Adaptive Transmission And User Scheduling For Large-scale MIMO Downlink System Over Rician Fading Channel

In recent years,massive multiple-input multiple-output(MIMO)has been widely recognized as one of the key techniques to improve capacity and reliability in wireless communication systems,and has attracted significant attention in both industry and academia.Massive MIMO is an effective method to improve transmission rate and coverage.It can greatly reduce the transmission power,improve the system spectrum utilization,and realize green communication by using low-cost and low-power devices.Note that most prior works on massive MIMO adopt the Rayleigh fading channel model.Although this channel model assumption greatly simplifies the analysis,it can not capture the characteristics of the line-of-sight(LOS)component between the transmitter and the receiver.This is especially the case in millimeter-wave communication systems in which LOS propagation dominates.Therefore it is of great importance to consider the more general fading channels which take into account the LOS conditions,i.e.,Ricican fading channels.In this thesis,the adaptive transmission and user scheduling algorithms are investigated for massive MIMO downlink systems over Rician fading channel.Firstly,the downlink adaptive transmission exploiting only statistical channel state information(CSI)is implemented for single cell full-dimension MIMO(FD-MIMO)downlink transmission systems over uncorrelated Rician fading channels with uniform planar antenna(UPA)array at the BS.Based on the asymptotic properties of the channel's LOS component,the optimal beamforming vector for each user is derived by maximizing an approximation of the ergodic sum rate.Furthermore,a low complexity three-dimensional(3D)beamforming transmission algorithm is introduced.Simulation results show that the downlink adaptive transmission algorithm performs well in achieving considerable sum rate while requiring much less CSI at the BS.Next,based on the above adaptive transmission algorithm for downlink FD-MIMO system,user scheduling algorithms taking into account both achievable rate and fairness are investigated.An approximation of user's signal-to-interferenceplus-noise ratio(SINR)and a lower bound of user's average signal-to-leakageplus-noise ratio(SLNR)are derived.Then,by maintaining a priority vector,a SINR and a SLNR based user scheduling algorithms which take into account both the achievable rate and fairness are proposed.Furthermore,in order to reduce the computation complexity of scheduling algorithms,two simplified user scheduling algorithms are proposed.Simulation results show that the proposed user scheduling algorithms can make good trade-off between the achievable sum rate and fairness,and the simplified user scheduling algorithms can greatly reduce the computation complexity while cause slight performance degradation.Then,the downlink adaptive transmission and user scheduling algorithm for massive MIMO downlink transmission systems under correlated Rician fading channels are investigated.Using Mullen's inequality,a closed-form lower bound of each user's average SLNR is derived.By maximizing this derived lower bound,the statistical beamforming transmission is derived.Furthermore,based on the beamforming transmission algorithm,three user scheduling algorithms exploiting only statistical CSI are proposed: 1)maximum SLNR:schedule the user with the maximum SLNR;2)most dissimilar: measure the similarity between different users and schedule the user that is most dissimilar to the already selected users;3)maximum signal-to-interference factor: treat the interuser interference as uncorrelated noise to each user's useful signal and schedule the user with the largest signal-to-interference factor.Simulation results show that the proposed user scheduling algorithms perform well in achieving considerable sum rate.Finally,the above downlink adaptive transmission and user scheduling algorithm are extended to the FD-MIMO system under correlated Rician fading channels.First,the design of beamforming vector for each user is derived by maximizing the lower bound on the average SLNR and low-complexity sub-optimal algorithm to obtain the beamforming vector is proposed.Then,based on this beamforming vector,three user scheduling algorithms are proposed with only statistical CSI: maximum SLNR,most dissimilar and maximum signal-to-interference factor.Furthermore,a tractable approximation of the ergodic sum rate is derived.Simulation results show that the sub-optimal beamforming algorithm performs almost the same with the optimal algorithm,and the proposed user scheduling algorithms perform well in terms of achievable sum rate while requiring much less computation complexity compared to the the sum rate based user scheduling.