The Study Of Intelligent Intrusion Detection System Based On Neural Network In Linux

This paper analyses the present research status and significance of intrusion detection technique, after that, discusses the intelligent intrusion detection System based on neural network in the Linux environment. Firstly, it analyses the intelligent intrusion detection technique and its regular mathematical model, and secondly, describe the component modules of system, which includes, data collecting module, data analyses module, data responding module and others assistant module. At last, discusses the design of intelligent intrusion detection system in the Linux operating system.On the basis of the analyzing of the core structure, working mode and the network datagram capture, the main task of data collecting and characteristic analysis in Linux is to design the NIDS data collecting module. The data collecting module reads the configuration files first according to the data which need to be collected, and then read the data from data sources according the configuration files, standardize the collecting data, prepare for the next data processing. At last, do the filter process, if the data meetings the requirements of data receive center then sent it.Otherwise, discard those data and return the beginning read the data source again. After data collecting, system analyses the network protocol in order to finish the feature extraction, prepare the input data for the neural network. The extracted characteristic value input into the neural network, training and test the network. Finally, the whole system has been tested. The test result show that intelligent intrusion detection system for Linux has the very good performance in efficiency and accurate.