While mainly a Matlab and C exercise in its realisation, this project is based on theories of both speech-optimised coding—also referred to as vocoding—which is an optimisation of standard audio signal coding for the specificity of speech, and joint source and channel coding, a novel theory of general signal coding, that goes in exactly the opposite direction as the one established by the Shannon theorem of separation.
Given the conditions of the project, I had to decide, jointly with Arturo Veloz and Jean-Marc Boucher, the two professors responsible for this project, and Fernando Villavicencio, the exchange student associated with the project, on a “close-to-real-life” application. Considerations of practicality and feasibility led us to decide to choose and include the vocoder developed by Fernando for his MSc project, as well as the North American CDMA standard IS-95A into our application. The application would support experimentation on joint coding over fading channels, as required in the project outline (see Appendix A).
This dissertation aims at reporting the work achieved during the project, and the results obtained. It is organised in three chapters. Chapter 1 will present background information and report some state-of-the-art in speech coding and joint source-channel coding. In turn, Chapter 2 will present in greater detail the components chosen in the project: Fernando Villavicencio’s TCENLP vocoder, and the IS-95A CDMA standard. Finally, Chapter 3 will present implementation details of the project, as well as the results we could obtain from our study.
Unfortunately, the implementation of a working system proved to take up most of the time resource allocated for the project so that while implementation details are numerous, results are few. While our attempt at a scheme of joint coding proved a failure, it taught the author a valuable lesson. Furthermore, the project leaves a working implementation that may be re-used later on for more ambitions projects.