The Stability of the Internet Traffic Features
Bin Zhang, Yuandong Mao, Mei Zhang, Yun Yu, Guoquan Jiang, Bo Deng · Advances in intelligent systems research/Advances in Intelligent Systems Research · 2016
In this paper, we present a statistical analysis of traffic features at the packet level.We show that all traffic features demonstrate similar approximately power-law distribution for different time and interval at minute time scale except for the packet size.We observe that feature entropy and independent feature symbols number for fixed packets number are relatively stable in a short time interval, which is very useful for traffic anomaly detection.