Research On Intrusion Detection System Based On Wavelet Neural Network

With the development of the Internet,there are high demands for internet applications. The network services, such as E-Bank and E-Commerce arebecoming the part of life. And Network attacks are increasing. People have realized the importance of network security.As a kind of active measure of Information Assurance,IDS acts as an effective complement to traditional protection techniques.The dynamic security circle,including poicy,protection,detection and response,can greatly contribute to improving the assurance ability of information systems and reducing the extent of security threats.Network Intrusion Detection System is becoming a great developing direction.The IDS has a lot of intelligence,it correlates artificial intelligence closely.It is the combination of computer science,mathematics and so on.It includes many advanced thchniques of current computer field,such as data mining,neural network,etc.So the research for IDS means the research for other techniques too.Utilizing the ability of time-frequency analysis of the wavelet transform in signal processing and approximation of the neural networks towards any nonlinear function, a method of intrusion detection based on wavelet neural network is proposed.Replacing ordinary neural network activation functions by wavelet transform,the method can effectively improve the efficiency and speed of the network training samples.In the simulation,it reflects a good convergence and learning ability.This method is proved to be suitalbe for Intrusion Detection System. Through Combining A Wavelet Neural network with SNORT technology,they exchange what one has for what one needs, it can make the intrusion detection system have more perfect functions than singe detection way,in that,not only distinguish exactly the kind of intrusion,but also have a detection capability to the unknown part. The conclusion shows this new intrusion detection system model have a further value in research and application.