Analysis and design of space-time coded systems for wireless communication

Space-time coding is an efficient way of achieving transmit diversity, which allows high-speed data transmission over wireless channels. It has been already incorporated in the standardization process of third generation wireless communication systems. However, for a successful integration of space-time coding in future wireless communication applications, numerous problems related to this new family of codes are yet to be addressed. Motivated by these problems, this dissertation focuses on different aspects of the analysis and design of space-time codes and space-time coded systems. Performance analysis for two different classes of space-time codes, namely trellis codes and block codes, is carried out through the derivation of an exact pairwise error probability (PEP) expression. This leads to the definition of dominant factors on the performance over fading channels and gives better insight into the inner mechanisms of these codes. Based on the derived PEP, analytical error rate performance results are also demonstrated. Besides the performance evaluation, many critical aspects related to design and decoding of space-time codes are addressed in this work. Contrary to the pioneering works which assume perfect channel knowledge at the receiver, decoding techniques without requiring explicit channel state information are discussed and a maximum-likelihood sequence estimator for space-time block coded systems is proposed. The proposed scheme is essentially a quadratic incoherent receiver, which does not require any kind of channel knowledge and also outperforms the pilot-symbol assisted techniques which are widely used in practice today. On the code design, motivated by the fact that currently available space-time block codes suffer loss in throughput rate in some cases, non-orthogonal codes are introduced which outperform the current codes in throughput rate. The application of space-time block codes to frequency-selective channels is also discussed in terms of performance and complexity based on a comparative study of different currently available approaches.