Skip to Main Content
In this paper, we propose one method to transmit any kind of possible information in an acceptable rate through speech channels while people are talking on the mobile or fixed phones. The paper gives detail description of data embedding and decoding, the information to be transmitted is embedded into the echo of the original speech at the transmitting end, and is decoded through cepstrum analysis at the receiving end. Compared to echo watermarking technologies, the method would have more advantages:(1) By segments division, the transmitting capacity can be improved;(2) The original speech signal will not be corrupted by the echoes;(3)The method do not need complex Error Control strategy. Further, the paper gives the properties analysis: Correct Decoding Probability (CDP), Data Rate, and the effect of front and rear noises. The simulation results show the methods can embed data with any length into speech signal. Finally, parameters selection and applications are also discussed in this paper.