MQTT_UAD: MQTT Under Attack Dataset. A public dataset for the detection of attacks in IoT networks using MQTT protocol
José Aveleira‐Mata, Héctor Aláiz‐Moretón, Martín Bayón-Gutiérrez, María Teresa Gacía-Ordás, Natalia Prieto-Fernández, Isaías García-Rodríguez · Data in Brief · 2025
Internet of Things (IoT) systems have become increasingly popular in recent years, leading to a growing need to ensure their security. The diversity of devices and protocols, combined with the limited computational power of many of these devices, makes them especially vulnerable to a wide range of threats. Detecting cyberattacks is therefore essential to ensure the safe operation of systems that IoT devices. In this work, we present MQTT UAD (MQTT Under Attack Dataset), a dataset specifically designed to analyze security in networks based on the MQTT protocol. This dataset has already been used to train intrusion detection models, but it had not been published independently or described in technical detail. Due to the interest it has generated, we provide a clear name and a full description of how it was created, processed, and structured. To build the dataset, we set up a realistic test environment where different attacks against the MQTT protocol were simulated, including Denial of Service (DoS), Man-in-the-Middle (MitM), and Intrusion attacks. Network traffic was captured during these tests, and the malicious frames were tagged. Finally, we validate the MQTT UAD for developing machine learning models aimed at detecting attacks in MQTT networks.