In this thesis, we investigate the following three fields on multi-input multi-output (MIMO) systems with limited feedback. End-to-end distortion: The first part of the thesis presents the joint impact of antenna numbers, source-to-channel bandwidth ratio, spatial correlation and time diversity on the optimum expected end-to-end distortion in an outage-free MIMO system. In particular, based on the analytical expression for any signal-to-noise ratio (SNR), the closed-form expression of the asymptotic optimum expected end-to-end distortion at a high SNR is derived, comprised of the optimum distortion exponent and the optimum distortion factor. The simulation results illustrate that, at a practical high SNR, the analysis on the impacts of the optimum distortion exponent and the optimum distortion factor explains the behavior of the optimum expected end-to-end distortion. The results in this part could be the performance objectives for analog-source transmission systems as well as a guidance on system design. Analog channel feedback: In the second part of this thesis, we propose to apply orthogonal space-time block codes (OSTBC) with linear analog channel feedback. Since MIMO channel information is a sort of analog source vector, relative to quantized channel feedback, linear analog feedback has the advantages such as outage-free, self channel adaptation and low complexity. It is proved that the linear analog transmission method with OSTBC can achieve the matched filter bound (MFB) on received SNR. In comparison with the linear analog transmission method with circulant space-time block coding (CSTBC), the method with OSTBC performs better with respect to received SNR and mean-squared error. In comparison with the random vector quantization methods with different modulation schemes, the simulation results show that with respect to average direction error, the linear analog transmission method with OSTBC performs over any RVQ method with specific modulation scheme in the regimes of relatively high SNR and low SNR; with respect to average mean-squared error, it performs always better than the RVQ methods. We also evaluate the effect of applying the linear analog channel feedback with OSTBC to multiuser MIMO downlink beamforming. It is shown that the linear analog channel feedback with OSTBC can make the system approach the optimum performance within a short latency. Layered multiplexing: In the third part of this thesis, with respect to the systems with short blocks, a new layered multiplexing strategy is proposed to adapt an uncertain channel by Walsh layer-time coding, successive interference canceller and HARQ signaling. As illustrated by simulation results, with respect to its high success rate, good performance on average latency and lower computational complexity, this strategy would be a good replacement to the widely-used adaptive QAM modulation strategy.
