MQTT-IoT-IDS2020

消息队列遥测传输 (MQTT) 协议是物联网 (IoT) 机器对机器通信中最常用的标准之一。可用物联网设备和使用的协议数量的增加加强了对新的和强大的入侵检测系统 (IDS) 的需求。但是，构建 IoT IDS 需要有数据集来处理、训练和评估这些模型。 MQTT-IoT-IDS2020 是第一个模拟基于 MQTT 的网络的数据集。数据集是使用模拟的 MQTT 网络架构生成的。该网络包括十二个传感器、一个代理、一个模拟摄像头和一个攻击者。记录了五个场景：（1）正常操作，（2）攻击扫描，（3）UDP扫描，（4）Sparta SSH暴力破解，（5）MQTT暴力破解。保存原始 pcap 文件，然后提取特征。从原始 pcap 文件中提取了三个抽象级别的特征：（a）数据包特征，（b）单向流特征和（c）双向流特征。数据集中的 csv 特征文件适合机器学习 (ML) 使用。此外，原始 pcap 文件适用于更深入地分析 MQTT IoT 网络通信和相关攻击。

Message Queuing Telemetry Transport (MQTT) is one of the most widely used standards in machine-to-machine communication for the Internet of Things (IoT). The growing number of available IoT devices and their utilized protocols has heightened demand for new, robust intrusion detection systems (IDS). However, developing IoT IDS requires datasets for processing, training, and evaluating such models. MQTT-IoT-IDS2020 is the first dataset that simulates MQTT-based network environments. It was generated using a simulated MQTT network architecture consisting of twelve sensors, a broker, a simulated camera, and an attacker. Five scenarios are recorded: (1) Normal operation, (2) Attack scan, (3) UDP scan, (4) Sparta SSH brute-force attack, (5) MQTT brute-force attack. Raw pcap files are saved first, followed by feature extraction. Three levels of features are extracted from the raw pcap files: (a) Packet features, (b) Unidirectional flow features, and (c) Bidirectional flow features. The CSV feature files in the dataset are suitable for machine learning (ML) applications. Additionally, the raw pcap files enable in-depth analysis of MQTT IoT network communications and related attacks.

搜集汇总

背景与挑战

背景概述

MQTT-IoT-IDS2020是首个模拟基于MQTT协议的物联网入侵检测数据集，专为训练和评估入侵检测系统而设计。它包含五个场景的数据，涵盖正常操作和多种攻击类型（如扫描和暴力破解），并提供原始pcap文件和提取的csv特征文件，适用于机器学习和网络通信深度分析。数据集由学术机构于2020年发布，支持物联网安全研究。