| Wireless communication networks are on the rise all through the globe,It's important to make the most out of this preexisting widespread infrastructure by extending its services and functionalities.By studying the data transmission over voice channel,voice and data can be transmitted over voice channel,which also lays the foundation for the voice confidential communication.Because of the presence of the vocoder on voice channel,data transmitted over voice channel is limited to have the characteristic of voice.Therefore data needs to be processed for voice-like modulation before transmission.The existing voice-like modulation algorithms can be divided into two directions: data is modulated into voice-like signal which is suitable for transmission over voice channel by using modulation methods;according to the mapping relationship,data is mapped into one-to-one corresponding voice-like symbols.In this paper,algorithm of data transmission over voice channel is designed,and simulated over voice channel which based on the idea of modulation algorithm.By analyzing the time domain waveform of voice,it's composed of different frequencies sine wave and a small amount of noise,so orthogonal frequency division multiplexing modulation algorithm can be used to realize the conversion of data to voice-like signal.In OFDM,it divides the channel into several sub-channels which has a narrow frequency band.Data is also converted to parallel low-rate data which has the corresponding number of sub-channels.On each sub-channel,data is modulated by using carriers,and then IFFT is used to transform signals from frequency domain to time domain.Finally signals were converted to speech signals by using parallel-to-serial conversion.In addition,signals modulated resemble speech signals on the waveform by using FFT.Through simulations,it is found that the algorithm can achieve better performances of data transmission over certain vocoders,and the error correction code can improve the BER of data transmission.However,this algorithm does not satisfy the robustness of data transmission over all vocoders.Therefore,it is necessary to design a voice-like modulation algorithm that satisfies non-specific voice channels.By using traditional digital modulation methods,data can be transmitted over the voice channel by adopting appropriate parameters.Thus considering non-specific and multiplevocoders cascades,the traditional modulation method can be used to design and implement speech-like modulation algorithm.Firstly,we can analyze the frequency response of the voice channel.Secondly,we should select the favorable frequency to be the carrier's frequency.Then,we will select an appropriate parameter and modulation method to implement the speech-like modulation algorithm which called Hermes.Using transcoding to guarantee the signals have a fixed pitch frequency.Simulations show that the algorithm can achieve a robust transmission over non-specific vocoders,and it also has a better transmission performance over different vocoders cascades.In OFDM algorithm,when data rate is 2.4 kbps,BER can reach to 0.25% on EFR by using4QAM;when encoding rate is 12.2 kbps in AMR,BER can reach to 0.69% by using BPSK,and BER can be 0.03% after adding error correction code.It can be seen that the algorithm can implement robust data transmission over vocoders which has a high encoding rate.In Hermes,when data rate is 2 kbps,BER can reach to 0.01% on FR,and when encoding rate is 12.2 kbps in AMR,BER can reach to 0.04%.What's more,when data rate is 2 kbps,the algorithm can achieve a BER of 0.70% under the condition that AMR and EFR is cascaded.Therefore,the Hermes algorithm can achieve robust transmission of data over non-specific vocoders and cascaded vocoders. |
