Development of IoT Threat Intelligence System

Abstract

The rapid proliferation of IoT device has brought innovative opportunities as far as it has unveiled new attack venues for cyber-attacks. As a result of limited computational powers and weak security configurations, IoT-based systems fall easy victims to attacks like Unauthorized Access, Denial of Service (DoS) attack, and Distributed Denial of Service (DDoS) attack. We present in this work an IoT Threat Intelligence System design and development to support real-time monitoring, detection, and threat classification of IoT-targeted cyber-attacks. Based on key traffic features extracted from TON_IoT Datasets, a Random Forest classifier is trained to predict threat. We achieved accuracy of 99.5% with near-perfect precision, recall, and F1-measure scores against all types of attacks to demonstrate overwhelming reliability and generalizability. Detected threats are punted temporarily into a lightweight SQLite database and streamed in real time via MQTT protocol. A React log panel provides visualization of log streams, threat patterns, and system activity to provide administrators and users with a form of situational awareness.