An Efficient Massive MIMO Channel Reconstruction Method Based on Virtual Scatterer Map

Abstract

In this letter, we introduce a novel concept called the virtual scatterer map (VSM) to reduce both the pilot overhead and computational complexity of massive MIMO channel reconstruction. The VSM stores the locations of the dominant virtual scatterers that can be seen by users at some predetermined user grid points spread over the cell. The VSM can be constructed by offline measurements or ray tracing simulation. Based on the VSM, we propose an efficient channel reconstruction method with reduced pilot overhead and computational complexity. Specifically, for a target user, the virtual scatterers (prior information about the scattering environment) are estimated by clustering and weighting those from neighboring user grid points in the VSM, and then a VSM-assisted turbo variational Bayesian inference (Turbo-VBI) algorithm is proposed to efficiently fuse the prior information provided by the VSM and the observations from a few uplink pilots to achieve low-overhead, low-complexity, and high-accuracy channel reconstruction. In addition, the paths (virtual scatterers) missed by the VSM can be automatically detected by the VSM-assisted Turbo-VBI algorithm. Simulation results show that our proposed method can effectively reduce the pilot overhead.