Research On Key Technologies Of Multiuser MIMO Downlink

With the development of society, the demand for high-speed wireless data services has increased. MIMO has become one of the key techniques of the next generation wireless communication systems. Different from point-to-point communication of single-user MIMO, multiuser MIMO is a point-to-multipoint communication system. It can not only increase the data rate and reliability of the single link, but also improve the entire system capacity.In multiuser MIMO downlink communications, in order to eliminate the multi-user interference, the base station needs to precode the transmitted signals. While in the user end, it needs to perform the channel estimation and signal detection to receive data correctly. Multiple users and multiple antennas of the system increase the computational complexity of signal processing of the transmitter side and receiver side. How to achieve a trade-off between performance and complexity is a key issue in algorithm design.The aim of this thesis is to try to solve the above problem. It consists of three parts1. We propose two low complexity precoding algorithms for multiuser downlink. One is based on generalized zero-forcing and Gram-Schmidt orthogonalization, the other is based on LQ composition. Owing to avoiding SVD operation, the two methods have lower complexity than the conventional BD algorithm, while there is no any degradation in sum throughput.2. Two ordered successive interference cancel (OSIC) algorithms based on grouping are proposed. In the two methods, the received signals are divided into several groups. Maximum likelihood idea is used to reduce the error in each group. Interference cancel is performed among groups. Analyses and simulations show that the two methods not only suppress the error propagation problem existing in the conventional OSIC algorithm, and improve the detection performance, but lower the complexity by avoiding the pseudo inverse and ordering operations. With more antennas and higher order modulation, the advantage of low complexity of these two methods can be more obvious.3. A practical channel estimation algorithm based on minimum mean-squared error (MMSE) is proposed. In the proposed scheme, only the autocorrelation matrix of the received signal samples and the estimated noise power are utilized to obtain the channel matrix information. The drawback of the conventional MMSE algorithm that the knowledge of channel statistics and SNR is needed to be known beforehand is overcome. The proposed algorithm has a lower computational complexity than the conventional LMMSE algorithm. The simulation results show that it can achieve almost the same performance as the conventional LMMSE algorithm.