Digital Twin-Enabled Security Situation Awareness for Cloud-Edge Computing: A Dynamic Mapping and Predictive Analysis Approach

Abstract

Security situation awareness (SSA) is essential for proactive defense in cloud-edge computing, but traditional methods struggle with heterogeneous entity mapping, multi-source data fusion and emerging threat prediction. Digital Twin (DT) offers a new technical path to address these bottlenecks. This study proposes a DT-Enabled SSA framework (DT-SSA), which builds a high-fidelity virtual mirror of cloud-edge physical systems to achieve full-cycle SSA covering dynamic mapping, real-time perception, fusion analysis and predictive early warning. The framework includes four core modules; an adaptive feature alignment-based multi-scale mapping algorithm enables accurate physical-virtual matching, and a GNN-LSTM hybrid model realizes real-time situation analysis and threat prediction. Experiments on a real cloud-edge testbed show DT-SSA achieves 97.1% situation assessment accuracy and 93.5% 5–10-minute threat prediction accuracy, with 8.3ms latency. It outperforms traditional methods by 42.8% higher threat prediction lead time and 19.6% lower false warning rate, providing a novel solution for cloud-edge security governance.

Previous article in issue Next article in issue