Turbo Channel Coding for AMR Speech Codec

• Main Authors: Wei-Chih Tzeng, 曾偉誌
• Other Authors: 簡福榮
• Format: Others
• Language: zh-TW
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/8u88a8

碩士 === 國立臺北科技大學 === 電機工程系研究所 === 95 === European Telecommunication Standards Institute (ETSI) initiated a standardization program in October 1997 to develop an Adaptive Multi-Rate (AMR) speech codec for Global System for Mobile Communication (GSM). Depending on channel conditions, the AMR codec offers substantial improvement in error robustness by adapting bit rates between speech and channel coding. In April 1999, the Third Generation Partnership Project (3GPP) adopted the AMR codec as the mandatory speech codec for the third generation WCDMA (Wideband Code Division Multiple Access) system. Three different channel coding schemes are incorporated with the AMR codec and compared in this thesis. The first scheme encodes the bitstream with the convolutional code and decodes it back using the hard-decision Viterbi algorithm. The second scheme encodes the bitstream with the same convolutional code but decodes it back by a soft-decision Viterbi algorithm. The last one exploits the turbo code to encode the bitstream and decodes it back by using the log-maximum a posteriori (log-MAP) algorithm. Both unequal error protection (UEP) and equal error protection (EEP) strategies are also included for comparison. The experimental results show that the performance of the UEP is better than that of the EEP in low signal-to-noise ratio (SNR) environment, and the third scheme that exploits turbo code performs best among all of the three channel coding schemes.