Document is current

Article History

Received: 20 November 2024

Revised: 9 February 2026

Accepted: 12 February 2026

First Online: 10 June 2026

Declarations

:

: The authors declare no competing interests.

: The needs for data security and privacy in industrial settings were carefully taken into account when designing the suggested DTGNN-FIDS framework. The Edge-IIoTset, ToN-IoT, and IoT-23 benchmark datasets, which are publicly accessible and anonymized and devoid of sensitive operational data or personally identifiable information, were used for all experimental evaluations in this study. Instead of using raw payload information, the suggested system mainly uses network-level metadata (such as traffic statistics, communication patterns, and resource indicators) in real-world deployments. This design decision lowers privacy threats and exposes as little sensitive industrial information as possible. Additionally, by limiting needless data transfer between edge and cloud components, the event-triggered update system lowers the possibility of communication eavesdropping and data leakage. Standard industrial cybersecurity procedures presume safe execution environments at edge and cloud nodes, encryption of transferred data, and access control.

: Only a minimal amount of language editing and grammatical refining was done during text preparation using generative artificial intelligence (AI) technologies. The creation of scientific content, research ideas, technique, experimental design, data analysis, results, figures, and conclusions was not done by an AI system. The authors created all of the technical content, including the suggested DTGNN-FIDS framework, mathematical formulas, simulations, and performance analysis. All text was thoroughly examined and revised by the writers, who also accept full responsibility for the work’s integrity, originality, and accuracy.