Automatic Classification of Attacks on IP Telephony
Jakub Safarik, Pavol Partila, Filip Rezac, Lukas Macura, Miroslav Voznak
DOI: 10.15598/aeee.v11i6.899
Abstract
This article proposes an algorithm for automatic analysis of attack data in IP telephony network with a neural network. Data for the analysis is gathered from variable monitoring application running in the network. These monitoring systems are a typical part of nowadays network. Information from them is usually used after attack. It is possible to use an automatic classification of IP telephony attacks for nearly real-time classification and counter attack or mitigation of potential attacks. The classification use proposed neural network, and the article covers design of a neural network and its practical implementation. It contains also methods for neural network learning and data gathering functions from honeypot application.