The Improvement And Optimization Of The Algorithm Based On Arm Platform Mp3 Encoder

In the current information society, there is much voice information which containsmass of redundancy. It will take up vast resources in transmission and storage processesif the redundant is not removed. Therefore, voice compression technology has been paidwidely attention.In this dissertation, the development status of audio compression technology isintroduced firstly, and then the basic principle and structure of MPEG-1layer3(MP3)audio codec is investigated. Especially, several key modules such as psychoacousticmodel, sub-band filtering and quantization coding are carried on further study. Based onall of these, an implementation solution of MP3codec is proposed, which focuses onthe improvement for the psychoacoustic model, quantization module and sub-band filterbank module. Finally, MP3codes is optimized across-the-board according to thecharacteristics of ARM, and the optimization results are transplanted to the ARMplatform.The experimental results indicate that the improved and optimized algorithm is notonly characterized by excellent coding quality, but also lower computational complexity,so it needs lower requirements for hardware, cost, power consumption and size forsystem, and its performances can reach the expected design requirements.