A Data-Driven Digital Twin Approach for Enabling Secure Communication in 5G-IoT Networks

In 5G-IoT networks, the concepts of virtualization and network slicing expand the attack surface and introduce new vulnerabilities across the core network, software layers and orchestration platforms, aerial communication, and non-terrestrial networks. Consequently, there is a pressing need for efficient security solutions and tools capable of protecting 5G-IoT systems dynamically and autonomously against evolving threat landscapes. To address these challenges, we developed Secure5G, a data-driven Digital Twin (DT)-based framework designed to enhance the security and resilience of 5G-IoT networks. The framework enables autonomous, adaptive defenses that proactively identify and mitigate emerging risks, including new, overlooked, or previously unaddressed threats that may compromise the reliability and continuous operation of 5G-IoT infrastructures, aerial communication systems, and related services. The DT component in the proposed approach creates a virtual replica of the 5G-IoT environment, facilitating continuous monitoring, adaptive learning, and resilience testing under evolving attack scenarios. Simulation experiments on 5G-IoT demonstrate that the DT provides realistic insights into the dynamic threat landscape. The proposed system achieves stable performance, reaching an accuracy of 84%, which improves to 91% when the training dataset size is increased to 80%. The results suggest that the DT-based adaptive security approach is well-suited for real-time deployment, offering incremental continuous learning and model updating capabilities for 5G-IoT-based critical sectors including aerial communication.