The Research On Channel Estimation And Performance Analysis In Massive MIMO System With Pilot Contamination

With the development of 5G communication technology,massive MIMO system has received extensive attention and research due to its high data rate and effective suppression in channel fading.Massive MIMO system is equipped with a large number of antennas at the base station to communicate with users,and improving spatial multiplexing capabilities.In order to perform data transmission more efficiently,the base station needs to obtain channel state information between the users and itself.However,due to the increasing number of users and the limited channel coherence time,there must be user pilot multiplexing between different cells.The phenomenon that leads to pilot contamination between users.Pilot contamination affects the quality of channel estimation and becomes a bottleneck problem that further improves system performance.In this thesis,the uplink regular transmission model of massive MIMO system is used as the entry point.The regular transmission process is introduced.During the pilot transmission phase,Users use the pilot sequence to transmit the pilot signal.Through the received pilot signal,the base station uses the least square or minimum mean square error techniques to estimate the channel state information.During the data transmission phase,the base station detects the data signal sent by the user according to the obtained channel estimation,and then acquires the data signal.Due to the influence of pilot contamination and noise,the channel estimation always has errors,and the quality of the channel estimation will affect the data signal.The quality of the reception further affects system performance.In order to improve the quality of channel estimation,this thesis re-allocates the pilot signal and pilot power used by the users for the pilot signal transmission phase.The re-allocation of pilot sequences by sorting users reduces the negative impact of pilot contamination between different cell users on channel estimation.And in the case where the total pilot power in the target cell is constant,the target cell does the pilot power allocation through user grouping.In this thesis,the above-mentioned pilot sorting power allocation method,pilot power allocation and equalizing pilot power allocation are compared,and the differences of the three allocation methods on channel estimation are studied.Pilot power and data power are important factors affecting system performance improvement.Under the condition of the available pilot power and data power certain,the reasonable pilot-to-data power ratio design in the target cell will become an important factor for system performance improvement.In order to reasonably design the pilot-to-data power ratio of users in the target cell,this thesis proposes a user joint optimization algorithm to rationally design the ratio,and compares the design with the same pilot-to-data power ratio in the case of different pilot contamination levels.Then,the optimal pilot-to-data power ratio matching between users in the cell is sought to further improve system performance.In the current regular transmission mode,pilot transmission and data transmission are separately transmitted in the uplink transmission phase.Based on the regular transmission mode,the superimposed pilot transmission mode is analyzed.Thesis compares the regular transmission mode with the superimposed pilot mode in channel estimation and spectral efficiency.