| With the continuous development of communication technology,a variety of voice services have brought many conveniences to people’s lives,but also accompanied by security risks such as eavesdropping,information leaking.In particular,the occurrence of the Snowden incident further confirmed the insecurity of communication equipment.However,the existing methods of encrypted call cannot deal with the security threats such as backdoors and viruses in the communication equipment.To solve these problems,encryption and transmission should be separated from each other: the voice signal is encrypted before it enters the communication equipment,so that the voice signal always exists in the call equipment and the voice channel in the form of cipher text.However,due to the presence of the voice-coder in the voice channel,the encrypted signal of the voice signal can not be transmitted directly through the voice channel,but needs to be converted into a signal suitable for the transmission of the voice coder,which has been a problem for many years in the industry.Therefore,it is necessary to study the modulation methods that enable encrypted data flow to be transmitted through the voice channel,that is,to study speech-like modem technology and apply it to secure communication.This paper focuses on speech-like signal generation technology based on speech synthesis and digital modulation.In the speech-like signal generation scheme based on speech synthesis,the low modulation rate is mainly studied,and a speech-like modem method based on LPC model pole mapping is proposed.In the speech-like signal generation scheme based on digital modulation,how to reduce the error rate of the modulation algorithm and improve the transmission rate is mainly studied.Speech-like signal modem methods based on ASK,PSK and multi-mode mixing are proposed,and their transmission performance is analyzed.Finally,a real-time voice secure communication prototype system is designed and implemented,which verifies the feasibility of this scheme.The main work of this paper is as follows:1.The transmission characteristics of Tencent conference voice channel are analyzed.Because there is a coder in the voice channel,the transmission of non-voice signal will produce greater distortion.This paper conducts channel transmission experiments,demonstrates the transmission characteristics of Tencent conference voice channel,and provides a basis for the design of speech modulation algorithms.2.A voice-like modem method based on LPC model pole mapping is presented.On the basis of analyzing the LPC full pole model and the pole distribution of voiced syllables,the data stream is mapped directly to the conjugate pole parameters of the LPC model and the base frequency of the excitation signal.A six-way LPC cascaded voice-like modem scheme is designed,and the parameter range is determined by experiment.Through the actual voice channel transmission experiment,the proposed scheme achieves a transmission rate of 1.65 Kbps and a transmission error rate of 6.72%.Compared with the existing schemes of the same type,the modulation rate is greatly improved,which provides a new idea for related research.It provides a new idea for the research of speech-like modulation based on speech synthesis.3.To further improve the transmission rate,a speech-like modem based on ASK and PSK is studied.Analyses and compares voice-like modem methods based on 2ASK,R-ASK,DBPSK,DQPSK,and their performance.The transmission performance of these voice modulation methods is given through the actual voice channel transmission experiments.4.A multi-mode mixed speech-like modem method is presented.Class-like voice secure communication requires not only a low bit error rate but also a high modulation rate.Based on the advantages of ASK and PSK modulation,a multimode mixed speech-like modem method is designed by combining R-ASK with DBPSK and DQPSK.The maximum transmission rate of this method is 4.5 Kbps and the error rate is 1.86%.When the transmission rate is 3.0 kbps,the error rate is 0.55%.Compared with other similar schemes,this method has obvious advantages in modulation rate and error code performance.5.A real-time voice secure communication prototype system is designed and implemented.Codec2 compression algorithm is used to compress plain voice signal,ZUC-256 stream encryption algorithm is used to encrypt compressed voice data,LFM signal is used as the frame synchronization header signal,Port Audio is selected as the implementation framework,and real-time classified voice confidential calls are implemented based on Tencent conference,which verifies the feasibility of this scheme. |
