Audio-Signal Authenticating System Based on Asymmetric Signature Schemes

Abstract

In this paper, we propose a simple and highly secure signature scheme to protect the copyright of digital-signal. Employing the masking property of human ears that most of people can't hear the signal of frequency of 1-20Hz, we use the encoding scheme to change the energy of frequency to conceal a signature. The embedded signature values will not influence the quality of original audio-signal. We can raise the integrity of audio-signal because the scheme can definitely figure out if a signal is tampered or not. In designed method, we extract a characteristic value M from the audio-signal, then apply the RSA signature scheme to transform the characteristic into signature S and conceal it in audio-signal. In the requirement of verifying copyright, our scheme retrieves the signature S and uses RSA signature scheme to decrypt it and obtain a characteristic value M . By comparing the value M with the original M , if the signal was tampered, the characteristic M and the original M should be distinct; otherwise, they are the same. In order to authenticate the audio-signal, the proposed method can independently retrieve the characteristic value M quickly without applying the original audio-signal, even the signal was tampered. Therefore, the security of the scheme is improved.