing overfitting, and Chi squared aids in the solution of the feature selection issue by examining the relationship between features. The NSL-KDD and UNSW-NB15 datasets, two of the most popular datasets these days, will be utilized to test our model. The following metrics were employed to assess our proposed model: It accomplishes this by creating new synthetic examples in feature space that are near to the other points (that is, members of the minority class). Chi squared then assists us in selecting features by examining the relationship between features that utilize the three new features. The tree datasets NSL-KDD, UNSW-NB15, and TON-IOTthree of the most popular datasets these days—are utilized to test our model. The following metrics have been applied to assess our suggested model. We used 10 cross-validation, f1-score, accuracy, precision, and recall to our model. Our model approach outperforms the current IDSs in terms of accuracy, recall, and precision, scoring nearly 100% on all metrics (accuracy, precession, recall, and f1-score).

An R–A dual network detection model for abnormal behavior of running vehicles

Multiclass objects detection algorithm using DarkNet-53 and DenseNet for intelligent vehicles

Abnormal Driving Detection Based on Normalized Driving Behavior

A driver’s dangerous driving behavior detection method based on the YOLOV5-CA network

Network Abnormal Traffic Detection Model Based on Semi-Supervised Deep Reinforcement Learning

A Robust Model for Predicting Abnormal Behavior in Vehicular Networks using AdaBoost and Chi-Square

Dual-Modality Vehicle Anomaly Detection via Bilateral Trajectory Tracing

Driving Behavior Recognition Algorithm Combining Attention Mechanism and Lightweight Network

Violation Behavior Detection for Non-motor Vehicles

Research on Deep Learning Automatic Vehicle Recognition Algorithm Based on RES-YOLO Model

Research on Small Target Detection in Driving Scenarios Based on Improved Yolo Network

One-Dimensional Convolutional Neural Network Model for Abnormal Driving Behaviors Detection Using Smartphone Sensors

YOLO-ABD: A Multi-Scale Detection Model for Pedestrian Anomaly Behavior Detection

A Novel Driver Abnormal Behavior Recognition and Analysis Strategy and Its Application in a Practical Vehicle

YOLOv5s-MEE: A YOLOv5-based Algorithm for Abnormal Behavior Detection in Central Control Room

Detecting Socially Abnormal Highway Driving Behaviors via Recurrent Graph Attention Networks

Deep-Neural-Network-Enabled Vehicle Detection Using High-Resolution Automotive Radar Imaging

Research on Vehicle Detection in Infrared Aerial Images in Complex Urban and Road Backgrounds

Abnormal Traffic Detection for Internet of Things Based on an Improved Residual Network

A High-Precision Vehicle Detection and Tracking Method Based on the Attention Mechanism

A Study on Machine Learning-Enhanced Roadside Unit-Based Detection of Abnormal Driving in Autonomous Vehicles