Abstract:

The insufficient of current traffic identification technique was analyzed, and then a heuristic traffic identification method based on trusted list was proposed. This method uses a “memory” technique, adds the discerned connection into a trusted list. Moreover, a real-time optimized method was developed to make sure that the more active connections can be measured to primarily in the trusted list all the time, thus will accelerate the traffic identification speed and improve the accuracy. The experiment results show that this method can identify the encrypted traffic perfectly, and more effectively than existed methods.

Key words: peer-to-peer, traffic identification, trusted list, deep packet inspection