An Architecture for Network Intrusion Detection System based on DAG Classification

Abstract: Intrusion detection is an effective approach of dealing with problems in the area of network security. Rapid development in technology has raised the need for an effective intrusion detection system as the traditional intrusion detection method cannot compete against newly advanced intrusions. In this paper we proposed a feature based intrusion data classification technique. The reduced feature improved the classification of intrusion data. The reduction process of feature attribute performs by DAG function along with feature correlation factor. The proposed method work as feature reducers and classification technique, from the reduction of feature attribute also decrease the execution time of classification. For evaluation purposes, this model is applied to KDD �99 dataset.

Keywords: Network Intrusion Detection System, Directed acyclic graph, Classification, KDD�99 Data set, Support Vector Machine (SVM), Ensemble Technique, Neural Network.