A Method of Detecting Anomalies in IP Phone Traffic based on Ontology of Voip Messages

The increased use of VoIP technology heightens the potential threats to ordinary users. This study examines the peculiarities of implementing corporate VoIP telephony systems, identifies the main problems in protecting information within such systems, and suggests solutions. Special attention is given to the development of methods for speech analysis and natural language processing, enabling the creation of more accurate and effective systems for detecting anomalous traffic and potentially hazardous communications.An ontology for describing messages in VoIP IP telephony systems is proposed, formalizing main concepts as separate entities and detailing connections between them. The suggested method for detecting anomalies in IP telephony traffic relies on grouping VoIP messages through context-frequency analysis.Experimental studies of the proposed approaches were conducted, resulting in the implementation of a software subsystem for detecting anomalous messages based on the ontological approach in current corporate IP telephony systems.