Based On The Design Of Low Bit Rate Voice Algorithm Mcu

Speech Coding is taking use of the redundancy between speeches and the restrict of ear to get rid of the unnecessary information and attain the goal to reduce the bit rates or storage capacity., In the paper we just realize the compression of speech according to these two characters,the content of the paper is to complete their realizations in software and hardware respectively.First we introduce the theory of wave coding and describe the theory of ADPCM and the process of its realization;then we simulate it in software VC6.0. Based on this, Integrated Circuit is designed in Verilog HDL for this algorithm;In addition,the circuit has passed the simulation,verification,and testing using software Modsim SE6.1 ;finally we synthesize the circuit in Synplify Pro 7.3 and discuss the questions in design and process of solving problems. Due to the constrain of bit rate for this algorithm,it mainly applies to the occasions which require high quality for speech.In the other hand,we deeply analyze the critical technology of parameter coding algorithm,and specially discuss the theory of ACELP and the functions of all parts,during this,we analyze the performance efficiency for those important parts and optimize the program code,which provide the foundation for the following realization in hardware;Then we simulate it in software and compare the results between ADPCM and ACELP,and express the reasons and analyze them for their differences.In the last part,we discuss the hardware realization for ACELP in hardware design which is based on IP C.CORE;Several instruction cycles are needed when it executes multiply operation,while in this algorithm,multiply operation occupies big proportions, so we design a special multiply circuit. Moreover,a great many programs are optimized to improve the efficiency in Hardware. Ultimately we complete the realization in hardware;but in order to satisfy the application in practice,more work needs to be done.