Anomaly-based online intrusion detection system as a sensor for cyber security situational awareness system

Almost all the organisations and even individuals rely on complex structures of data networks and networked computer systems. That complex data ensemble, the cyber domain, provides great opportunities, but at the same time it offers many possible attack vectors that can be abused for cyber vandalism, cyber crime, cyber espionage or cyber terrorism. Those threats produce requirements for cyber security situational awareness and intrusion detection capability. This dissertation concentrates on research and development of anomaly-based network intrusion detection system as a sensor for a situational awareness system. In this dissertation, several models of intrusion detection systems are developed using clustering-based data-mining algorithms for creating a model of normal user behaviour and finding similarities and dissimilarities compared to that model. That information can be used as a sensor feed in a situational awareness system in cyber security. A model of cyber security situational awareness system with multisensor fusion capability is presented in this thesis. Also a model for exchanging the information of cyber security situational awareness is generated. The constructed intrusion detection system schemes are tested with different scenarios even in online mode with real user data