1,176,226 labeled network traffic records across 15 categories including normal behavior and 14 specific attack types. The data captures telemetry and network flows from 10 distinct IoT and IIoT devices, including Modbus sensors, DHT11 sensors, and ultrasonic sensors, integrated into a real-world testbed.
Use Cases
- Train a multi-class classification model to identify specific cyber threats using the Attack_type column
- Develop binary intrusion detection systems using the Attack_label feature to distinguish normal traffic from malicious activity
- Analyze protocol-specific vulnerabilities by examining features related to MQTT, Modbus, and HTTP traffic
Strengths
- 1,176,226 rows of network traffic data collected from a physical IoT/IIoT testbed
- 61 features including network layer headers and application layer protocol fields
- 14 specific attack labels such as DDoS_UDP, SQL_injection, Backdoor, and MITM
- Covers 10 physical IoT devices including Modbus sensors, Flame sensors, and water level sensors