New Datasets Released for Cybersecurity Research: UNSW-NB15 and BoT-IoT

Severity: Low (Score: 18.9)

Sources: research.unsw.edu.au

Summary

On April 12, 2026, UNSW Canberra released two significant datasets for cybersecurity research: the UNSW-NB15 and BoT-IoT datasets. The UNSW-NB15 dataset includes 100 GB of raw network traffic, featuring nine types of attacks such as DoS and backdoors, captured using the IXIA PerfectStorm tool. This dataset is intended for academic use in developing intrusion detection and threat intelligence systems. The BoT-IoT dataset, comprising 69.3 GB of data, simulates a realistic network environment with normal and botnet traffic, including DDoS and data exfiltration attacks. Both datasets are available for free academic use, with authors asserting copyright and requiring citation of related papers. They aim to support research in network forensics, IoT security, and privacy preservation. Researchers can contact the authors for more information regarding the datasets. Key Points: • UNSW-NB15 dataset includes 100 GB of traffic with nine attack types. • BoT-IoT dataset simulates botnet traffic and includes over 72 million records. • Both datasets are available for free academic use with citation requirements.

Key Entities

• Analysis (attack_type)
• Backdoors (attack_type)
• Data Exfiltration (attack_type)
• DDoS (attack_type)
• DoS (attack_type)
• Fuzzers (tool)
• Argus (tool)
• Bro-IDS (tool)
• IXIA PerfectStorm (tool)
• Tcpdump (tool)
• Keylogging (mitre_attack)
• Reconnaissance (mitre_attack)
• T1041 - Exfiltration Over C2 Channel (mitre_attack)
• T1056 - Input Capture (mitre_attack)
• ieee.org (domain)
• unsw.edu.au (domain)