The Research On The Suboptimal Transmission Scheme And Channel Feedback In Multiuser MIMO-OFDM System

MIMO-OFDM system is regarded as one of the most potential solutions for the future wideband wireless communication systems for its fully exploitation of the advantages of both MIMO and OFDM. In multiuser communication systems, the channel state information of users should be fed back to the base station so as to implement multiuser spatial multiplexing and extract multiuser diversity, the feedback overhead is a heavy burden in the system accordingly, so the scheme design of channel information feedback with high efficiency is an extremely important work for multiuser MIMO-OFDM systems, which is closely related with the transmission schemes in practical implementations. As the optimal transmission scheme is difficult to implement in practical systems, this thesis mainly researches suboptimal transmission schemes and limited feedback.Firstly, a random beamforming scheme with beam sets selection is presented, which preserves advantage of orthogonal random beamforming with low complexity and low feedback overhead, it can effectively improve the performance of orthogonal random beamforming for practical numbers of users.Secondly, the impact of channel estimation error on the sum-rate of ZFBF transmission is analyzed. The thesis derives the upper-bound expressions of the rate loss due to channel estimation error in the case of ideal feedback and limited feedback respectively, the results indicate that the ZFBF transmission scheme is robustness to channel estimation error. The thesis also presents the approximate expression of optimal training intervals in the case of ideal feedback, simulation results testify the validity of the expression.Finally, a time-domain feedback scheme for multiuser MIMO-OFDM is presented. Compare with limited feedback scheme in frequency-domain, the scheme has two advantages: it has much less feedback overhead and lower implementation complexity due to scalar quantization. Simulation results show that the scheme can efficiently reduce the feedback overhead in multiuser MIMO-OFDM systems.