The Design And Implementation Of SPIT Prevention Scheme Based On SIP

ABSTRACT:With the introduction of the IMS (IP Multimedia Subsystem), the number of VoIP (Voice over Internet Protocol) users is set to increase. Naturally, SPIT (Spam over Internet Telephony) is expected to become a very serious issue in the next years.SPIT, known as unsolicited bulk calls sent via VoIP networks, is an attack through the multimedia information to do the recipient merchandising, commercial advertising or malicious calls. It causes serious problems to the normal users'life and a waste of network resources. Public services are unable to provide the timely and effective services.The main work of this paper is as follows:Firstly, analyze SPIT prevention technologies of some organizations, such as IETF (Internet Engineering Task Force) and so on. Obviously, several SPIT prevention methods are being proposed, but the relevant research is still at a very early stage. Then, we design an advanced SPIT prevention scheme based on SIP (Session Initiation Protocol), because SIP solutions dominate today's enterprise market. The proposed scheme is applicable for detecting and filtering both machine-initiated and human-initiated SPIT. It is composed of three key mechanisms. The first one tackles first/initial contact problem by applying a discriminator system? based on the principle of Turing Test. The second one designs a SPIT classifier by using the PNN (Probabilistic Neural Network). Inspired by Honeypot, we present the third mechanism which is named Honeyphone. Finally, we implement these mechanisms as a prototype system. Experiment and analysis results show promising performances in terms of completeness, real-time and usability. At the same time, we conduct an analytical survey of already proposed SPIT prevention techniques.To sum up, the scheme's feasibility is validated. These three mechanisms can work separately, but this paper proposes a fusion method to combine the information from these mechanisms to get a more accurate decision.