Multisensor Fusion for Intrusion Detection and Situational Awareness
Abstract
Cybercrime damage costs the world several trillion dollars annually. And al-though technical solutions to protect organizations from hackers are being con-tinuously developed, criminals learn fast to circumvent them. The question is,therefore, how to create leverage to protect an organization by improving in-trusion detection and situational awareness? This thesis seeks to contribute tothe prior art in intrusion detection and situational awareness by using a multi-sensor data fusion. The model for multisensor data fusion system incorporateshuman cognition reasoning into a hybrid multisensor fusion, i.e. vertical fusion,horizontal fusion within a network segment, and horizontal fusion between thenetwork segments. The proposed model is able to reduce false positive alarmsfor intrusion detection, improve the detection of unknown threats, and provide coverage for the whole cyber kill-chain.