A Review on the Integration of Blockchain and Intelligent Learning Techniques for Threat Detection and Security Enhancement in IoT Networks

Abstract

The growth of the Internet of Things (IoT) has brought about security concerns owing to the massive global integration of diverse and constrained devices. Blockchain, Machine Learning (ML) and Deep Learning (DL) techniques have been proposed as effective ways to improve IoT security. This paper reviews the literature on the use of blockchain and smart learning techniques for security enhancement and threat detection in IoT networks. Blockchain offers a decentralized and immutable approach to secure data integrity, authenticating and controlling access to IoT devices, while ML and DL techniques allow intelligent monitoring of network data to detect anomalies and predict cyber-attacks. This study follows a systematic literature review approach, examining peer-reviewed journal articles and conference proceedings from 2019 to 2024. The analysis shows while blockchain, ML and DL technologies play a crucial role in enhancing IoT security, they each have limitations including scalability, computational overhead, data dependency and lack of flexibility against new cyber threats. Additionally, the majority of studies concentrate on either blockchain-based security or ML-based intrusion detection, with limited study on the integration of the two for real-time threat detection and mitigation. The study highlights this limitation and calls for the development of intelligent hybrid models that integrate blockchain technology with ML/DL to address scalability, adaptability and real-time security for IoT networks to ensure confidentiality, integrity and reliability.