Design And Implementation Of A Low-Rate Secure Voice Communication System

Voice is an important medium for human communication and exchange.Since the beginning of the new century,voice communication technology has been extensively researched and developed rapidly.Supporting voice interaction has become a necessary function of many mobile communication terminal devices.However,in many mobile voice communication systems,the ground parts other than the air interface are transmitted in clear text,which poses a huge security risk.Therefore,in recent years,the end-to-end encryption technology of mobile voice communication systems has received extensive attention and in-depth research.Among them,the realization of end-to-end voice secret communication based on low-rate voice compression coding and voice-like modulation technology has become one of the research hotspots.This thesis first implements a 500 bps speech compression coding algorithm based on the Enhanced Mixed Excitation Linear Prediction(MELPe)model.The MELPe coding model has shown great potential in the field of low-rate speech coding and is widely used.In order to obtain higher-quality decoded speech at a lower coding rate,the model has been further improved and the coding parameters have been deleted.The margin spectrum amplitude parameter and the non-pitch period flag of the system only retain the extracted line spectrum frequency parameter,pitch period,gain,bandpass voiced voice,and energy parameters to achieve low-rate speech coding,and the final design implements a 500 bps speech compression Coding algorithm and simulation experiment on its voice compression effect.The experimental results show that the 500 bps low-rate speech compression coding algorithm can synthesize speech MOS scores above 2.3 and the intelligibility is 100%,which has high synthesized speech quality.Secondly,in view of the common problem of speech communication systems-the problem of respiratory sound interference,this paper provides a breathing sound detection algorithm based on time-frequency characteristics,using short-term energy,short-term zero-crossing rate,low-frequency voice energy ratio And other parameters,to detect the breathing sound in the speech and make post-processing to improve the auditory effect.Tests show that this algorithm can effectively distinguish between normal speech and breathing sound,thereby eliminating breathing sound interference and optimizing call quality.It can be seen from the experiment that the speech processed by the breath sound detection algorithm in this paper has a good intelligibility;and the missed detection rate and the false detection rate of the breath sound are 4.2% and 0.3%,respectively,which significantly inhibits the breath sound in the speech effect.Finally,based on the 500 bps speech coding algorithm and breath sound detection algorithm implemented by the above design,combined with the genetic algorithm-based speech modulation and demodulation method,a low-rate speech secret communication system is constructed,and an experimental simulation is performed on the system.The test results show that the system can effectively eliminate breath sound interference,and it has high security performance under the selected GSM-EFR channel environment,and the decrypted voice has high intelligibility.