Research On P2P Traffic Identification And Monitor Technology

With the development of computer and network technology, Internet expands in scale constantly, and applications based on Internet have also been developed rapidly. Efficient, scalable P2P technology became the object of extensive research. Network applications based on P2P technology has significantly changed the Internet. The natures of free and open make piracy and illegal content on P2P networks to spread unlimited. With the large number of P2P applications, P2P traffic has become the main flow of Internet. A large number of downloading and video data traffic impedes the normal network operations carried out. Effectively to identify and monitor P2P traffic has been a hot research topic.This paper first describes the characteristics of P2P applications, the P2P network structure development and the necessity of P2P traffic monitoring technology. Secondly,after the review of the development of P2P traffic identification technology, this paper gives special introduction to identification technology based on packet content and behavioral traits of P2P data traffic.On the basis of time series analysis using autoregressive moving model, identification methods based on behavioral traits of network layer connections and application layer DNS service are given. The P2P download can be determined, if a sudden increase in the number of a peer's connecting IP addresses or the IP addresses which were not obtain by DNS protocol, using Autoregressive Moving model. The Experiments results prove the validity of these methods.Based on the research on key technologies mentioned above, this paper gives the design and realization of a P2P data traffic identification and monitor system. Practical operation effects show that the system has good practical value.