Low-Rate Voice Compression Codec System Design And Implementation Based On AMBE-2000

As society is developing, more information is to be transmitted in people's lives, which makes the limited channel resource becomes more precious. With the purpose of saving valuable bandwidth, how to compress the original information as possible is the interest of the researchers. Signal compression has become one of the key technologies in communication and information processing. In the multimedia communication field, voice compression technology has been developed for almost 20 years, and yielded substantial results. But the compressible room of voice signal remains great, there are broad prospects. So a kind of low-rate compression system is designed and realized in this thesis.Firstly, the Voice Compression and Coding basic theory is introduced in this thesis. According to the characteristics of the voice signal, and combining with the classic smooth signal processing theory, the basic skills and methods of voice signal processing is achieved. Given several practical vocoder model, Introduced a vocoder model of the basic concepts and compared with the typical voice compression algorithms, their respective advantages and application conditions are proposed. The technology and application characteristics of the multi-band excitation (MBE) is mainly analyzed. A kind of advance multi-band excitation algorithm (AMBE) is proposed in this thesis. The compression algorithm used in the voice compression chip is the AMBE algorithm proposed.Secondly, a hardware voice compression codec system is designed and realized in this thesis. Details of the system design and hardware PCB production process is introduced. The entire system considers the voice compression chip AMBE-2000 and TMS320VC5402 DSP as the core, and the external circuit is contributed with other function chips, which can realize the functions of voice signal acquisition, compression and decompression coding, encryption and decryption, signal transmission and so on. The interface design method between the function circuits is introduced, and time sequence of the system is analyzed in this thesis. The feasibility of the system is analyzed and demonstrated from the cost, functional and practical aspects. Finally, this thesis presents the experimental results, Oscilloscope measured using voice compression chip several major pins normal working hours waveform trajectory map. The effective use of a voice, "Bing Cheng Harbin" as a test, original voice and after hardware compression decompression after the speech into computer, read using MATLAB software can be compressed before and after the time-domain waveform and frequency-domain waveform contrast to the map. The results of analysis and research, draw conclusions and sum up the hardware circuit design and debug needs attention.