The Research And Design Of Intrusion Detection Based On Neural Network

With more and more site intruded by hackers, security expert found than only use crypt technology to build a security system is not enough. The Intrusion Detection is a new security technology, apart from tradition security protect technology, such as firewall and data crypt. IDSs watch the computer and network traffic for intrusive and suspicious activities, they not only detect the intrusion from the Extranet hacker, but also the intranet users.This paper open with some elemental conceptions and theories of IDS. The paper analyzes intrusion-detection technique in existing IDS models and IDS products, discovers they are limited and hard to meet IDS ' s needs which occupies real-time character, adaptability, accuracy and the ability of self-learning. Then study upon on neural network, the paper finds it is very suitable for the IDS in concept. An intrusion-detection system based on neural network will play a much role in the theory and practical if it can be designed and implemented. And the paper gives a detailed describing to the deducing of BP algorithm and its betterment arithmetic of Levenberg-Marquardt(LM) optimized algorithm.This paper introduces the neural network technology in IDS model, And put forward a detailed design scheme of intrusion-detection model based on neural network. Great emphasis was put in key modules. Lastly according experimental through training and intrusion procedure, we get a fairly analysis, which indicates the neural network has a very great advantage in intrusion detection. Finally, according to the result, the writer put forward some questions and some new ideas.