Integrated design of digital communications systems

Design of digital communications systems has been traditionally divided into three subsystems, namely, source coding, channel coding, and digital modulation. Each subsystem may be optimized separately with little, or even no, consideration of others. As a result, systems designed in this conventional way have high complexity, and are sometimes not practical. A more efficient approach to digital communications systems design is warranted. In this dissertation, new algorithms are developed to design digital communications systems in an integrated fashion, i.e., to design a system as an entire entity giving consideration to all the constituent components. The algorithm first focuses on transmission systems operating over fixed condition channels. It is then extended to the case of communications systems operating over time-varying channels. Particularly, a new design algorithm for integrated adaptive communications systems is developed in the case of communications systems operating over time-varying wireless channels.; In order to design a system as a whole, all constraints and key subsystems of the system are taken into consideration. First, all aspects of the design are characterized into three primary categories of restrictions, namely, nature of source information, system resource budget and user performance requirements. By using rate-distortion theory, a set of guidelines have been analyzed and summarized based on these three restrictions. From these guidelines, a general algorithm for integrated design of communications systems has been developed. This new algorithm, which provides judicious designs for each subsystem, is quite general and not particularly application-oriented. A large variety of existing implementation methods for each subsystem may be utilized in achieving a practical system design with low complexity.; Furthermore, in the case of digital communications systems operating over time-varying, adaptive subsystems are needed to combat the time-varying nature of the channels. The dynamic capacity of the channel may be segmented into several regions. A different operating state is associated with each region. Assuming that the adaptive system adapts fast enough within each adaptive interval, the system may be considered to operate over a fixed channel condition. As a result, the adaptive system operates within these predetermined states. Examples are given for speech and data transmission. The examples show that the adaptive systems designed according to the new adaptive algorithm yield superior performance compared to that of a system designed and implemented with a fixed configuration.