Post-Quantum 5G Slicing Security Framework Integrating AI-Driven Threat Mitigation in Global Telecommunications

Abstract

The rapid deployment of fifth-generation (5G) networks and the adoption of network slicing have significantly enhanced connectivity and service customization, but they have also expanded the cyber-attack surface of telecommunications infrastructure. At the same time, emerging quantum computing capabilities threaten the security of widely used classical cryptographic schemes such as RSA and Elliptic Curve Cryptography. Despite growing concerns, existing security frameworks for 5G slicing rarely integrate post-quantum cryptography with intelligent threat detection mechanisms. This study addresses this gap by developing a quantum-resistant security architecture for 5G network slicing that combines post-quantum cryptography and artificial intelligence–based anomaly detection. The framework incorporates NIST-standardized algorithms including CRYSTALS‑Kyber for secure key exchange and CRYSTALS‑Dilithium and FALCON for authentication. A hybrid post-quantum/classical TLS scheme was implemented to support secure key establishment with minimal latency overhead of 1.65 ms. Additionally, an AI-driven anomaly detection model using a stacked ensemble of Isolation Forest and XGBoost monitored slice orchestration behavior to detect cyber threats. Simulation results demonstrated high detection performance with a macro F1-score of 0.962 and AUC-ROC of 0.985. The proposed framework offers a practical and scalable approach for securing next-generation telecommunications infrastructure against both current cyber threats and future quantum attacks.