无蜂窝大规模MIMO系统信道估计中基于加权图的叠加导频分配

Weighted graph based superimposed pilot assignment for channel estimation in cell free massive MIMO systems

针对无蜂窝大规模多输入多输出(cell-free massive multiple-input multiple-output, CF-mMIMO)系统信道估计中导频污染问题,提出了基于加权图的叠加导频(superimposed pilot, SP)分配方案。首先分析导频污染对SP信道估计的影响,引入一种全新度量表示用户间潜在导频污染程度;其次根据接入点与用户间大尺度衰落系数完成用户加权干扰图构建,将系统吞吐量最大化问题转化为有容量最大k切割问题求解;最后采取顶点交换局部搜索算法实现SP次优分配。仿真结果表明:本文方案能够有效地改善SP信道估计误差以及系统吞吐量。

In order to solve the problem of pilot contamination in channel estimation of cell-free massive multiple-input multiple-output(MIMO)systems,this paper proposes a scheme of weighted graph based superimposed pilot assignment.Firstly,the influence of pilot contamination on superimposed pilot channel estimation is analyzed,and a new metric is introduced to represent the potential pilot contamination level among users.Secondly,the user-weighted interference graph is constructed according to the large-scale fading coefficient between the access points and the users,and the throughput optimization problem is transformed into the capacitated max k-cut problem.Finally,vertex swapping local search algorithm is adopted to realize the suboptimal allocation of superimposed pilot.Simulation results show that this scheme can effectively reduce pilot contamination,improve the estimation error of channel based on superimposed pilot and system throughput.