Research On Channel And Covariance Matrix Estimation For MIMO Wireless Communication Systems

As one of the key technologies of LTE, MIMO can improve the spectrum efficiency and increase the system capacity, which has a broad prospect of application. In MIMO wireless communication systems, channel state information plays a crucial role in coherent demodulation, space-time detection and adaptive transmission. The accuracy of channel estimation is closely related to the system performance. In order to improve the efficiency of spectrum, modern mobile communication systems reuse the same frequency in different cells, which will encounter serious impairment due to co-channel interference from neighboring cells. Therefore, this thesis will mainly focus on channel and interference covariance matrix estimation method for MIMO wireless communication systems, and the main work of this thesis including the following three parts:This thesis propose a approach based on compressive sensing for channel estimation for MIMO-OFDM systems equipped with 2-dimensional active antenna arrays, which jointly estimate the delay, elevation angle and azimuth angle of the three dimensional space-time channel. In order to reduce the complexity of the algorithm, a multi-step estimation scheme is proposed:first, the sparse of channel in delay-domain can be utilized and the positions of non-zero taps in time domain can be found by applying compressive sensing theory. And then, the channel can be regarded as sparse in angle domain and the azimuth angle and elevation angle of each path can be estimated by compressive sensing theory. Finally, the MIMO channel matrix can be estimated by substitution of estimated parameters into the path-based space-time channel model. Simulation results show that the proposed method achieves more effective channel estimation performance and reduces pilot overhead compared with least square estimation.This thesis propose to estimate channel frequency response matrix and interference-plus-noise spatial covariance matrix (SCM) jointly for MIMO-OFDM systems with co-channel interference under maximum a posteriori probability criterion, which effectively reduces the estimation error and bit-error-rate significantly after employing interference rejection combining at the receiver side. Further, two schemes are proposed to handle with SCM by utilizing the correlation and rank of SCM. Both the two schemes not only keep the semi-definite property of obtained SCM, but also decrease the the number of parameters to be estimated, which in turn improves the accuracy of SCM estimation. At the same time, the number of interferers and the length of interference channels are estimated under minimum description length criterion.The performance of MIMO systems employing interference rejection combining with imperfect channel estimation and interference spatial covariance estimation is investigated in terms of the system ergodic rate. By using random matrix theory, closed-forms of ergodic rates are presented without any numerical integra-tions, including the case of:a) ideal channel and ideal instantaneous SCM, b) estimated channel and ideal instantaneous SCM, c) estimated channel and estimated instantaneous SCM, d) estimated channel and ideal long-term SCM. Simulation results are provided to demonstrate the consistency between analytical results and Monte-Carlo simulations. Finally, performance analysis are given for MIMO systems with channel and covariance estimation errors.