Research And Implement Of IDS

With the development of network technique, more and more life and work of the people depend on the computer network. And because of the flaws and opening characteristic, its security confronts a baptism. Computer Intrusion Detection has been a more important area of the security technology, it is currently a hot theoretical research. Intrusion Detection System can collect and analysis the key information of the computer networks and systems, including the breach of security strategy signs of the incident or attack, then notice the system security administrator.Traditional intrusion detection system uses the pattern-matching approach to the realization of detection, so its' adaptive performance was not ideal. Most of the intrusion detection systems only compare the collect data with the security policies mode to found the attack. This pattern matching method can do efficiently in the known intrusion mode, but do badly in some of the latest unknown attacks. Using the data mining technology to extract data from a large number of networks connecting can update itself continuously through the studying of the invasion. Intrusion Detection is now hot issue research, it has great theoretical and practical value.With the application of further study of the data mining technology in the network intrusion detection system, In this paper, we put forward an improved scalable intrusion detection system, and gives the system model and process, architecture.We added the data pre-processing module in this papaer, used the data mining technology from intrusion rule set so that the rules can be updated automatically. Our test proved that the new model can meet the current network security Intrusion Detection System requirements.It has small changes in the sensitivity for the importation of attacks,and also has better adaptability.Research in this paper:On the basis of understanding the current development and status of the Intrusion Detection, we analyse the advantage and weakness of the traditional Intrusion Detection System. In this paper we mainly introduce the data mining methods and the intrusion detection model based on data mining. Finally we summed up the main thesis of this work and the existing problems, we pointed out the future direction of further study.The author's main research work:July 06---August 07 in De'an Computer Technology Co. Ltd. Jinan City, Shandong Province, for the development of modules within the network design and network security, we developed the network function modules, which has been put into operation now. During the inspection,I read a large number of books and information on the intrusion detection system and performed of a certain model of the study.