Optimizing Communications Systems: Finding Gains at the Least Marginal Cost

The analytical expression for the ultimate limit of communications efficiency has been known since Shannon's "A Mathematical Theory of Communication", published in 1948. Since then, the digital revolution has provided the signal processing complexity needed to close the gap between this limit and practical communication systems. This dissertation on optimizing communications systems investigates the tradeoff between communications efficiency and signal processing complexity in light of the current state of art techniques in communications theory. It begins with techniques of calculating information rates for non-linear satellite channels. Information rate calculations give the absolute limit in communications efficiency and are used as a reference point to the optimization problem. The second part of this dissertation uses information rate calculations to examine the complexity versus efficiency tradeoff of a physical layer approach known as faster-than-Nyquist (FTN) signalling, which has been a popular area of research recently. Our results show FTN does not show significant advantages over a much simpler OFDM technique, which is already widely implemented in modern communications systems. These results inspired the question of what method of optimizing communications systems could be obtained at the least marginal cost in complexity. While techniques such as MIMO still holds promise for future improvement at the physical layer, application specific optimization may give the greatest performance gain at the least cost in signal processing complexity. The final part of this thesis examines this statement and hopes to provide evidence of this conjecture.