北京邮电大学学报

  • EI核心期刊

北京邮电大学学报 ›› 1999, Vol. 22 ›› Issue (2): 67-71.

• 研究报告 • 上一篇    下一篇

MPEG声音编码的单片DSP实现

林胜, 门爱东   

  1. 北京邮电大学电信工程学院, 北京 100876
  • 收稿日期:1998-06-04 出版日期:1999-03-10
  • 基金资助:
    国家“九五”重点科技攻关资助项目

Single Chip DSP Implementation of MPEG Audio Encoding

Lin Sheng, Men Aidong   

  1. School of Telecommunication Engineering, Beijing University of Posts and Telecommunications, Beijing 100876
  • Received:1998-06-04 Online:1999-03-10

摘要: 为了开发价廉的单片定点DSP声音编码器, 对MPEG声音编码标准参考算法的运算量和存储量要求进行了比较深入的分析, 综合考虑了编码质量和处理器速度的要求, 结合计算机仿真结果, 指出了使用单片定点DSP实现的关键所在.基于AD公司的ADSP-2181, 充分利用其为声音处理优化的硬件结构, 设计实现了实时MPEG声音第2层编码的软硬件方案.测试结果表明, 在采用基于MAC的滤波器精度扩展和改进的心理声学模型算法的条件下, 编码质量和实时性都能得到保证.

关键词: 压缩编码, 声音处理, 数字信号处理

Abstract: To develop a low-priced audio encoder using single fix-point DSP.The requirements of calculation and memory of MPEG audio encoding reference algorithm are analysed. Taking both the encoding quality and speed of processor into consideration, along with the result of simulation, the critical problem of single fix-point DSP implementation are point out. Fully using the property of the hardware optimized for audio processing, we have designed the software and hardware of a real-time MPEG audio layer 2 encoding implementation based ADSP-2181 of Analog Device Corp. The test result shows that both encoding quality and real-time property can be achieved when MAC based precision extension filter and modified psychoacoustic model are used.

Key words: compressed encoding, audio processing, digital signal processing

中图分类号: