Research Of Intrusion Detection System Based On Multi-Agent Of Radial Basic Function Neural Network

People's daily life and work increasingly depend on computer network and information technology because of their fast developments, and the requirement of the network security also becomes higher and higher. But there are a plenty of the current protection technologies of the network security, such as firewall, access controlling and data encryption and so on, all these technologies can not entirely protect the network security and prevent the hacker's attack. So, researching and developing new intrusion detection system, which can initiatively and dynamically protect the network, invalidly become a new primary direction of the network security.Neural network's outstanding abilities of parallel computing, non-linearity, self-adaptive, handling distortion data, anti-jamming, make itself a very significant tool applying in the intrusion detection field. BP network, one of the most widely used neural networks, is already applied in the intrusion detection field. In comparison with BP network, RBF neural network is not an error back propagation method but an entire front propagation method. Its training time is short and it is not easy to converge to the local minimum. The function approximation ability, pattern recognition and classification ability are better than BP network. Therefore intrusion detection system based on RBF will overcome shortcomings of the traditional intrusion detection system.The agent (Agent) is a function entity to feel the changes of environment and react to the changes. The agent's advantages above make it become a new research area of intrusion detection technique while applying to intrusion detection system. In this article, the author proposed intrusion detection system model based on Multi-Agent of radial basic function neural network through applying to two technology in the intrusion detection field, because of these advantages of radial basic functions and agent technology. Module of this model is agent based on RBF neural network and every agent can cooperative each other. Its intelligence, distribution, real-time response and interactive features provide comprehensive network defense and systems protection for users. In this article the author deeply research structure design, algorithm improvement, and emulator test based on agent of neural network. This model has the good usability and the extendibility and is an effective method of developing security management system.