| Summary: | 碩士 === 中華大學 === 資訊工程學系(所) === 96 === As Voice over IP (VoIP) is gaining popularity, VoIP has become an important communication tool in the present time. Thus, the security problem of Spam over Internet Telephony (SPIT) will become an important issue. SPIT cannot be treated as the same way of anti-spam problem, since the SPIT is a real-time service; it is more difficult to guard the SPIT than anti-spam. Besides, the lower cost of VoIP communication combined with traditional phone fraudulence will bring more serious social problems. In order to solve the SPIT problem, we proposed an effective SIP-based spam detection system (Social Network based SPIT Detection System；SNSDS). In proposed architecture, we adopt the idea of the relationship of trust between peoples, and integrated the six degrees of separation with the concept of the social network. Also, we apply breadth-first algorithm to search and calculate the confidence degree of caller, and compared with a predefined threshold to decide whether the caller is a legitimate VoIP Callers or no. In the experiment, we can find that the proposed architecture has better performance compared with the method presented by Ram.
|
|---|
