적응형 다중 비트율 음성 부호화기를 위한 효율적인 대수코드북 검색법An Efficient Algebraic Codebook Search Method for ham Speech Coder
본 논문에서는 적응형 다중 비트율 (AMR: Adaptive Multi-Rate) 음성 부호화기의 구현 시 계산량을 가장 많이 차지하는 대수 코드북 검색과정의 계산량을 줄임으로써 효율적인 AMR 음성 부호화기를 구현하였다. 대수 코드북 검색의 계산량을 줄이기 위하여 기존의 AMR 음성 부호화기에서 사용하고 있는 깊이우선 가지 검색법 (depth first tree) 검색 방법을 개선한 고속 코드북 검색 알고리즘을 제안하였다. 제안된 방법은 검색과정에서 최적의 여기신호로 선택될 가능성이 적은 트리를 제거하여 검색의 복잡도를 줄이는 방법으로 트리 선택을 위한 추가의 계산량이 필요없으며 검색에 필요한 계산량은 기존의 깊이우선 가지 검색법에 비해 현저한 감소를 이루었으나 약간의 음질 저하가 있었다. 제안한 방법을 적용하여 AMR 음성 부호화기의 12.2 kbps 모드를 TeakLite DSP를 사용하여 구현한 결과 기존의 방법에 비해 약 40%의 계산량을 감소할 수 있었다.
In this paper, we efficiently implement the AMR speech coder by reducing the complexity of algebraic codebook search. To reduce the computational complexity of the algebraic codebook search, we propose a fast algebraic codebook search method that improves conventional depth first tree search method used in AMR speech coder algorithm. The proposed method reduces the search complexity by pruning the trees which are less possible to be selected as an optimum excitation. This method needs no additional computation for selecting the trees to be pruned and reduces the computational complexity considerably compared to the original depth first tree search method with slightly degradation or speech qualify. Applying our method to the implementation or AMR speech coder with 12.2 kbps mode by using the TeakLite DSP, we reduce the search complexity about 40% compared to the conventional method.
In this paper, we efficiently implement the AMR speech coder by reducing the complexity of algebraic codebook search. To reduce the computational complexity of the algebraic codebook search, we propose a fast algebraic codebook search method that improves conventional depth first tree search method used in AMR speech coder algorithm. The proposed method reduces the search complexity by pruning the trees which are less possible to be selected as an optimum excitation. This method needs no additional computation for selecting the trees to be pruned and reduces the computational complexity considerably compared to the original depth first tree search method with slightly degradation or speech qualify. Applying our method to the implementation or AMR speech coder with 12.2 kbps mode by using the TeakLite DSP, we reduce the search complexity about 40% compared to the conventional method.
- Publisher
- 한국음향학회
- Issue Date
- 2003-02
- Keywords
ACELP; 적응형 다중 비트율; 대수코드북; 고속검색
- Citation
한국음향학회지, Vol.22, No.2, pp.129
- ISSN
- 1225-4428
- Appears in Collection
- EE-Journal Papers(저널논문)
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.