What question did this study set out to answer?

To develop a framework that detects network traffic anomalies effectively by addressing limitations of traditional methods.

Network abnormal traffic behavior detection and feature adaptive extraction method based on graph neural network

View Full Paper

Ask AI

Bookmark

Authors

This framework detects traffic anomalies in networks, showing improved feature extraction and accuracy implications.

To develop a framework that detects network traffic anomalies effectively by addressing limitations of traditional methods.
Proposes a Dynamic Graph Neural Network for adaptive traffic anomaly detection.
Introduces four core modules: dynamic diagram construction, spatial feature extraction, temporal feature extraction, and anomaly detection.
Employs time series sliding window and graph attention mechanisms for dynamic graph construction.
Utilizes hierarchical graph convolution for enhanced feature representation.
Achieves an F1-Score of 93.80% on the CIC-IDS-2017 data set, outperforming comparison methods.
Demonstrates effectiveness in capturing network behavior evolution over time.
Ablation experiments confirm the importance of each core module.

Huang et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark