Optimization Study Of Uplink Pilots For Massive MIMO Systems

Along with the rapid popularization of smart devices,mobile data traffic has grown quickly in recent years.In massive multiple-input multiple-output(MIMO)systems,base stations(BSs)are usually equipped with excess antennas while comparing with the number of active terminals.By exploiting these extra antennas,massive MIMO technology can increase spectral efficiency 10 times or more and simultaneously improve energy efficiency on the order of 100 times.In addition,it has the ability to reduce the latency on air interface,simplify multiple access layer,reduce hardware costs and increase system robustness.Therefore,massive MIMO technology has become a key research hotspot for 5th generation mobile communication system(5 G).In massive MIMO systems,the accurate channel state information(CSI)for both uplink and downlink is crucial to achieve a high-quality communication.However due to coherence interval length limitation,the number of orthogonal pilots is usually insufficient compared with that of active terminals.In this case,the orthogonal pilot assignment cannot be carried out.When one pilot sequence is reused by several terminals in the system,pilot contamination problem emerges.And the accurate uplink and downlink CSI cannot be obtained by BSs under pilot contamination.Research shows that even with an unlimited number of BS antennas,pilot contamination does not vanish and becomes one of the major bottlenecks limiting system performance.This paper focuses on the optimization study of uplink pilots for massive MIMO systems.Specifically,the main contents and innovation points are as follows:Firstly,this paper discusses the optimization study of uplink pilots for single-cell massive MIMO systems with spatially independent channels.With non-orthogonal pilots,closed-form expressions of the channel estimation error and system capacity are derived for both least squares(LS)and minimum mean square error(MMSE)estimation.For optimal channel estimation error performance,a pilot signal reconstruction scheme is analyzed in the case of sufficient and insufficient pilot length.With sufficient pilot length,pilots can be reconstructed to satisfy the orthogonal condition.With insufficient pilot length,an orthogonal procrustes principle based(OPP-based)method is proposed to reconstruct pilots such that the correlation is as small as possible.To improve the quality of service(QoS)of cell edge users,a pilot assignment method is formulated by considering the quality of pilots.Specifically,it will always try to avoid the situation that the pilot with low quality is assigned to the terminal with small large-scale fading gain.For optimal system capacity performance,a step-by-step pilot power allocation scheme is further proposed.At last,simulation results demonstrate the superiority of the proposed approaches.Secondly,this paper discusses the optimization study of uplink pilots for single-cell massive MIMO systems with spatially correlated channels.A spatial correlation between two channel response vectors is established by the large-scale fading variables and the angle of arrival(AOA)information when the amount of BS antennas tends to infinity.When the channel AOA intervals associated with distinct terminals are non-overlapping,the corresponding channel propagation vectors can be considered orthogonal.In this case,expressions of the system capacity are derived.For optimal system capacity performance,a low-complexity pilot assignment algorithm corresponding to the distinct channel AOA information is proposed.On the one hand,orthogonal pilots should be allocated to the terminals with high AOA similarities.On the other hand,orthogonal pilots should be allocated to the terminals with large AOA spans.By this means,the intra-cell interference caused by pilot reuse can be reduced as much as possible.At last,simulation results prove that the proposed pilot assignment algorithm can obtain a noticeable performance gain with limited BS antennas.Thirdly,this paper discusses the optimization study of uplink pilots for multi-cell massive MIMO systems with spatially correlated channels.Compared with single-cell circumstances,pilot contamination in multi-cell circumstances is usually more severe.When the number of orthogonal pilots is less than that of active terminals in one cell,expressions of the system capacity are derived.To mitigate intra-cell interference,a bottom-up user grouping method is formulated.Specifically,it progressively merges the individual terminals with the largest AOA similarity.After the user grouping method is adopted in all cells,a group matching algorithm is proposed to divide the user groups in the system into different collections.Then a graph coloring based pilot allocation algorithm is proposed to alleviate inter-cell interference.On the one hand,ensure that one pilot cannot be reused in one group.On the other hand,two terminals in different groups may reuse the same pilot only when the potential interference is small enough.At last,simulation results show that the proposed approach can effectively improve the complexity-performance trade-off with finite number of BS antennas.