A METHOD FOR DETECTING FRAUDULENT CALLS IN TELECOMMUNICATION NETWORKS USING DNA

摘要

The application describes a method for detecting a behavior of interest in a surrounding including at least one telecommunication network, by using approach of Immune Engineering. The method comprises steps of analyzing the behavior of interest by building a characterizing data string called DNA which comprises two or more data sub-strings characterizing fragments of the behavior of interest and called genes. Further, there are selected two or more data fragments called detectors and being substantially close to the genes, and the detectors are applied to identify similar to them data fragments if appearing in information concerning real time activity in the surrounding. Based on the identification results, the method detects whether the behavior of interest is present in the newly incoming information. The method is used to detect fraudulent attacks on telecommunications networks. Analysis is performed, based on "historic" information obtained from usage records databases relating to telephone sessions, faxes, data communication, login sessions, bandwidth usage etc. This is used to classify newly incoming usage call records as "normal" or "fraudulent" behavior. IMAGE