A Signature-based Intrusion Detection System for the Internet of Things

Internet of Things (IoT) is envisioned as a transformative approach with a wide range of applications in various sectors such as home automation, industrial control, and agriculture. It promises innovative business models and improved user experience. However, as evidenced by recent attacks such as the Mirai botnet, IoT networks and systems remain very vulnerable and require stronger protection mechanisms. Furthermore, due to processing, memory, and power constraints of typical IoT devices, traditional Internet security mechanisms are not always feasible or appropriate. In this work, we are concerned with designing an Intrusion Detection System (IDS) for protecting IoT networks from external threats as well as internal compromised devices. Our proposed design adopts a signature-based intrusion detection approach and involves both certralised and distributed IDS modules. Using the Cooja simulator, we have implemented a Denial of Service (DoS) attack scenario on IoT devices. This scenario exploits the RPL protocol, which is widely used for routing in low-power networks, including IoT networks. In particular, we have implemented two variants of DoS attacks, namely “Hello” flooding and version number modification. As shown by simulation results, these attacks may impact the reachability of certain IoT devices and their power consumption.