南华大学学报（自然科学版）2025，Vol.39Issue(2)：42-51,10.DOI:10.19431/j.cnki.1673-0062.2025.02.006
基于内存增强特征提取与时序预测的网络流异常检测
Network Traffic Anomaly Detection Based on Memory-augmented Feature Extraction with Temporal Prediction
周萍 1王超 1尹慰民 1马勇1
作者信息
- 1. 南华大学电气工程学院,湖南衡阳 421001
- 折叠
摘要
Abstract
To address the issue of memory pollution in network traffic anomaly detection,a new unsupervised network traffic anomaly detection model(memory-augmented detection for temporal network traffic,MDTN)is proposed,which combines a Transformer-based feature extraction module,a memory module,and a prediction-based temporal-dependen-cies extraction network.The memory module employs FIFO(first in first out)memory re-placement and KNN(k-Nearest Neighbor)strategy to enhance the generalization ability of the model and robustness to memory poisoning.The anomaly scoring method fuses recon-struction error and prediction error,and it enlarges the gap between normal and abnormal data.Evaluation results on four real network traffic datasets show MDTN over existing state-of-the-art baseline methods on AUC-ROC and AUC-PR.关键词
网络流异常检测/基于Transformer特征提取/内存网络/时间依赖性提取Key words
network traffic anomaly detection/transformer-based feature extraction/memory networks/temporal-dependencies extraction分类
信息技术与安全科学引用本文复制引用
周萍,王超,尹慰民,马勇..基于内存增强特征提取与时序预测的网络流异常检测[J].南华大学学报（自然科学版）,2025,39(2):42-51,10.
