Research And Application Of Traffic Identification Technology On IP Network

With the wide application of multimedia and P2P, technologies of network traffic identification and control are becoming a hotspot in network traffic management area. It is guaranteed that key network flow can run smoothly on network by network traffic control technologies. Network traffic identification technologies support essential conditions for network traffic control by traffic identification, packet classification and so on. Technologies of identifying different network traffic and classifying network packets have been researched in this dissertation by effective packet classification algorithms.Different technologies of network traffic identification and packet classification algorithms have been researched in this dissertation based on domestic and overseas results of researching in this area. These technologies have been applied to classic network traffic identification system. The main works done in this dissertation are as the following:1.Protocol data unit,coming from different layers in TCP/IP architecture, has been analyzed and studied. Traffic identification technologies on different layers have been summarized. Five kinds of classic P2P traffics have been analyzed through deep content inspection technologies on application layer. They can be identified by their traffic character signatures, which have been discovered.2.Classic packet classification algorithms have been studied in this dissertation. EABV algorithm is proposed under the requirement of traffic identification technologies. This algorithm takes the advantage of ABV algorithm in classifying packets in large scale ruleset, extends the function in updating ruleset dynamically. It is more appropriate to apply EABV algorithm in network identification environment where all kinds of traffic change in high frequency.3.A prototype system of BT traffic identification has been designed and implemented based on netfilter framework. EABV algorithm is applied in the system. The signature of BT traffic has been discovered by deep content inspection technology, then BT traffic can be recognized. In this system, the difference between EABV algorithm and time linear algorithm is compared. The correctness of EABV algorithm is proved, and the efficiency of EABV algorithm is validated by the experiment.