IoTminer: Semantic Information Extraction in the Packet Payloads
Yaxin Zhang, Min Huang, Qiang Li, Limin Sun · GLOBECOM 2022 - 2022 IEEE Global Communications Conference · 2022
Nowadays, massive Internet-of-Thing (IoT) devices are connecting with cyberspace, yet they suffer increased attack risks from known vulnerabilities to low-hanging exploitable manners. A proactive defense can help security professionals to discover potential risks, where IoT device identification is a necessary requisite. However, existing approaches suffer from coarse-grained and manual labor. In this work, we propose an automated semantic extraction approach, called IoTminer, which generates IoT device annotation from the packet payload. Specifically, IoTminer leverages relations between device types, vendors, and products to mine relevant entities for an annotation tuple (type, vendor, product). Further, we have implemented a prototype of IoTminer and conducted a real-world experiment to validate its efficacy. Results show that our IoTminer generates IoT device information at a fine-grained level, achieving 91.33% precision, 93% recall, and 90% F1 score. Moreover, the IoTminer can discover new IoT devices compared with state-of-the-art tools.