Detection of network attacks in software-defined networks using isolating forest algorithm

This paper proposes an approach for detecting network attacks in software-defined networks. The specifics of such networks in terms of security are taken into account, and a modified isolating forest algorithm is taken as the basis for the developed approach. The results of experimental studies where the optimal parameters of the isolating forest algorithm and the extended algorithm of the isolating forest are chosen are presented. Based on the results of the studies, a conclusion is made about the effectiveness of the isolating forest for network attack detection in software-defined networks