Characterizing Peer-to-Peer Traffic across Internet1

In this paper we, to the best of our knowledge, for the first time, launch a peer-to-peer network traffic measurement across the Internet backbone in China. Different from the existing studies, our data are derived from core routers on the Internet. Our study focuses on the three periodic peak value groups in the aggregation flow traffic, the heavy-tailed property in the distribution of traffic by different hosts and the long range dependence (LRD) in p2p traffic. Moreover we propose a novel data mining approach, ADTE to differentiate signaling and data traffic. Our study also shows that Napster dominates p2p traffic. In addition, we observe that the aggregation of p2p traffic is less than 1% of the total Internet traffic, much different from the observations in other studies. We analyze the possible reasons and wish that our conclusion could be a good guideline for the software developers.