Enhanced Variable Rate Speech Codec Algorithm

To meet the needs of communication development for higher speech quality and greater system capacity, CDMA systems employ variable rate speech coding techniques to obtain more flexibility. Starting with the principal concepts of speech compressing coding, this paper introduces the algorithm and characteristics of the enhanced variable rate codec (EVRC) in detail.The algorithm of this paper employs three coding rate. It comprises two parts, coder and decoder. The coder applies preprocessing, linear prediction analysis and long-term prediction analysis to input speech signal, and abstracts parameters such as linear spectrum frequency, pitch and so on. Combined with fixed codebook and adaptive codebook, these parameters are encoded in different ways according to current frame's coding rate, which is decided using a rate decision algorithm. The decoder decodes received data packets in terms of their rate types, and rebuild synthetic speech signal, which is then post filtered to get better speech quality. Basing on relaxed code excited linear prediction algorithm, this paper decides the coding rate according to the character of speech signal, so that fairly good communication quality and low coding rate can be obtained. This method is very valuable in reality.The main task is to research on this algorithm and to realize the system, and then do more work on preprocessing, rate decision and objective speech quality evaluation. In this paper, a speech enhancement method is proposed, which can enhance speech signal well even in very noisy environment. Fuzzy classifier is used to detect the word boundary with high precision in noisy environment. An object evaluation is introduced and compared with traditional Mean Opinion Score (MOS).