Applied to an algebraic codebook search conducted on an algebraic code-excited linear-prediction (ACELP) speech coder, two improved versions of reduced candidate mechanism (RCM), designated as Fixed-G1-RCM and Fixed-2Track-RCM, are presented in this study for further search performance improvement. It is mainly derived from two major research findings in a piece of our prior work. The first finding is that a pulse with a high contribution in a track is more likely to serve as the optimal pulse in the optimal vector pertaining to the track, and the second is that the speech quality can be well maintained at a search accuracy above 50%. In this proposal, the trade off can be tuned between the search accuracy and the search complexity so as to reach a nearly consistent speech quality. With this presented search algorithm implemented on a G.729A speech codec, it is experimentally demonstrated that either Fixed-G1-RCM (N = 3), or Fixed-2Track-RCM (M = 2, N = 3, or M = 3, N = 4) can provide a highly superior search performance relative to a global pulse replacement method (iteration = 2) and an iteration-free pulse replacement method.
- Algebraic code-excited linear-prediction (ACELP)
- Algebraic codebook search
- Reduced candidate mechanism (RCM)
- Speech codec