Joint source channel coding for entropy based differential speech coders
This thesis examines techniques which improve the performance, robustness and efficiency of digital speech transmission through error prone channels. In particular, we have achieved performance improvements for differential speech coders by replacing traditional single-stage linear prediction (LP) w...
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Theses and Dissertations |
| Published: |
2008
|
| Subjects: | |
| Online Access: | http://hdl.handle.net/10356/3254 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Nanyang Technological University |
| Summary: | This thesis examines techniques which improve the performance, robustness and efficiency of digital speech transmission through error prone channels. In particular, we have achieved performance improvements for differential speech coders by replacing traditional single-stage linear prediction (LP) with switched LP. The efficiency of differential speech coders in terms of average bit rate has also been improved by replacing
fixed-length codes with variable length codes (VLCs). The technique of joint
source-channel coding (JSCC) which exploits residual redundancy in the encoded bit stream has also been applied to improve the robustness of differential speech coders operating in an error prone environment. |
|---|