Voice Compression Codec Research And DSP Implementation In Trunking Communication System

In the digital trunking communication system, because of the frequency resources are so limited, it is necessary to send the voice signal by compression coding. Meanwhile, in order to enable users to get good receivinig voice, the compression algorithms should have good voice quality and shorter time delay characteristics.Among so many voice compression standards, the performance of G.729 voice codec which is established standards by ITU-T (International Telecommunication Union) in 1996 is very eye-catching. It uses the CS-ACELP coding scheme and can compress the sample rate of 64kbps voice signal (PCM) to 8kbps rate, almost without any quality losing. Meanwhile the algorithm delay is only 10ms, therefore it is desirable voice codec algorithm in digital trunking communication system. In recent years, digital signal processing algorithms and devices (DSP chip) are developing rapidly, so it implements the speech encoder.In this paper, it shows a voice codec system design which based on the G.729 standard by using the DSP chip-TMS320C5402 as a core processor. First we introduce the current popular voice codec technology, especially parametric coding technology; then describe the G.729 voice codec algorithm in detail, and use this algorithm as the core voice algorithm of the system designe; and then introduce the system design from the software and hardware comprehensively, including the connection between the various chips, driver programs and DSP main programs design; Finally, Aimed at the hardware system characteristics, we improve the G.729 standard from the the program structure and algorithm method, and successfully implemente in the system. By comparing the voice waveform between original and reconstruction, it proves the correctness of the system codec from the objective; meanwhile, through listening by the human's ears, it proves the feasibility of the system codec from the subjective.