The fast rate of cross-domain Digital Twins (DT) ecosystem growth in 6G-based scenarios poses unresolved security and privacy issues into the scope of existing frameworks. This study examines the inherent constraints of existing methods and presents TwinGuard-Sec, a new federated blockchain-based AI system expressly aimed at providing a set of standardized security and privacy of data in a heterogeneous realm of DT. The methodology comprises a dual-layered systematic architectural framework, comprising an AI-governed threat intelligence unit and zero-knowledge identity verifications and a distributed ledger technology layer that is domain-interoperable with lightweight consensus algorithms ensuring synchronous operation in real time. The framework fills three essential gaps in research including: absence of standardized cross-domain security protocols, inadequate privacy preserving mechanisms applied to sensitive inter-organizational data sharing and lack of scalable consensus algorithms to be used in DT-specific needs. We show on the rigorous test of a comprehensive 6G virtual twin testbed that includes 50 distributed nodes and five application domains (smart mobility, e-health, industrial IoT, smart cities, and autonomous systems) that the performance is significantly improved: 27.4% increase in threat detection accuracy (reaching 95.0% vs. 76.4% base) can be improved, better privacy preservation with a differential privacy parameter = 0.94 (62% improvement), 21.2% reduction in latency down to 147 ms. The system achieves Precision = 0.968, Recall = 0.959, and F1-score = 0.963 (macro-average), with AUC-ROC = 0.989 across eight attack categories. These results confirm that TwinGuard-Sec is an innovative means of ensuring the safety of cross-domain DT coordination, equipping it with both theoretical frameworks and implementation channels of the next generation intelligent infrastructure systems.
Building similarity graph...
Analyzing shared references across papers
Loading...
Mrim M. Alnfiai
Reemiah Muneer Alotaibi
Faiz Abdullah Alotaibi
Scientific Reports
King Saud University
Taif University
Imam Mohammad ibn Saud Islamic University
Building similarity graph...
Analyzing shared references across papers
Loading...
Alnfiai et al. (Tue,) studied this question.
www.synapsesocial.com/papers/69fbefef164b5133a91a424c — DOI: https://doi.org/10.1038/s41598-026-49490-3