Statistical Based Intrusion Detection Framework using Six Sigma Technique

Abstract

Summary This paper presents our statistical based intrusion detection framework for computer networks. This framework uses the six sigma technique to identify the thresholds for the critical network parameters. With the help of raw network data, the thresholds identified are used to differentiate normal, uncertain and abnormal behavior due to network intrusion. This is then used for efficient detection and response. We also present a methodology of six sigma control analysis for intrusion detection in a network. Performance evaluation of our statistical based intrusion detection approach with related intrusion detection approaches conducted using the benchmark DARPA data are very promising.