Network abnormal state monitoring and diagnosing method based on AI cloudization

• Main Authors: RUAN WEIBIAO, WEI BENJIE, AN YUSHUN
• Format: Patent
• Language: Chinese; English
• Published: 09.04.2024
• Subjects: ELECTRIC COMMUNICATION TECHNIQUE; ELECTRICITY; TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHICCOMMUNICATION

The invention belongs to the technical field of network security situation awareness, and discloses a network abnormal state monitoring and diagnosing method based on AI cloudization. Collecting m groups of simulation characteristic data, wherein the simulation characteristic data is corresponding network flow data when a network fault occurs in a simulation process; the simulation feature data are preprocessed, and correlation weighting is carried out on the preprocessed simulation feature data; a network fault diagnosis model for predicting the network state is trained on the basis of the simulation feature data subjected to correlation empowerment, and the network fault diagnosis model is constructed on the basis of GCN; the terminal router inputs network flow data collected in real time into the network fault diagnosis model, and obtains a network state corresponding to the prediction label; the cloud sends the network state to the terminal router; the problem that the abnormal state of the network cannot