Systematic Review and Comparison of Anomaly Based Network Intrusion Detection Systems Based on Efficiency
Uma Subramanian, Hang See Ong · 2013
Currently, anomaly based network intrusion detection (ANID) is the solution for novel and sophisticated attacks. This review focuses on the comparison of Anomaly based Network Intrusion Detection Systems (ANIDS) based on efficiency. A collection of ANIDS that were trained and tested using KDD cup 99 dataset in the period 2002 to 2012 (May) are considered for this review paper. A total of 258 papers were reviewed.Among the ANIDS using kdd’99 dataset 70.58% are machine learning based. It is observed that the fuzzy based ANIDS constitute 37.5 %. It is concluded that Fuzzy based ANIDS gave good Detection Rate (DR), and SVM based ANIDS gave good False Alarm Rate (FAR) consistently based on the efficiency comparison performed.