A hybrid intelligent intrusion detection system to recognize novel attacks

We propose a hybrid intelligent intrusion detection system to recognize novel attacks. Current works in intrusion detection solve the anomaly detection and the misuse detection. The misuse detection cannot recognize the new types of intrusions; while the abnormal detection also suffers from the false alarms. The mechanism to detect new forms of attacks in the systems will be the most important issue for intrusion detection For this purpose, we apply the neural network approach to learn the attack definitions and the fuzzy inference approach to describe the relations of attack properties for recognition This study concentrates the focus on detecting distributed denial of service attacks to develop this system. Experiment results will verify the performance of the model.