Estimation of wireless communications channels using superimposed training: Approaches, analysis and applications

Wireless communications technology has emerged as the key information transmission platform due to its flexibility in supporting user roaming. Typical wireless channels such as mobile wireless channels, indoor radio channels and underwater acoustic channels, are both time- and frequency-selective. Multipath propagation and limited bandwidth are the two main causes of frequency selectivity, which leads to intersymbol interference. The time variation of the channel is mainly due to the relative motion between the transmitter and receiver, as well as due to oscillator drifts and phase noise. In this dissertation, we concentrate on channel estimation and detection for various wireless systems, including single-input multiple-output (SIMO) time-invariant system, SIMO time-varying system, multiple-input multiple-output (MIMO) time-invariant as well as time-varying systems.; Our focus is on superimposed training for channel estimation where a periodic (nonrandom) training sequence is arithmetically added (superimposed) at a low power to the information sequence at the transmitter before modulation and transmission.; The dissertation is organized into eight chapters. Chapter 1 provides a brief introduction to the topics considered in this dissertation. Chapter 2 presents a performance analysis for the first-order statistics-based approach of [40]. The issue of training power allocation is also considered. Chapter 3 presents a semi-blind DML (deterministic maximum likelihood) approach using superimposed training in a time-invariant SIMO (single-input multiple-output) system. The computer simulation environment is also introduced, including the computer simulation of multipath channels. Chapter 4 provides a performance analysis for the iterative channel estimation in the semi-blind approach described in Chapter 3. The Cramer-Rao lower bound (CRLB) on the channel estimate for SISO systems is also derived. Chapter 5 concentrates on solving the training synchronization problem which is untouched in some prior works and preceding chapters. Computer simulations show that the proposed approach is very selective. Chapter 6 focuses on channel estimation and detection using semi-blind DML approach in time-varying SIMO channels. Jakes' model is used to generate the time-varying channel coefficients in computer simulations. A complex exponential-based expansion model and a polynomial model are used to estimate the time-varying channel. Chapter 7 extends the approaches used in Chapter 6 to MIMO (multiple-input multiple-output) systems. Diversity at the transmitter is introduced to increase the data rate. Finally, we summarize our work and discuss potential future work in Chapter 8. (Abstract shortened by UMI.)...