Real-time Multi-channel Voice Recording System Based On TMS320C5509

DSP (Digital Signal Processor) has been widely used in high-speed controling , image processing, communications technology, wireless, voice processing, network, equipment, medical equipment, instrumentation and home appliances, and other fields; DSP provide a highly efficient and reliable infrastructure hardware for digital signal processing. The design uses the TMS320C5509 which is specifically applicable to the voice signal processing.From the end of the 1930s when PCM and channel vocoder theory was advanced, the speech coding has developed nearly 70 year. And in the past 20 years, with the development of computer and microelectronics technology, high-quality voice compression coding technology has begun to be used in large-scale.The G723.1 standard can provide the high quality of the reconstruction of voice at two lowe encoding rate ,5.3 kbit / s and 6.3 kbit / s.Railway communication system plays a critical role in insuring the smooth operation of the railway system.At the same time, the accurate recording of the voice item during the communication of the railway is of vital importance when determination the responsibilities and evidence against law is needed.Present recording system mostly adopts tape recording method or recording system based on the PC, which has not considered voice compression to save storage space. This paper will introduce a digital real-time multi-channel voice recording system used in rail communication, which adopts DSP as its processor and compresses the voice data by G723.1 to save storage space and stores the compressed voice data by database on PC.This paper introduces the hardware and system design, systematically analyses the principles of G723.1 and the C source code. Then, we take a series of measures to optimiz the C source code on the basis of special hardware and instructions of 5509. The efficiency of the code has been greatly improved, so system can deal multi-channel voice real-time.