An Enhanced 6G-Enabled IoT Network Security Framework Using Edge AI and Machine Learning

Abstract

The 6G wireless network can combine with the Internet of Things (IoT) to enable terahertz (THz) communication, reconfigurable intelligent surfaces (RIS), network slicing, and AI-native radio access networks (RANs), which would connect trillions of heterogeneous resource-constrained devices. Such enormous size, coupled with the demands of ultra-low latency and mass connectivity, significantly increases the attack surface that IoT ecosystems face today, and makes it difficult to meet the latency, bandwidth, and privacy demands of 6G. This study provides a comprehensive discussion of the current state of security research in Edge Artificial Intelligence and Machine Learning approaches to securing 6G-enabled IoT networks, and is explicitly contextualized with respect to previous surveys of 6G-IoT security, while uniquely focusing on introducing the reader to the field of Edge AI and ML 6G-IoT security. We review recent works dealing with federated learning-based intrusion detection, deep and hybrid learning architectures, deep reinforcement learning for adaptive defense, TinyML for extreme-edge inference, explainable AI for auditability, and blockchain-assisted zero trust access control. On top of the identified gaps, we present an Enhanced 6G-IoT Security Framework, consisting of a closed-loop, federated detection-and-response workflow that incorporates a five-layer architecture, namely the perception, edge-AI, 6G network, cloud-intelligence, and application layers. Comparative tables represent techniques, threat taxonomies, and design trade-offs, as well as architecture. The study ends by highlighting many open challenges, such as the lack of 6G-IoT representative datasets, adversarial robustness, energy-latency-accuracy tradeoffs at the extreme edge, as well as the need for quantum-safe secure information, models with explainability, and sustainable secure information. The study provides a basic research roadmap for developing more trustworthy 6G-IoT systems in the future.