DataSense: CIC IIoT Dataset 2025 for Real-Time Sensor-Based Attack Analysis

Overview

DataSense: CIC IIoT Dataset 2025 is a real-time sensor-based benchmark dataset for attack analysis in Industrial IoT (IIoT) environments. It is developed by the Canadian Institute for Cybersecurity (CIC) and described in an Electronics journal article as a resource for anomaly and intrusion detection with multi-objective feature selection.

IIoT Testbed

• Built in a dedicated IoT and IIoT laboratory at CIC with around 40 interconnected devices.
• Includes more than 15 types of industrial sensors, many built from scratch using Arduino boards alongside real industrial sensors.
• Also incorporates network equipment, IoT devices, edge devices, and attacker systems to form a realistic multi-layer IIoT environment.

Architecture and Data

• Testbed organized into five layers: IoT/IIoT Layer, Network Infrastructure, Edge Layer, Cloud Layer, and Attacker Layer.
• Dataset combines synchronized sensor readings with corresponding network traffic captures.
• Designed to support research on anomaly detection, intrusion detection, and resilience mechanisms for IIoT systems.

Use Cases

• Developing and evaluating machine learning models for IIoT anomaly and intrusion detection.
• Investigating feature selection strategies that balance detection accuracy and resource usage.
• Studying attack scenarios in realistic multi-layer IIoT deployments.

Access and License

The dataset is hosted by the Canadian Institute for Cybersecurity and linked from the official CIC DataSense page. The associated Electronics article provides citation details and experimental context; users should follow CIC terms of use and publisher policies.