Application In Campus Network Of Intrusion On Detection System Based On Data Mining

With the continuous development of network technology, many colleges and universities are conducting campus information construction, followed by the campus network security issues become increasingly important. Campus network is characterized by the user traffic flow, Internet a long time, on-line penetration rate in the high complexity of network applications, causing massive access to a variety of servers. Campus network security issues has become a can not be ignored and the letter to be resolved issues. Intrusion detection technology since the early 20th century, 80 has been proposed, after 20 years of continuous development from the initial idea of a valuable study and simple theoretical models, the rapid development of a wide variety of practical prototype system, and in the past 10 years have emerged in many commercial intrusion detection system products, to become the field of computer security an indispensable important security protection technology. At present, the relevant research area of network security has become a hot topic. In recent years, intrusion detection systems are increasingly being used in the network to ensure network security. However, there is currently on the market of commercial NIDS, mostly more complicated, more difficult to grasp, but also more expensive in terms of for the campus network can not bear.The topics for the campus network characteristics and requirements to design and construct a new data mining-based intrusion detection system Snort. Main tasks include: layout of the campus network to install Snort server and software; according to actual demand to configure intrusion detection systems; developed entirely based on the Snort Web applications; the network security problems facing the Snort rule set and add to the rule base; Detection Intrusion Detection system performance.Data mining techniques applied to intrusion detection to intrusion detection as a data analysis process, using data mining techniques to build the system features automatic mode, to improve the accuracy of intrusion detection systems, scalability and adaptability. An improved K-Means algorithm is used to build cluster analysis module plug-ins and pre-detection engine plug-ins.