Design And Implementation Of Network Traffic Identification Analysis System

With the rapid development of Internet technology and network applicationservices, network management, maintenance and inspection technology has broughtgreat difficulties. Although associated with the management of networking technologyis constantly improving, but in the network control, management, analysis and designpeople get more and more difficult, so it is necessary to research the sources ofnetwork traffic and the characteristics. Identification and analysis of the network canhelp companies quickly and efficiently find the source of network traffic andcharacteristics, thus providing the basis for the maintenance and adjustment of thenetwork.This paper first introduces the research background and significance, then reviewthe current research. In Chapter II describes the outline of a typical trafficidentification technologies, including identification technology based ports, loadcontent recognition technology based on data reported traffic, identificationtechniques based on the statistical characteristics, traffic identification technologybased on machine learning. in Chapter III introduces association rule miningtechniques, in particular the classic association rule mining Apriori algorithm, analysisthe advantages and disadvantages of Apriori algorithm, because of Apriori algorithmis high compute complexity, propose an improved Apriori algorithm. This article alsodescribes the needs of the system, and analysis the system test results, the final workof this paper summarizes the work carried out in the next prospect.The purpose of this paper is to design a network traffic identification analysissystem, on the one hand, the system monitors network traffic through, we will be ableto grasp the real-time operation of the entire network; on the other hand, statistical analysis of the results of the system through different stages, we can operationalefficiency of network resources, network performance, and more comprehensiveunderstanding the network configuration.We chose.Net+SQL Server2008to develop technical solutions, using C#asthe development language, identification technology based on the use of packet trafficload content to designed a prototype system for an insurance company. The prototypesystem also uses improved algorithms based on Apriori algorithm for data mining,data extraction and analysis the characteristics, this can help businesses find thesources and characteristics of network traffic, providing a solid theoretical basis to thecompany’s network maintenance and adjustment. The test results of the system showthe Network Traffic Identification Statistical Analysis System has a higher recognitionrate, can effectively tap the inherent characteristics of network traffic, and has strongstability and robustness, conform the design requirements.