Research on Voip-Over- Vpn s Classification And Mitigation

Abstract

With the development of IP technology, VoIP(Voice over IP) service is the most popular voice communication mode in the world. At the same time, more and more individual and enterprise users use VPN s to encrypt transmitted content to protect communication privacy. However, the operators and CSPs(Carrier Service Provider) are interested in their subscribers who used VoIP-Over- Vpntechnology, since it will result in their huge revenue loss while the subscribers escaped the high voice charging. To solve this problem, this paper develops a mechanism to detect and mitigate the VOIP-over- VPN traffic to protect the revenue for the telecommunication industry. In this paper, an advanced packet matching engine is introduced to extract traffic packet features and flow features from encrypted VOIP traffic, and input them into an intelligent detection engine based on heuristic algorithm to identify and classify voIP-over- VPN traffic. Based on the identification and classification results, an Action engine is introduced to block, shape, charge or report the event records. Through plenty of experiments and practices, the VOIP-over-VPN detection and mitigation solution proposed in this paper can achieve more than 95% detection accuracy, and can be alleviated according to different mitigation policies.