Research Of Speech Codec System Based On DSP

Speech signal processing is an important part of modern communication research, speech compression coding technology as the key of it now has achieved a great development. Therefore, it is of great importance to study, understand and inquire the voice coding-related technology..Through access to relevant literature, the current major speech coding algorithms were compared and analyzed. Based on in-depth study of key technology for speech coding, the paper analyzed the principle of the G.729 algorithm. Being a representative algorithm in low bit rate speech coding, G.729 was adopted as the core coding algorithm in the current study. For its characteristics of algorithm, the DSP chip was used as the core processor of the system, and a versatile and scalable voice codec hardware platform with modular characteristics was developed, and the software on this platform was also designed. In order to overcome the shortcomings of the standard algorithm of G.729, i.e. low efficiency and long execution time, the paper focused on streamlining the algorithms, code and others to make the source code optimized. The optimized algorithm improved the coding efficiency, while ensured high-quality voice output,, realizing the real-time speech signal processing. In view of the above characteristics, in addition to the use as a speech codec, the system is more suitable for speech signal processing education, providing students a hardware and software platform for learning and practicing. Finally, the performance of the system was analyzed and evaluated, and the results showed that the indicators of the speech codec system meet the design requirements, and the system is stable.