Securing Disaster Management Systems: A Comparative Study of ML and DL-Based IDS for Mirai Botnet Detection

The Internet of Things (IoT) has become a corner-stone of modern infrastructure, supporting essential systems like Disaster Management Systems (DMS). However, as IoT technologies advance, they also become more susceptible to cyber threats, compromising privacy, confidentiality, and system availability. Among these threats, the Mirai botnet is particularly dangerous due to its capacity to disrupt IoT networks. This research evaluates three models to build an Intrusion Detection System (IDS) to effectively detect Mirai botnet attacks, thereby ensuring the availability of DMS. The IDS is designed to accurately identify three key Mirai attack variants, preventing the formation of botnets and protecting critical IoT services. The results show that XGBoost achieved a high accuracy of 99.98%, rapid detection time, and a low False Positive Rate (FPR), the proposed IDS-based XGBoost model offers a robust solution for strengthening IoT security.