| ABSTRACT |
The rapid proliferation of Internet of Things (IoT) devices has introduced significant security vulnerabilities, creating an expanding attack surface often exploited for botnets and distributed cyberattacks. To address this challenge, this thesis presents the development of an intelligent IoT Network Security Monitoring System designed to detect, analyse, and mitigate threats in real-time by leveraging the OpenWrt operating system on a Raspberry Pi 5 (1), integrated with an Intel BE200 Wi-Fi 7 module (2) via a Waveshare PCIe to M.2 E Key HAT (3), to establish a secure gateway. The system employs a hybrid artificial intelligence engine that combines a supervised Random Forest model for identifying known attack signatures with an unsupervised Isolation Forest model for detecting zero-day anomalies, implemented within a stateless architecture to optimise performance. Key features include high-speed packet inspection, automated mitigation via dynamic firewall blocking, instant Telegram notifications, and a comprehensive web-based dashboard for real-time visualisation of network health. Experimental results demonstrate the system's ability to successfully identify and block various attack vectors, such as port scans and flood attacks, with high precision, illustrating that enterprise-grade intrusion prevention can be effectively achieved on affordable hardware to provide a practical security framework for modern smart environments.
|