DALHOUSIE NIMS LAB IOT 2024 DATASET

This dataset comprises data captured from IoT devices, a real-life usage scenario which we call Active as experienced by everyday users. Our setup focuses on capturing realistic data through this scenario, providing a comprehensive understanding of IoT devices. Rigorous testing and system evaluations were conducted to ensure the collection of realistic data at connection points in real time. All captured data is benign, free from any malware indications. The capturing structure was meticulously designed to exclusively record outbound and inbound traffic for each device, minimizing the risk of noise or unrelated capture.

The dataset comprises of nine popular IoT devices namely
Amcrest Camera
Smarter Coffeemaker
Ring Doorbell
Amazon Echodot
Google Nestcam
Google Nestmini
Kasa Powerstrip
Samsung 32 inch Smart Television (TV)
Amazon Smartplug.

The traffic collected from these devices are stored as individual .pcap file. For our research, we extract flows from these .pcap files using flow analysis tools precisely (Tranalyzer and NFStream). Therefore, the flow dataset of Tranalyzer2, NFStream (Default and Extended) for each of the each device are also contained with this dataset.

The difference between NFStream Default and Extended are with respect to feature sets and not the number of flows. Within the main folder of the dataset, you will find four folders namely (PCAP, TRANALYZER, NFSTREAM DEFAULT AND NFSTREAM EXTENDED. Each folder contains the scenario (Active) with which within the Active folder contains respective files named after the devices.

The Pcap files are labelled as active.iteration.pcap.

The Tranalyzer files are labelled as active.iteration.txt.

The NFStream files are labelled as active.iteration.csv.

Depth explanation is provided in our paper.