Font Size: a A A

Underwater Acoustic Digital Speech Communication System Based On AI Speech SDK Library

Posted on:2022-06-20Degree:MasterType:Thesis
Country:ChinaCandidate:X L GuoFull Text:PDF
GTID:2518306350982859Subject:Information and Communication Engineering
Abstract/Summary:PDF Full Text Request
Underwater acoustic channel has the characteristics of low information carrier propagation rate,severe propagation attenuation,time varying,space varying and frequency varying,which leads to the realization difficulty on high-speed and long-distance communication,the transmission speed of traditional speech communication still reaches several kbps after applying various efficient compression coding algorithms,for the reasons above,it's quite a great choice to realize low-speed underwater acoustic speech communication(UASC)by applying speech recognition and synthesis.Spread spectrum communication system has strong anti-jamming ability,good communication concealment and easy to realize CDMA,which satisfies needs of most UASC working scene,so that this thesis design and implement the UASC system based on speech recognition/synthesis and spread spectrum communication,the realization of speech recognition and synthesis is based on Baidu AI Speech SDK Library,and the spread spectrum communication system adopts DSSS.Main contents of this thesis include the research and analysis of key algorithms of direct sequence spread spectrum communication and the design and development of communication system client.For algorithm parts,this paper describe and analysis modulation mode,PN sequence,synchronization methods and time domain decision feedback equalization,results show that m-sequence is the excellent candidate,BPSK has the finest reliability,BPSK/CSK hybrid mode has the highest transmission rate,T-DFE and source encoding work together to effectively improve the anti-jamming ability,DSSS signals without source coding will cause serious error propagation in the process of decision feedback equalization at low SNR,and synchronization is a necessary signal processing step in actual system.Based on the performances simulation results above,the spread spectrum communication system client is developed on Android OS,speech recognition/synthesis is integrated by Baidu provided SDK package,communication signal process is implemented by calling Java class or native code library by JNI interface,the final version of client includes speech processing,communication signal processing,signal acquisition and playback,working process log display,which work together to complete the UASC system process.Finally this paper validates the function and performance of mentioned algorithms and system client by real phone debugging and pool experiment,results show that performance of algorithms in actual underwater environment basically accords with simulation analysis,the client function verification is passed,the reliable transmission and communication rate basically meet the design requirements.
Keywords/Search Tags:underwater acoustic digital speech communication system, spread spectrum communication, speech recognition and synthesis, Android application development
PDF Full Text Request
Related items