Channel Estimation For MIMO Wireless Communication Systems In Double Selective Fading Channels And Spatially Correlated Channels

MIMO techniques are key schemes for achieving high data rate and performance infuture mobile communication systems. The need for high block broadband transmissionrates and high terminal mobility results in that wireless MIMO communication channelspossess two characteristics: time-selectivity and frequency-selectivity. Therefore, doublyselective fading channel model will be encountered in future MIMO communication systems.However, there still exists certain spatial correlation and pin-hole effect between transmittingand receiving antenna array elements, which induces wireless MIMO fading channel modelto be more complexity. Accordingly, there are also numerous difficulties to be resolvedamong MIMO spatio-temporal signal processing referred to. It is great important to estimateand track MIMO channels, for that the accuracy of channel estimation has much effect onthe performance of the whole communication system. As the number of antennas increases,channel estimation becomes challenging because the number of unknowns increases, and thepower is split at the transmitter.In this dissertation, the research mainly focuses on the topics of channel estimation forMIMO wireless communication systems in double selective fading channels and spatiallycorrelated channels. The main contents are as follows:The problem of channel estimation and tracking for MIMO time-varying frequency-selective channels is proposed based on training sequences and Sequential Monte Carlo filter.Relying on the time evolution of time-varying wireless channel, a low order autoregressivevector model approximates the MIMO channel variation and facilitates time-varying channeltracking. Sequential Monte Carlo Filter is utilized to track MIMO doubly selective channels,and MMSE-DFE schemes are employed for symbol detection, which is compared with ex-tended Kalman filter method.MIMO time-varying frequency-selective channel estimation is investigated using pilotsymbol aided modulation (PSAM) for data block transmissions. Firstly, a complex exponen-tial basis expansion model is introduced to represent doubly selective MIMO channel; thenbased on basis expansion channel model, MIMO doubly selective channels are estimatedusing PSAM. Finally, based on this model, the optimality in designing PSAM parameters isdone according to the minimization of channel estimation MSE and symbol detection MSE,the maximization of average channel capacity schemes, mainly including placement of pi-...