A SIP Security Testing Framework

Session Initiation Protocol (SIP) has emerged as the predominant protocol for setting up, maintaining, and terminating Voice over Internet Protocol (VoIP) sessions. In spite of the security mechanisms that it offers, several attacks are being made on the SIP architecture. In this paper we take a proactive approach and highlight the importance of testing SIP from a security perspective. We first give a brief introduction to some of the most common attacks on SIP. We then describe a framework to effectively test several security aspects of a SIP network and thereby help mitigate such attacks. We also present a genetic algorithm that we developed and used to generate data in our fuzz testing. Finally, we present the results of some tests performed on popular SIP devices using our framework.