This dissertation considers issues of blind parameter estimation for data acquisition in digital communication systems. Signal processing for data acquisition in digital communication systems mainly consists of synchronization, channel estimation, and equalization. While blind channel estimation and equalization techniques for linear modulation schemes have received intensive attention, neitherblind (adaptive) synchronization for linear modulation nor blind equalization for non-linear modulation schemes have received nearly as much attention. We propose blind adaptive estimation schemes for several synchronization parameters for single and multi carrier systems, and study blind adaptive equalization for a particular non-linear modulation scheme.
The main approach we use in blind synchronization parameter estimation is dispersion minimization, which attempts to restore the dispersion free nature of the digital signals. We consider blind timing offset recovery based on dispersion minimization (chapter 2), carrier phase offset recovery based on dispersion minimization (chapter 3), and carrier frequency offset recovery of multicarrier systems based on fourth power minimization as a variation of dispersion minimization (chapter 4). Due to the multimodality and the suboptimality of these approaches, performance analysis, for example, classification of desirable and undesirable local minima, initialization strategies, and achieved MSE performance, is conducted for each application. Finally, the effect of a multipath channel on an FSK signal is analyzed and several blind adaptive schemes to mitigate the distortion due to a multipath channel are proposed.