Quantum-Secured Federated Learning for Privacy-Preserving and Adaptive Attack Detection in 6G IoT
S. Abarna, Rajasri Kathavarayan, Kanchana Kavaratty Raju, Shofia Priyadharshini Dayalan, Jeyakarthic Mohan, Vijayabaskaran Pavadaisamy Sivakolundu · 2025
The advent of sixth-generation (6G) networks introduces unprecedented capabilities for Internet of Things (IoT) ecosystems, empowering devices with ultra-fast connectivity and facilitating data-driven applications across sectors. The highly interconnected nature of 6G IoT devices also expands the attack surface, necessitating advanced security frameworks that protect device data and prevent unauthorized access. This chapter proposes a novel Privacy-Preserving and Adaptive Attack Detection System using Quantum-Secured Federated Learning (QFL) to secure 6G-enabled IoT networks. The QFL framework combines federated learning (FL) with quantum-resistant encryption to create a decentralized and adaptive attack detection mechanism, maintaining data privacy and robust security across devices. In our approach, IoT devices locally train anomaly detection models on real-time data, capturing suspicious patterns without exposing raw data. These local model updates are quantum-encrypted and securely combined at a central server, resulting in a global model that continuously evolves to detect emerging attack vectors. The combination of quantum-resistant cryptography ensures resilience against potential quantum computing threats, safeguarding IoT devices from eavesdropping and tampering.