Integration of Blockchain and Machine Learning for Robust Intrusion Detection in IoT Network

This thesis explores the integration of blockchain technology and machine learning techniques to enhance cybersecurity in IoT networks. An intrusion detection system (IDS) framework is proposed, leveraging the decentralized, immutable, and transparent features of blockchain with the predictive capabilities of machine learning. The study utilizes the CICIoT2023 dataset, which has 33 diverse attack types on 105 devices, to evaluate five machine learning models: Random Forest, MLP, CNN, and RNN and a Hybrid RF/MLP Model. The Random Forest and MLP models demonstrated superior performance across multi-class, grouped, and binary classification tasks. Blockchain technology is integrated to securely log detected anomalies, to ensure data integrity and transparency. The proposed IDS framework significantly outperformed existing solutions and addresses the common challenges in IoT security. This research highlights the potential of blockchain and machine learning integration in creating robust, scalable, and flexible cybersecurity solutions for IoT ecosystems.