Behavioural Approach to Network Anomaly Detection for Resource-Constrained System – Presentation of the Novel Solution – Preliminary Study
Mariusz Pelc, Dawid Galus, Magda Żołubak, Štěpán Ožana, Wojciech Chlewicki, Katarzyna Cichoń, Michał Podpora, Aleksandra Kawala‐Sterniuk · IFAC-PapersOnLine · 2019
In this paper some preliminary results of network anomaly detection for resource-constrained systems have been presented. All described algorithms are embedded into a three-tier system application. The presented solution is a response to nowadays trends in computer/network security domain that indicate gradual shifting from signature-based anomaly detection typically based on some selected statistical measures toward systems that are based on behavioural analysis of computer systems and/or networks. The presented system requirements, explain relevant terminology as well as we include some selected results and their analysis. The solution is potentially destined for very resource-constrained systems (IoT specifically), all tests were carried out in an example network to additionally analyse their effect of the presented approach on the computer resources.