Publication

Abstract : Intrusion detection is now a significant part in computer and network security. Various intrusion detection approaches are presented to secure the network, but the performance of the system is reduced. Thus, to improve the detection rates and decrease false alarm rates in intrusion detection is important. The crux of an efficient intrusion detection system is its ability to differentiate between normal and potentially harmful activity. Earlier, developers had used coded rules and blocking specific activities for safeguarding the system. However, in view of the current and future threats, automated and adaptive detection systems are required to safeguard the system. In this paper, an adaptive intrusion system is proposed based on Artificial Immune Systems (AIS). The AIS is based on the Human Immune System (HIS). HIS can detect and defend against harmful and previously unknown invaders, so an Intrusion Detection System (IDS) based on the same principles is proposed. The KDD-cup dataset is used that is a benchmark for evaluating the security detection mechanisms. The Principal Component Analysis (PCA) is applied to transform the input samples into a new feature space.