Classification of Peer-to-Peer Traffic Using Neural Networks

摘要

We present applications of data mining, and in particular, neural networks, to classify Peer-to-Peer (P2P) traffic in IP networks. we captured Internet traffic at a main gateway router, performed pre-processing on the captured data, selected the most significant attributes, and prepared a training data set to which the neural network algorithms were applied. We built several models using a combination of various attribute sets for different ratios of P2P to Non-P2P traffic in the training data. We observed that the accuracy of the model increases significantly when we include the attributes "Time", "Src IP addr" and "Dst IP addr" in building the model. By detecting sequence of packets and communities of peers, we achieved classification accuracy of higher than 95%. Consequently, we recommend that: (a) the classification must be done within the authority of the Internet Service Providers (ISP) in order to detect communities of peers, and (b) the model needs to be frequently trained to ensure the correctness of the classification algorithm. Our approach is based only on information in the IP layer, eliminating the privacy issues associated with deep packet inspection.