Towards self adaptive network traffic classification

摘要: Abstract A critical aspect of network management from an operator’s perspective is the ability to understand or classify all traffic that traverses network. The failure port based classification technique triggered interest in discovering signatures on packet content. However, this approach involves manually reverse engineering applications/protocols need be identified. This suffers problem scalability; keeping up with new applications come everyday very challenging and time-consuming. Moreover, traditional developing once using them different networks low coverage. In work, we present a novel fully automated payload content (PPC) system addresses above shortcomings. Our learns application where desired. Furthermore, our adapts as for changes. Based real traces several service providers, show capable detecting (1) tunneled wrapped applications, (2) use random ports, (3) applications. it robust routing asymmetry, important requirement large ISPs, has high precision (>97%). Finally, easy deploy setup performs real-time.