Joint Tx/Rx MMSE Beamforming Design for Multi-user MIMO-OFDM SDMA Downlink System under Channel Estimation Error

• Main Authors: Wei-Kai Lin, 林偉凱
• Other Authors: Hsiang-Feng Chi
• Format: Others
• Language: en_US
• Published: 2006
• Online Access: http://ndltd.ncl.edu.tw/handle/50600253533201458027

碩士 === 國立交通大學 === 電信工程系所 === 94 === The techniques of MIMO-OFDM SDMA have recently considered one of the candidates for the broad-band multiple-access wireless systems. In this thesis, we focus on the research of beamforming in the applications of multi-user MIMO-OFDM SDMA downlink wireless system. Many researchers have used the joint Tx/Rx beamforming approach in the MIMO systems. However, they only considered the single user system. In addition, their methods are proposed under the perfect channel estimation assumption, which is no true in the real world. In this thesis, we will aim at multi-user MIMO SDMA systems to propose a MMSE beamforming algorithm, which has better immunity to the channel estimation errors than the previously published methods. The joint Tx/Rx beamforming technique can transform a mutually cross-coupled MIMO transmission system into a set of independent eigen subchannels so that subchannel signal can be decoded without the interference from other antennas. In our proposed MIMO-OFDM SDMA downlink system, by using null-space technique in the base station (Tx side), the multi-user MIMO downlink system can be simplified into several independent single user MIMO subsystems. The MMSE beamforming algorithm is further applied to each single user MIMO subsystem achieving the spatial diversity gain. In addition, to combat the frequency-selective channel caused by multi-paths, we focus the aforementioned problems in the MIMO-OFDM systems. However, all the previously proposed methods are derived under the assumption of perfect channel estimation. But in real wireless systems, the channel estimate always contains estimation errors which may cause significant performance degradation when these methods are used. The performance loss can be imputed to the multi-user interference induced by the imperfect null-space matrix and the inter-stream interference caused by imperfect Tx/Rx beamforming. In this thesis, without the assumption of perfect channel estimation, we will consider the joint Tx/Rx beamforming design for multi-user MIMO-OFDM SDMA downlink system which is more robust to channel estimation error than the previous methods. Finally, the performance of the proposed schemes will be evaluated by computer simulations.