IoT-RPL 2021: Cyber Attack Dataset Based on RPL Routing for IoT
收藏资源简介：
The Internet of Things (IoT) has emerged as a central focus within computer science research, with the Routing Protocol for Low Power and Lossy Networks (RPL) serving as a pivotal standard for IoT routing. Devices within IoT networks are characterized by their extensive connectivity, pervasive presence, and constrained processing capabilities. Identifying routing attacks on 6LoWPAN-based IoT devices poses a significant challenge due to network intricacies. Various techniques, including anomaly detection, have been proposed to detect and classify such attacks by analyzing network traffic characteristics. This study specifically addresses routing attacks targeting the widely utilized RPL protocol in 6LoWPAN-based IoT systems. The dataset (IoT-RPL) comprises ".csv" files documenting four distinct routing attacks—Blackhole Attack, Flooding Attack, DODAG Version Number Attack, and Decreased Rank Attack—obtained from the Cooja simulator. This dataset facilitates the development of Intrusion Detection Systems (IDS) for RPL-based IoT networks using Artificial Intelligence and Machine Learning methods, eliminating the need for attack simulations. Simulating these attacks under realistic conditions is crucial for testing protection mechanisms. This study offers an alternative to traditional intrusion detection systems, emphasizing the importance of identifying relevant attack attributes, analyzing network traffic data, and ensuring dataset balance and representativeness. In conclusion, this research represents a significant advancement in IoT security, providing a new dataset to support the application of artificial intelligence and machine learning methods in detecting routing attacks on IoT devices.
物联网（Internet of Things，IoT）已成为计算机科学研究的核心热点领域，低功耗有损网络路由协议（Routing Protocol for Low Power and Lossy Networks，RPL）作为IoT路由领域的关键标准，发挥着举足轻重的作用。IoT网络中的设备具备连接范围广泛、部署无处不在且处理能力受限的典型特征。由于网络结构复杂，针对基于6LoWPAN的IoT设备开展路由攻击检测是一项极具挑战性的任务。现有研究已提出包括异常检测在内的多种技术方案，通过分析网络流量特征实现此类攻击的检测与分类。本研究聚焦于基于6LoWPAN的IoT系统中广泛应用的RPL协议所面临的路由攻击问题。本次发布的IoT-RPL数据集包含若干.csv格式文件，记录了从Cooja模拟器中获取的四类典型路由攻击：黑洞攻击（Blackhole Attack）、泛洪攻击（Flooding Attack）、DODAG版本号攻击（DODAG Version Number Attack）以及秩递减攻击（Decreased Rank Attack）。该数据集可为基于RPL的IoT网络入侵检测系统（Intrusion Detection Systems，IDS）的开发提供数据支撑，支持人工智能（Artificial Intelligence）与机器学习（Machine Learning）方法的落地应用，无需研究者再自行开展攻击模拟实验。在真实场景下模拟攻击对验证防护机制的有效性至关重要，本研究提供的数据集为传统入侵检测系统方案提供了替代选择，强调了筛选有效攻击属性、分析网络流量数据以及保障数据集平衡性与代表性的重要性。综上，本研究在IoT安全领域取得了重要进展，提供了全新的数据集，可为人工智能与机器学习方法应用于IoT设备路由攻击检测提供有力支撑。
