Research On Pilot Contamination Of Massive MIMO

MIMO(Input Multiple Output Multiple)is a key technology which can greatly improve the spectrumefficiency while keeping the power efficiency constant.In the current 4G system,MU-MIMO(MIMO Multi-User)and cooperative MU-MIMO are constrained by channel state information acquisition and information exchange between the base stations,which affects the spectrum and power efficiency.The massive MIMO system,which is equipped with enough antennas at the base station,can serve multiple users occupying the same time-frequency resource.As the number of base station antenna tends to infinity andthe transmitting power of the user can be kept arbitrarily low,the capacity of a single user is only limited by the interference from the users which use the same pilot(i.e.,pilot contamination)with the multiuser channel tending to be orthogonal and the Gaussian noise and uncorrelated inter cell interference disappearing.With the increase of the base station antennas,pilot contamination has become the main factor that limits the capacity of the massive MIMO system.This is the most challenging problem for system design.On the one hand,pilot contamination reduces the accuracy of the uplink channel estimation,thereby affecting the performance of the downlink precoding.On the other hand,as the number of base station antennas can never be infinite,the interference of the adjacent cells,especially the boundary users,also greatly influences system performance.In this paper,new algorithms are proposed to decrease the influence of pilot contaminations in the massive MIMO systems based on the characteristics of different scenarios.The main work and results are as follows:1)While analyzing the cause of pilot contamination,we get that pilot reuse in the adjacent cells results in the fact that base station cannot distinguish the user with the same pilot mode.In this paper,a new pilot scheme based on pseudo-random code is proposed,using different phase pseudo-random sequence to distinguish the pilot using the same orthogonal codes between different cell users by scrambling,which reducesthe pilot contamination.The formulaof the mean square error of the expected channel estimation under the pilot design schemeis also derived.Numerical simulation results show that this new pilot design scheme based on pseudo-random sequence can effectively improve the downlink transmission capacitywith low complexity,in the consequence of the reduce of the influence of the pilot contamination.2)Analyzing the system downlink signal to interference plus noise ratio formula,we can draw a conclusion that the main factor affecting the performance of the downlink system is the large scale fading as the number of base station antenna tends to infinity.Considering that different characteristics of large scale fading coefficients between different users and base stations,and that the boundary interference is much stronger than others in the neighing cell,a dynamic pilot allocation algorithm based on signal to interference ratio is proposed with all cells reuse the same pilot.Theoretical analysis shows that this pilot allocation algorithm can obtain optimum downlink signal to interference plus noise ratio.In addition,the pilot design algorithm in the presence of additional orthogonal pilot groups is also proposed.Numerical simulation results show that the proposed dynamic pilot allocation can effectively improve the downlink performance of massive MIMO systems.