Development And Realization Of The Algorithm Of Speech Coding Based On DSP

With the rapid development of Multimedia communication technology and Internet technology, Application of Digital speech Coding is more widely in recent years, especially in the area of Visible telephone, IP Internet telephone, Digital cellular mobile communication, Integrated server digital network, Public Switched Telephone Network and packet transfer Voice System, etc. Digital speech Coding can reduce rate of bits, for the purpose of provided more channels to transfer image, document, computer files and other Data stream, but the operation must assure quality of speech. So ITU introduce G729, a voice compressing criterion, Conjugate-Structure Algebraic- Code-Excited Linear -Prediction (CS-ACELP) and G723.1 criterion, 6.3kbps å'Œ 5.3kbps rate speech Coding in 1996. Two criterions have the advantages of high speech quality, low time delay and good stability, which can be applied in the area of digital speech communication and be worth researching highly.First of all, the development course and tendency of Speech Compress Coding has been introduced. On the basis of analysis and research to theory of G.729 criterion, G.723.1 criterion arithmetic, realization of the algorithm of speech coding's system based on TMS320VC5509 is narrated. Due to large operation when arithmetic is run in real-time, several methods of optimized codes was put forward. Loaded optimized codes in TMS320VC5509 and simulated, two arithmetic's performance was compared.Whole paper is made up of five chapters. In the first chapter, summarizes the classification, standard and tendency of Speech Coding;in the second chapter, theories of G729 criterion, G723.1 criterion arithmetic are analyzed;design of hardware and software speech coding's system based on TMS320VC5509 is given in the third and forth chapter;in the last chapter, the main conclusions of the design are summarized.