Fuzzy logic for the detection of anomalous behavior in IP telephony systems

Abstract

This research exposes the main frauds and attacks detected according to the literature review, in the IP telephony infrastructures. The data mining techniques (Data Mining), most used in recent years, are explained and used to detect anomalous behavior in data networks, such as Fuzzy Logic techniques. In addition, several studies and experiments were carried out with the Open Source data mining tools that are used worldwide, such as the WEKA -Waikato Environment for Knowledge Analysis- and the KEEL-Knowledge Extraction based on Evolutionary Learning-, combining several algorithms of fuzzy logic with others belonging to Decision Trees and Decision Rules, which experimentally showed the algorithms to be implemented for the detection of anomalous behaviors.for the realization of the experiments, traffic samples, of two types, according to the raised by the literature.