数字信号在语音信道中的传输算法研究

发布时间：2018-05-26 18:18

本文选题：数字信号 + 语音信道　；参考：《西安电子科技大学》2014年硕士论文

【摘要】：由于语音通话安全防范措施的缺失,近年来移动端到端的语音通信安全问题已受到越来越多的关注。通过加密技术对语音信号进行加密保证了语音信息在移动端到端间通信的安全性以及直接性,由于加解密过程中的操作容易性、加密速度性以及加密稳定性等原因,对于数字语音信号的加密技术是主要的语音信息的加密方式。需要说明的是,将模拟信号转换成数字信号是进行数字语音信号加密前的必要工作。由于通信系统中的数据信道无法保证数字信号传输的短时性和实时性,并且会与国际网络出现互通问题,因而需要借助语音信道来进行数字信号的传输。语音信道是多用来传输话音信号的模拟信道,模拟话音信号与数字信号在内容组成、频带范围等方面存在很大差异,尤其是,模拟语音信号有数字语音信号所没有的类似真实语音的特征。语音信道的关键部分是声码器,只有具备类语音特征的信号才能较好的通过声码器。因此,如果一般的数字语音信号没有类似真实语音的特征,那么这种数字语音信号在通过语音信道的声码器时就有可能被当作噪声而丢弃,从而在接收端对数字信号进行恢复时产生较大的误码,进而对合成语音造成严重失真。所以说,如何进行调制传输使数字信号在语音信道上可靠传输,并且保证语音信号的安全性,这才是移动端到端语音信号保密通信的关键技术。对于在语音信道中传输数字信号过程中出现的问题,我们主要做了以下几方面的贡献:(1)利用混合激励线性预测,也就是MELP(Mixed Excitation Linear Prediction)方式,我们可以通过语音压缩编码的方式来对语音信号进行语音压缩的编码,在编码完成后,再进行SM2椭圆曲线公钥密码算法方法的加密,这样便保证了语音信号传输过程中的安全性。(2)对于类似真实语音调制过程中存在的各种各样的问题,我们提出了在发送端对所需要发送的数字信号进行信道编码、正交频分复用(OFDM,Orthogonal Frequency Division Multiplexing)类语音调制和快速傅里叶变换(FFT,Fast Fourier Transform)插值这三个模块的处理,通过以上三个模块的处理,我们便可以优化类似真实语音信号调制的过程,从而使所需传输的数字语音信号可以具备类似真实语音的特征。也就使语音信号可以在GSM语音信道中可以可靠、稳定、安全的传输。特别地,本文对OFDM类语音调制和FFT插值所需要的参数设置方法进行了设计和取值分析。(3)在此基础上,本文给出了在GSM语音信道中抗声码器压缩的数字信号安全传输的系统实现方案,并讨论了该系统应用于3G和4G系统的理论可行性。(4)最后,在MATLAB仿真环境下对本文所设计的系统分别在GSM语音信道中和3G语音信道中进行了仿真,并对仿真结果进行了分析。从我们仿真的结果可以看出,在GSM语音信道的仿真环境中,我们利用MELP语音信号压缩编码技术将模拟语音信号转换为数字语音信号,与此同时将转换后的数字语音信号通过SM2加密以及类似真实语音的调制,最后该数字语音信号能够以近似真实语音信号的形式在GSM语音信道中传输,接收端收到的语音信号的误码率也非常小。同样,在3G语音信道中,采用相同的方法同样可以将数字语音信号转换为近似真实语音的信号形式,而且在接收端收到的信号的误码率同样非常小。因此,我们对数字信号在GSM语音信道中安全可靠的传输提供了一种非常可行的算法,并且也同样适用于3G语音信道。
[Abstract]:In recent years, the security of voice communication in mobile end to end has been paid more and more attention due to the lack of security precautions for voice calls. Encryption technology to encrypt the voice signal to ensure the security and immediacy of the voice information in the mobile end to end communication. Because of the easy operation and encryption in the process of encryption and decryption, the encryption technology is easy to encrypt and encrypt The encryption technology for digital voice signals is the main way to encrypt the main voice information. It is necessary to explain that converting analog signals into digital signals is a necessary work before encrypting digital voice signals. Because the data channel in the communication system can not guarantee the short transmission of digital signals, the data channel in the communication system can not guarantee the short transmission of the digital signal. It is time and real-time, and interworking with international network, so the voice channel is needed to transmit the digital signal. The voice channel is the analog channel used to transmit voice signals, and the analog voice signal and the digital signal are very different in content composition, frequency range and so on, especially, analog voice signal The key part of the voice channel is the characteristic of the real voice. The key part of the speech channel is the vocoder, only the signal with the class voice feature can pass the vocoder. Therefore, if the general digital voice signal is not similar to the true voice, then the digital voice signal is in the voice channel sound. The code device may be discarded as noise, resulting in a larger error in the recovery of the digital signal at the receiving end, and thus causing serious distortion to the synthetic speech. So, how to transmit the digital signal reliably on the voice channel and ensure the security of the voice signal, which is the end to end of the mobile phone. The key technology of the speech signal secret communication is the following contributions: (1) we can use the hybrid excitation linear prediction, that is, the MELP (Mixed Excitation Linear Prediction) mode, and we can use the voice compression coding to make the speech sound. The signal is encoded by the speech compression. After the coding is completed, the SM2 elliptic curve public key algorithm is encrypted, which ensures the security of the speech signal transmission. (2) we propose a digital letter to be sent at the transmitter to the various problems in the similar real voice modulation process. Channel coding, OFDM, Orthogonal Frequency Division Multiplexing class speech modulation and fast Fourier transform (FFT, Fast Fourier Transform) interpolation, the processing of these three modules, through the processing of the above three modules, we can optimize the process of similar real voice signal modulation, thus making it necessary The transmitted digital voice signal can have the characteristics similar to the real voice. It also makes the speech signal reliable, stable and secure in the GSM voice channel. In particular, this paper designs and analyzes the parameter setting methods needed for the OFDM class speech modulation and FFT interpolation. (3) this paper gives the article in this paper In the GSM voice channel, a system implementation scheme for secure transmission of digital signals compressed by a vocoder is proposed, and the theoretical feasibility of the system used in the 3G and 4G systems is discussed. (4) finally, the system designed in this paper is simulated in the GSM voice channel and 3G voice channel under the MATLAB simulation environment, and the simulation results are carried out. From the simulation results, we can see that in the simulation environment of the GSM voice channel, we use the MELP speech signal compression coding technology to convert the analog speech signal to the digital voice signal. At the same time, the converted digital voice signal is encrypted by SM2 and similar to the real voice modulation. Finally, the digital voice signal is used. It can be transmitted in the GSM voice channel in the form of approximate real speech signal, and the error rate of the speech signal received by the receiver is also very small. Similarly, in the 3G voice channel, the same method can be used to convert the digital voice signal to the form of the approximate true voice, and the error rate of the signal received at the receiving end is the same. The sample is very small. Therefore, we provide a very feasible algorithm for the secure and reliable transmission of digital signals in the GSM voice channel, and it also applies to the 3G voice channel.
【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN912.3

【参考文献】