Intelligent detection method for malicious traffic of Internet of Things system and application

• Main Authors: LIN KUNDA, XU XIAOLONG
• Format: Patent
• Language: Chinese; English
• Published: 08.10.2021
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; ELECTRIC COMMUNICATION TECHNIQUE; ELECTRICITY; HANDLING RECORD CARRIERS; PHYSICS; PRESENTATION OF DATA; RECOGNITION OF DATA; RECORD CARRIERS; TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHICCOMMUNICATION

The invention relates to an intelligent detection method for the malicious traffic of an Internet of Things system and application. The method comprises the following steps: designing and applying brand new standardization processing for network traffic, obtaining a vectorization result, and extracting a vectorization result by taking each feature extraction network corresponding to each preset vector type as input; enabling the input to sequentially pass through a to-be-trained network designed and formed by a multi-level feature connection layer, a fusion layer and a classification layer, executing training to obtain a malicious traffic detection model, i.e., realizing the malicious traffic detection of target network traffic by using the model. In the whole design scheme, the time sequence features, the short-time statistical features and the byte features of the traffic are fused, so the detection model is more powerful than other models, the experimental performance is better, and the robustness is highe