An Empirical Study of High-Risk Vulnerabilities in IoT Systems

Internet of Things (IoT) systems are increasingly widespread across various fields. Concurrently, vulnerabilities in IoT devices are continuously emerging, potentially leading to severe consequences, such as information leakage, system failure, or even resource abuse. For IoT system developers, understanding the characteristics of these critical vulnerabilities is crucial for safeguarding the security of IoT systems. However, existing studies on IoT vulnerabilities have not specifically focused on such severe or critical vulnerabilities, i.e., high-risk vulnerabilities. Moreover, previous works analyzed IoT vulnerabilities based on unofficial information sources, such as online reports, GitHub issues, or open-sourced projects. To fill this gap, this article presents the first large-scale empirical study on high-risk IoT vulnerabilities based on the well-known vulnerability data source, i.e., the national vulnerability database (NVD), which is maintained by the U.S. government. We constructed a database consisting of 1739 IoT-related vulnerabilities archived over the last two decades (from 1999 to 2023), including 1076 high-risk vulnerabilities for analysis. We classified the high-risk vulnerabilities into four categories, consisting of 25 different weakness types. We further collected 11 detection tools and summarized their capabilities in detecting IoT vulnerabilities. Our study sheds lights on new findings and insights for developers to secure the IoT system.