Abstract
The profound integration of informatisation and industrialization is amplifying security concerns regarding digital twins within industrial IoT (IIoT) network protocols. Current techniques for detecting vulnerabilities in network protocols, primarily based on feature mutation and fuzz testing, encounter drawbacks including reliance on expert experience and incapacity to address unknown protocols. This work concentrates on the automated analysis and formulation of vulnerability detection algorithms to tackle the issues of vulnerability detection for digital twins in IIoT protocols. A method for detecting network protocol vulnerabilities is proposed, utilising a combination of Generative Adversarial Networks (GANs) and mutation algorithms. A network protocol analysis model utilising GANs is employed to thoroughly extract information from message sequences, delineate message formats and associated attributes, and ascertain the structure of the network protocol.