Abstract: In order to improve the detection rate of the intrusion detection system and reduce the false alarm rate,the misuse detection technology and anomaly detection technology were combined to overcome the single technical defect,and the improved evolutionary neural network was taken as a detection engine.Firstly,the genetic algorithm was improved to overcome the defect of the real-code poor global optimization,reduce the complexity of computation,and improve the speed of genetic algorithm evolutionary convergence.The combination of improved genetic algorithm and BP neural network LM algorithm further overcome the defects of slow training and being easy to fall into local optimum in the learning phase of neural network.Thereby,the capabilities of the neural network classification and pattern recognition increase.Using KDDCUP99 dataset as training and test data sets,experimental results show that the intrusion detection hybrid model based on evolutionary neural network can achieve significant improvement in the extraction speed of data feature rules,detection accuracy and recognizing new types of attacks.

Key words: Intrusion detection,Misuse detection,Anomaly detection,Genetic algorithm,Evolutionary neural network