A Network Traffic Identification Method Based on Finite State Machine

Identifying network traffic accurately and efficiently is crucial for network management and security. Though several identification methods were proposed recently, none of them is shown to be better than all others and recognized as a de facto one. In this paper, an analysis of existing methods is given, and then a Perfect Protocol Finite State Machine (PPFSM) that can describe different identification methods is put forward. Based on these works, we proposed a traffic identification method named Network Flow Sequence Identification (NFSI), which utilizes the characteristics of PPFSM states transformation to identify network traffics. Finally, experiment results of identifying Skype voice traffic showed the accuracy and efficiency of NFSI.