Skype traffic identification based on flow characteristics
Hongzhi Liu · Caai Transactions on Intelligent Systems · 2010
Most of the Skype traffic identification models are limited to Skype communication mechanisms and static payload characteristics.No net flow characteristics are considered in identification algorithms.To overcome this limitation,a hierarchical Skype traffic identification model based on naive Bayesian classification was developed.Flows were analyzed according to network connection modes.Results were then obtained according to real-time flow characteristics,such as packet size,average inter-packet gap and burstiness of bandwidth consumption.The validity of the algorithm was proven by testing conducted on the Beijing China Unicom backbone network.