Data-Driven Intelligence for Characterizing Internet-scale IoT Exploitations
| Authors | |
|---|---|
| Year of publication | 2018 |
| Type | Article in Proceedings |
| Conference | 2018 IEEE Globecom Workshops |
| MU Faculty or unit | |
| Citation | |
| Web | https://ieeexplore.ieee.org/document/8644468 |
| Doi | http://dx.doi.org/10.1109/GLOCOMW.2018.8644468 |
| Keywords | network monitoring;darknet;IoT;cyber security |
| Description | While the security issue associated with the Internet-of-Things (IoT) continues to attract significant attention from the research and operational communities, the visibility of IoT security-related data hinders the prompt inference and remediation of IoT maliciousness. In an effort to address the IoT security problem at large, in this work, we extend passive monitoring and measurements by investigating network telescope data to infer and analyze malicious activities generated by compromised IoT devices deployed in various domains. Explicitly, we develop a data-driven approach to pinpoint exploited IoT devices, investigate and differentiate their illicit actions, and examine their hosting environments. More importantly, we conduct discussions with various entities to obtain IP allocation information, which further allows us to attribute IoT exploitations per business sector (i.e., education, financial, manufacturing, etc.). Our analysis draws upon 1.2 TB of darknet data that was collected from a /8 network telescope for a 1 day period. The outcome signifies an alarming number of compromised IoT devices. Notably, around 940 of them fell victims of DDoS attacks, while 55,000 IoT nodes were shown to be compromised, aggressively probing Internet-wide hosts. Additionally, we inferred alarming IoT exploitations in various critical sectors such as the manufacturing, financial and healthcare realms. |
| Related projects: |