基于RISC的MPEG-4音频解码软件优化  

Software optimization of MPEG-4 audio decoding based on RISC processor

作　　者：梅优良[1] 刘鹏[1] 周建[1] 陈科明[1] 

机构地区：[1]浙江大学信息与电子工程学系,浙江杭州310027

出　　处：《浙江大学学报（工学版）》2007年第4期603-606,692,共5页Journal of Zhejiang University：Engineering Science

基　　金：国家"863"高技术研究发展计划资助项目(2002AA1Z1140);霍英东教育基金资助项目(94031)

摘　　要：为了提高MPEG-4先进音频编码(AAC)的解码效率,提出了在通用32位精简指令集计算机(reduced in-struction set computing,RISC)上实现MPEG-4 AAC低复杂度框架解码的软件优化技术.解码过程可以分成比特流解码部分和运算部分.应用存储器分层结构加速比特流解码;基于运算过程、对象和乘法运算的优化技术提高运算部分的解码效率.结果表明,在35 MHz的处理器频率上实现了立体声实时解码,提高了解码效率.In order to improve the efficiency of the MPEG-4 advanced audio coding （AAC） decoding procedure, software optimization technology was proposed for implementation of MPEG-4 AAC low complexity profile audio decoding on general-purpose 32-bit reduced instruction set computing （RISC）. The decoding procedure was divided into bitstream decoding section and computation section. Hierarchical memory architecture was applied to accelerate bitstream decoding, and the decoding efficiency of computation section was improved by the software optimization which was implemented based on the processes, targets of computation and multiply operations. Great decoding efficiency was achieved with the result of decoding stereo audio in real time at the maximum processor speed of 35 MHz.

关 键 词：MPEG-4 先进音频编码 精简指令集计算机 瞬时噪声整形 

分 类 号：TN912.3[电子电信—通信与信息系统]