一种大模型智能体网络行为监测识别方法、装置、设备及介质
By capturing network traffic, reconstructing sessions, and extracting features, and combining clustering algorithms to reconstruct the task chain, the problem of monitoring the network behavior of large-scale intelligent agents in encrypted communication scenarios is solved, achieving efficient and accurate identification and tracking, and meeting the needs of network security management.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2026-04-21
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies cannot effectively monitor the behavior of large-scale intelligent agents in encrypted communication scenarios, resulting in high operational complexity, high computational consumption, and easy leakage of user communication privacy, making it difficult to meet the needs of network security management.
By capturing network traffic, reconstructing network sessions, extracting protocol semantic fingerprints, flow-level dynamic rhythm features, and intent-driven behavior graphs, and using clustering algorithms to reconstruct the entire task execution chain, we can achieve accurate identification and tracking of network behavior of large-scale intelligent agents and avoid decryption operations.
It improves the accuracy and completeness of network behavior monitoring of large-scale intelligent agents without decrypting network communication content, adapts to monitoring needs in encrypted communication scenarios, avoids privacy leaks, and provides clear monitoring and identification results.
Smart Images
Figure CN122069118B_ABST
Abstract
Citation Information
Patent Citations
Encrypted traffic application activity identification method based on wavelet transform
CN111626322A
Time sequence graph construction method and device, equipment and medium
CN114547491A