Vehicle-to-Vehicle Flooding Datasets using MK5 On-board Unit Devices

Area of Science:

• Computer Science
• Network Security
• Cybersecurity

Background:

• Vehicular networks are crucial for modern transportation, enabling novel services.
• Network availability is paramount, and attacks like flooding can disrupt services and endanger lives.
• Existing datasets for vehicular flooding attacks often rely on simulated data, lacking real-world complexity.

Purpose of the Study:

• To generate realistic datasets of flooding attacks in vehicular networks.
• To analyze the complexity of these attacks using real vehicular devices.
• To provide a foundation for developing robust detection mechanisms.

Main Methods:

• Utilized three realistic MK5 On-board Unit (OBU) devices to generate attack datasets.
• Employed a machine learning algorithm for initial data analysis.
• Evaluated performance using Accuracy, F1-Score, Precision, and Recall metrics.

Main Results:

• Successfully generated datasets capturing realistic vehicular flooding attack scenarios.
• Machine learning analysis provided initial insights into dataset complexity.
• Quantified detection performance using standard machine learning evaluation metrics.

Conclusions:

• The generated datasets offer a valuable resource for studying vehicular network security.
• Real-world data is essential for understanding and mitigating flooding attacks.
• Further research can leverage these datasets to develop advanced intrusion detection systems.