CoAP_UAD: CoAP under attack dataset - A comprehensive dataset for CoAP-based IoT security research.

Area of Science:

• Cybersecurity
• Internet of Things (IoT)
• Network Protocols

Background:

• Lightweight protocols like the Constrained Application Protocol (CoAP) are crucial for resource-constrained Internet of Things (IoT) devices.
• However, their design and reliance on UDP expose CoAP networks to unique protocol-level attacks.
• Effective intrusion detection is vital for maintaining the reliability and safety of these IoT deployments.

Purpose of the Study:

• To introduce CoAP_UAD, a novel, publicly available dataset for studying and benchmarking intrusion detection systems for CoAP-based IoT networks.
• To provide a standardized resource for evaluating defenses against protocol-semantic attacks, abstracting away device-specific vulnerabilities.

Main Methods:

• Developed a realistic testbed emulating constrained IoT devices and CoAP deployments.
• Executed diverse protocol-oriented attacks, including cross-protocol interaction, message manipulation, and Denial of Service (Block size amplification).
• Captured network traffic at the router, labeling each frame/flow and exporting data in CSV format for reproducibility.

Main Results:

• Created CoAP_UAD, a dataset featuring labeled network traffic from a realistic IoT testbed.
• The dataset includes various attacks specifically targeting CoAP protocol behaviors.
• Data is formatted in CSV for ease of use in intrusion detection system research.

Conclusions:

• CoAP_UAD offers a valuable resource for advancing research in IoT network security.
• The dataset enables the development and benchmarking of intrusion detection methods tailored for CoAP.
• It supports the creation of more robust and secure constrained IoT environments.