Network Intrusion Detection System Model

Network brings convenience and promptness to human beings, while it also causes risks for information security. As more and more serious problems of network security appears , scientists bring forward the Policy Protection Detection Response model (PPDR model) in the 1990s, in which Network Intrusion Detection System (NIDS) functions as the key part. There are a lot of threatening factors and means in real situations changing quickly so that the mass of the present NIDS can' t change their rules to figure out whether the action is an intrusion. Therefore many NIDS do not work under the practical circumstances for most of the intrusions.According to different demands in various applying fields, several kinds of revision models and methods are proposed to deal with the problem. This paper lists some representative models of NIDS and algorithms which judge if the actions are intrusions and to what degree they adapt to PPDR model, then puts forward a new NIDS theory model-Self Learning network Intrusion Detection System MODEL (SL_NIDS).The main work of this paper can be divided into three parts: (l)This paper puts forward a new NIDS theory model (SL_NIDS) which can update the rules that judge which action is an intrusion, and presents the structure of network in which the model will run, and build the experimentation circumstances. (2)This paper shows a knowledge getting method based on neural network rule extraction that integrates the neural network and expert system, which separates the studying rules and the using rules. This coincides with the internationaltendency that the intrusion rules will be normalized.(3)A model of module is made possible that supervises and analyzes network packets which work between transport layer and network layer. We have inserted an overanxious layer in the existing protocol stack. We can rebuild the network packets and encryption and decryption data and many other important operations of network security in the layer. In addition, that can elevate the efficiency of the whole system.