Research And Implementation Of Identification For P2P Streaming Media Traffic

In recent years, P2P (peer-to-peer), as a bran-new technology, has developed whip and spur. Interrelated investigation indicates that P2P traffic has taken up 60% to 80% of the total Internet traffic stealthily, and it has become the killer application of Broad Band Internet. P2P operation has been increasing constantly, so it consumes much network bandwidth, even arouses congestion of network and reduces the performance of other operations. On the other hand, for the enterprise, inner employees use enterprise network to listen music or watch P2P Streaming Media on the internet, this will make tremendous waste of network resource. Further more, using unsafe network environment to get P2P applications may open up a back door to the enterprise network, make viruses and baleful codes be able to stand aside safety check up and enter into the corporation. Therefore, effective identification of P2P traffic is very necessary and takes a very important position in the network management.There are many kinds of P2P applications, and among them there are 2 kinds of applications which occupy the most network bandwidth. They are P2P file-sharing applications and P2P streaming media applications, and the latter developed so rapidly in the recent 2 years. With regard to the former, many researchers have made a lot of fruitful research; But for the latter, since it emerged just in 2005, so there are so limited research about it. This paper will explore and research how to identify the traffic of P2P streaming media.Because the current P2P streaming media softwares are all commercial, so their protocols are not open, we can not understand their communication principles and implementation detail through their protocols. But since P2P streaming media technology is the result that combines the P2P network communication technology and the streaming media technology, so if we want to identify the traffic of P2P streaming media, we must understand the application principles and the identification methods of the traditional P2P software. In Chapter 2 of this paper, it is described that the principle of P2P streaming media, its current research situation, the detection technology of P2P's traffic and so on. From that we can know the P2P traffic recognition technology, which has been most extensively applied in commerce, is based on application layer signature.If the recognition technology based on the application layer signature is used, the first step is to find out the signatures. But all the P2P streaming media softwares are commercial, and their protocols are not open to public, so it's hard to find out their signatures. In Chapter 3, 5 known P2P file-sharing protocols are first analyzed, and several common characters about their application layer signatures are summ up. This can guide the exploration to the application layer signature of P2P streaming media softwares with unknown protocols. From some sniffing experiments and the summarized common characters, in the last part of Chapter 3, an automatic excavation algorithm for application layer signature is put forward. And this algorithm achieved positive experimental results.However, because of the limitation of this automatic excavation algorithm and the defect of application layer signature recognition technology, the experiment result in Chapter 3 was not enough. In Chapter 4, in order to make up for the shortfall of application layer signature recognition technology, a "Traffic Recognition Algorithm of Streaming Media Based on Connective Characters" is put forward, which can identify P2P streaming media software by connective characters. The algorithm need not test the special content in the payload, so the problem has been relieved a lot that the signatures are hard to find out. In the last part of Chapter 4, this recognition algorithm is validated through an experiment in an actual network environment. The Chapter 3 and Chapter 4 are the main part of this paper, and the later is more important.