UNSW-IoTraffic: Consumer IoT Packets, Flows and Protocols
Consumer IoT network traffic from 27 devices over about 203 days: 95.5 million packets and 4.9 million flows, released as packet, flow and protocol files.
View DatasetShowing 12 of 63 datasets
Consumer IoT network traffic from 27 devices over about 203 days: 95.5 million packets and 4.9 million flows, released as packet, flow and protocol files.
View DatasetNine attack captures for network intrusion detection, with a documented total of 27,170,754 packets, 115 statistical features per packet and separate binary labels.
View DatasetPublisher-reported 49,902,287 labeled IoT/IIoT records with 88 columns; the inspected attack log contains 302 sessions across 14 attack types.
View Dataset10,129 observations with 19 columns, including four sound channels and an occupancy target, from a seven-node indoor sensing installation.
View DatasetSmart home traffic captured under normal operation and seven multi-stage attack scenarios across heterogeneous end devices. CC BY 4.0, published 2026.
View DatasetAbout 4.6 million labelled bi-flows with 80 CICFlowMeter features, covering 15 attack families plus benign traffic in 16 single-class files. CC BY 4.0.
View DatasetInternet of Medical Things IDS dataset with BLE, IP packet, and IP flow captures. CSV and pickle formats, 2.7 GB, for ML-based healthcare IoT intrusion detection.
View DatasetMQTT IoT IDS dataset from a simulated network with 12 sensors, broker, camera, and attacker. PCAP and CSV features support MQTT intrusion detection research.
View DatasetIoT IDS dataset for distinguishing normal and malicious ICMP/Ping traffic generated from an ESP-01s embedded device. PCAP, Zeek logs, and labelled CSV files.
View DatasetFree CC0 synthetic dataset: 4,000 rows of Modbus, OPC UA and DNP3 flows labelled for SCADA intrusion detection. 21% Attacks. Reproducible from its seed.
View DatasetReal IoT botnet traffic dataset from 9 commercial devices (webcams, routers, thermostats) authentically infected by Mirai and BASHLITE. Over 7M records, 115 statistical features. CSV format. Benchmark for deep-learning-based IoT anomaly and botnet detection.
View DatasetLarge-scale distributed IoT IDS benchmark with traffic captured at individual device interfaces across 78 heterogeneous smart city IoT devices using the Gotham testbed. PCAP and CSV. Published January 2026 on Zenodo. Designed for federated learning and decentralised IDS research.
View Dataset