Anomaly Detection

Time Series for Anomaly Detection The file is a Matlab data file. It contains 3 time series, representing the packet rate of 3 different traffic traces, related to inbound traffic of the UNINA Network. The traces were collected in year 2004. The packet rate was sampled with a period of 2 seconds and each trace lasts 2 hours. These data have been used for studies on volume-based anomaly detection and are related to time intervals during which no anomalies were observed on the UNINA network by the NOC operators. In other words, they can be considered anomaly-free. When refering to our Anomaly Detection Dataset, please cite the following reference: A. Dainotti, A. Pescapè, G. Ventre, "A cascade architecture for DoS attacks detection based on the wavelet transform", Journal of Computer Security, Volume 17, Number 6/2009, Pages 945-968.

用于异常检测的时间序列数据集（Time Series for Anomaly Detection）：该文件为MATLAB数据文件，包含3组时间序列，分别对应UNINA网络入站流量的3种不同流量迹线，表征其数据包采样速率。该批流量迹线采集于2004年，数据包速率以2秒为采样周期，每组迹线的持续时长为2小时。此类数据曾被应用于基于流量体量的异常检测相关研究，其对应的时间段均为网络运营中心（Network Operations Center，NOC）运维人员未在UNINA网络中观测到异常的时段，换言之，这些数据可被归类为无异常样本。若需引用本异常检测数据集，请援引以下参考文献：A. Dainotti、A. Pescapè、G. Ventre 发表于《计算机安全期刊（Journal of Computer Security）》2009年第17卷第6期、第945至968页的论文《基于小波变换的DoS攻击检测级联架构》。