Space-Time Coding Techniques For Wireless Communication Systems

With the development of economy, advancement of society and improvement of people's standard of living, new generation wireless communication systems are expected to provide higher transmission rate, offer better quality and support faster moving terminals. The most effect way to meet such requirements is the MIMO technology, which deploys multiple antennas at transmitter and/or receiver. Recent research in information theory has shown that larger gains in the capacity of channels are feasible in MIMO systems. To make full use of the capacity of MIMO channels, space-time coding is a set of practical signal techniques when there is no channel state information (CSI) at the transmitter. Space-time coding introduces the correlations in both time and space domains when signals are transmitted by different antennas, which can greatly enhance the data transmission rate and improve the performance without bandwidth expansion or transmit power augment. This dissertation concentrates on the space-time codes and detections for MIMO systems, and it includes:In this dissertation, a new family of space–time block codes called diagonal block orthogonal algebraic space-time (DBOAST) block codes is proposed. The DBOAST block code can transmit at a normalized rate of 1 symbol per channel use with higher code advantage and larger capacity of the equivalent channel including the space-time code than those of diagonal algebraic space-time (DAST) block code. When using sphere detector, the detection problem for the DBOAST block code can be divided into two independent detection problems with half of the original dimensions, so that the computational complexity can be reduced. In the case of no CSI at the receiver, combining Cayley transform with the DBOAST block code, a new differential unitary space-time block code is given.Integrating trellis-coded modulation (TCM) with threaded algebraic space–time (TAST) block code, trellis-coded modulation space-time (TCMST) block code is presented, which has the advantage of full-diversity and small detection delay. Based on TCM pre-processing, the minimum distance between different codeword matrixes of the TCMST block code is maximized, so the average bit error rate of the TCMST block code is lower than that of the TAST block code at the same signal noise ratio. In addition, trellis sphere detection algorithm for the TCMST block code is given, which reduces the number of symbols within each candidate symbols set and gains higher detection efficiency in the iterative process by utilizing TCM set partition principle and trellis transfer restrictions.The necessary condition for symbol transmit rate which maximum diversity and channel capacity of linear space-time code are attained is given and a new systematic design framework of space-time block code is proposed. The space-time block code constructed by the design framework is capacity lossless and full-diversity, that is, it can achieve the full transmit diversity and the capacity of the equivalent channel including the space-time code equals the MIMO system capacity. It is proved that the proposed design framework subsumes many of the existing capacity lossless and full-diversity space–time codes.Signal detections of MIMO systems are investigated. Sphere detection algorithms are emphasized. Based on the analysis of the factors influencing computational complexity in sphere detection algorithms, a QR factorization with exchanged columns pre-processing sphere detection algorithm is brought forward, the new algorithm optimizes the detection order of components to reduce the average complexity of iterative search while the pre-processing is hardly increased the computational complexity. A low-complexity near-optimum sphere detection algorithm is presented by estimating detection ranges of each component with the lower bound of each component's error probability, which greatly reduces average computational complexity at the cost of a little performance degradation.