Application Research Of Speaker Recognition On Identity Authentication

In the information age, information security has an important position. On par with nuclear technology and space technology, it has become one of the three backbone technology in the new era. A nation without information security, there is no political, economic, and military security. Network, as the main carrier of information, whose security is the core of information security. Network security provides protection for the country’s information and communication security, identity authentication technology is the first gate of the network security.Identity authentication technology based on biological characteristics to provide a more simple and convenient than traditional authentication technology, which avoids many people to remember complex passwords, and also avoids the USB Key and other physical certificates to be copied or stolen. Voiceprint in identity authentication based on biological characteristics has long been left out than DNA and fingerprints because its recognition rate is not high, it is difficult to build a one-to-one correspondence between voiceprint characteristic parameters and speaker. However, as speech recognition technology booming, voiceprint recognition technology is in the ascendant, which breaks the embarrassing situation for the identity authentication based on biological voiceprint, identity authentication based on speaker recognition will usher a bright prospect.This paper through the systematic research to speaker recognition technology, on the basis to improve traditional DTW identification algorithm, designs a text-dependent speaker identification system, combining with hardware and software to test its usability. Finally, with the introduction of a trusted third party, this paper applies speaker recognition to communication authentication and proposes two authentication schemes based on speaker verification which is safe and effective.