The Design And Implementation Of Multi-channel Voice Processing System Based On G.729A

In recent years, with the wide application of new technologies in the related field ofelectronic communications, computers, signal processing, the voice codec technology hasbeen developed rapidly and achieved a series of breakthroughs. In practical engineering, inorder to solve the communication quality and communication rate which are two mutualrestraint issues, we need to design high-quality, lower-cost voice codec.From the practical point of view, this paper designed the design of the four-channel voicesignal processing system based on G.729A codec algorithm, the purpose is to exchangereal-time requirements of the system by reducing voice signal quality.In terms of hardware,elaborated hardware design principles and implementation process for each part of thesystem,this system chooses the TMS320C6416as the encoding and decoding of speech signalprocessor,and it is able to coding and decoding four-channel voice signal.In terms of Codecsalgorithms,by analyzing the core codec algorithm, in order to reduces the co mplexity of thealgorithm,this paper makes two optimization in the algorithm level,the first is acceleration ofthe multi-level search,the second is acceleration of the fixed codebook search. By voicequality assessment and system processing performance tes ting,the results show thatcomparing with the original algorithm,under the condition to ensure voice quality,theoptimized G.729A codec algorithm rate up by5.6times,can meet the requirements ofreal-time multi-channel voice processing system.