A live broadcast scene-oriented encrypted video traffic identification method, system, device and storage medium
By combining Sr-Sketch and causal inference models with DNS packet analysis, the problem of identifying encrypted live video traffic was solved, achieving efficient and accurate video type identification.
CN116647716BActive Publication Date: 2026-06-12UNIV OF JINAN
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- UNIV OF JINAN
- Filing Date
- 2023-05-30
- Publication Date
- 2026-06-12
AI Technical Summary
⚠Technical Problem
Existing technologies struggle to effectively identify and differentiate encrypted live video traffic, especially in live streaming scenarios. Traditional methods are costly or rely on high-quality training datasets and cannot achieve fine-grained video scene recognition.
⚗Method used
By employing the Sr-Sketch data structure and peak point features combined with a causal inference model, and through DNS packet analysis and traffic feature extraction, we can identify live video traffic and determine its type.
🎯Benefits of technology
It enables early and rapid identification and accurate classification of encrypted live video traffic, improves identification efficiency, overcomes the dependence on the quality of the training dataset, and can identify specific live video types.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN116647716B_ABST
Abstract
The application belongs to the field of network security and management, and provides a live broadcast scene-oriented encrypted video traffic identification method, system, device and storage medium. The method comprises the following steps: identifying the DNS request domain name of the obtained DNS data packet to obtain traffic data belonging to the live video type domain name; performing a matching operation on the domain name of the traffic data and the domain name in the prior knowledge base; if the traffic data can be matched, the traffic data is live video stream data, and the stream level features of the traffic data are extracted; otherwise, the persistence and frequency of the traffic data in each period are calculated, the importance of the traffic data is calculated according to the persistence and frequency of the traffic data in each period; when the importance in the final period is higher than the importance threshold, the traffic data is regarded as live video stream data, and the stream level features are extracted; and the video type is identified based on the stream level features and a causal reasoning model.
Need to check novelty before this filing date? Find Prior Art