Providing IoT host-based datasets for intrusion detection research

Abstract

The high number of vulnerabilities in Internet of Things devices has created malware-prone networks. A type of malware that imposes a serious threat to the Internet security is known as botnets. This malware exploits some vulnerabilities of IoT devices to infect them and perform large-scale Distributed Denial of Service attacks, affecting many users who depend on their services. This work presents the construction of an experimental environment to generate a dataset that contains data from a real IoT device that was infected by botnet malware in a laboratory. The dataset can be used to support the development of defence tools for IoT devices to identify botnets, as it contains network traffic and host-based features, such as, CPU and memory usage. The dataset and network environment files are available for the research community.