Linear Precoding Technologies For MIMO Systems In Mobile Communications

MIMO techniques are key schemes for achieving high data rate and performance in 3G mobile communications and beyond. In this dissertation, we mainly focus on the topics of linear space-time precoding. The main contents are as follows:1. Design of Linear dispersion codes in flat Rayleigh fading channels. Firstly, we study the linear dispersion codes design in flat fast fading channels. We analyze the pairwise error probability of linear dispersion codes in fast and quasi-static fading channels, and compare the diversity order and code gain of linear dispersion codes in fast and quasi-static fading channels. It is proved that the performance of linear dispersion codes is better in fast fading channels. We deduce the minimum code gain bound of linear dispersion codes in fast fading channels, and determine the optimization criterion to search optimal code according to the minimum code gain bound. Simulations demonstrate that the linear dispersion codes obtained by computer search have good performance. Secondly, a linear dispersion codes transmit scheme with symbol interleaver is proposed for general time correlated flat fading channels. By introducing a symbol interleaver after encoder of linear dispersion codes, the original system model can be regarded as equivalent system model in fast fading channels. Thus, the optimization of linear dispersion codes in general time correlated flat fading channels is turned to the optimization for flat fast fading channels.2. Design of space-frequency linear dispersion codes for MIMO-OFDM systems. Grouping space-frequency linear dispersion code is proposed for MIMO-OFDM systems in frequency selective fading channels. We analyze the pairwise error probability of grouping space-frequency linear dispersion codes and give thesubcarrier grouping method based on the channel frequency domain correlation. For MIMO-OFDM systems in frequency selective channels, the optimization design of grouping space-frequency linear dispersion codes is turned to the optimization for flat fast fading channels.3. Precoding based on code distance allocation for spatial multiplexing MIMO systems. Based on ML detection, subchannel code distance allocation precoding is proposed for spatial multiplexing MIMO systems. We discuss the precoding algorithms in two cases. When full channel information is available at the transmitter, we give the structure of linear precoding matrix. The analysis shows that pairwise error probability of precoded spatial multiplexing MIMO systems is determined by the minimum code distance allocated to each subchannel. The construction method and optimization criterion of code distance allocation matrix is given. Equal gain allocation and linear gain allocation precoding methods are discussed. In spatial fading correlated MIMO channels, we also give the subchannel code distance allocation precoding algorithms when spatial fading correlation is known at the transmitter.4. Linear precoding for V-BLAST based on OSIC detection. For spatial fading correlated MIMO channel, we first propose decorrelation and power allocation joint precoding to improve BER performance of V-BLAST systems. Another direct matrix precoding algorithm is also proposed for V-BLAST systems using OSIC detection to optimize precoding matrix directly based on pairwise error probability. We give the search methods for optimal precoding matrix for V-BLAST system with two and more transmit antennas respectively...