REAL-TIME POLICY FEEDBACK FOR DYNAMIC ACCESS CONTROL IN 5G SMART CITIES

Abstract

5G-enabled smart cities integrate diverse IoT ecosystems like smart meters, traffic sensors, and CCTV, demanding robust access control amid latency spikes and anomalous device behaviors. This paper introduces a policy-driven closed-loop framework building on Blockchain-based Access Assurance Policy Framework (BAAPF), incorporating real-time AI-monitored feedback to dynamically adjust micro-segmentation policies—tightening during anomalies or relaxing under normal conditions. Leveraging Ethereum smart contracts for tamper-proof policy enforcement, the system ensures zero-trust multi-tenant access without domain leaks, supporting OT/IT convergence in resourceconstrained 5G networks.The architecture features anomaly detection via ML models analyzing network metrics and device telemetry, triggering automated policy updates with sub-100ms latency. Evaluations use NS-3 simulations for 10,000+ device scalability and a PoC on private testnets, benchmarking against static blockchain schemes across smart-metering data flows, traffic analytics, and video feeds. Results show 40% faster anomaly response, 25% reduced overhead, and enhanced resilience, validating feasibility for real-world deployment. Future extensions target 6G integration and federated learning for policy optimization