VoIP中说话人审计的研究与应用计算机科学与技术专业论文.docxVIP

哈尔滨工业大学工学硕士学位论文 哈尔滨工业大学工学硕士学位论文 - - II - Abstract VoIP(Voice over Internet Protocol) is a very popular technology among the current Internet application fields. It is used to manage the real-time delivery of voice information over Internet. VoIP involves sending the compressed voice information in digital form in discrete TCP/IP packets rather than by using the traditional circuit-committed protocols of the public switched telephone network (PSTN). Because IP Network has the potential to support innovation, and like the other new technologies that deliver voice communications over Internet, VoIP improve the access of communications services, and reduce the cost of ownership around the world including many developing countries. VoIP can be applied in Telephone Bank, or used for collaboration by two people who are between long distance. However, it also can become facility for criminal. Law enforcement agencies are authorized to tap and record the suspect voice conversations. Speaker Recognition is a kind of biometric technology that uses an individual’s voice for recognition purposes on the basis of individual information included in speech signals. Combining legal interception on VoIP Speaker auditing can assist security department to investigate crime on VoIP. This paper focuses on researching and designing the method of VoIP speaker auditing. This paper presents an approach to combining VoIP auditing and speaker recognition, and an application of using speaker recognition on VoIP voice data is described. This paper deals with how to obtain VoIP voice data when audio communication: signaling protocol analysis and voice data traffic analysis. After converting the raw voice data to wave files, the speaker recognition can use them to identify the speaker and the result shows it is an effective way. This paper also discusses the influence of CODEC compression. Keywords VoIP, SIP, protocol analysis, speaker recognition, vector quantization - - I - 目 录 摘要 I HYPERLINK