An intrusion detection system based on neural network

Intrusion detection system (IDS) is a technology forwarded for guaranteeing the computer system security and thus finding and reporting unauthorized or abnormal phenomenon. Neural network possesses the abilities including self-adaptive, self-organizing and self-learning. Utilizing the capacities such as recognition, classification and induction can make the IDS adaptable to the dynamic changes characteristics of user´s behavior. The paper advocates a network IDS model based on BP neural network. Objective analysis is implemented for the experiment results through the training and detection process. Good results are gained in detection rate, false alarm rate, Omission rate.