Channel Charting-Based Pilot Assignment over Spatially Correlated Channels

Abstract

Massive multiple-input multiple-output (MIMO) is a promising technology to improve spectral efficiency (SE). However, because of limited pilot resources for channel estimation, pilot reuse among a large number of users is inevitable. It causes pilot contamination and will limit the system's performance. This paper proposes a channel charting (CC)-based pilot assignment (PA) method that assigns orthogonal pilot sequences to adjacent users on CC greedily to alleviate pilot interference with spatially correlated channels, where adjacent users on CC have stronger angular correlation. Moreover, a feature is explored from the second-order channel statistics for CC to extract more effective angular correlation among users. Pilot contamination is further mitigated by choosing the first assigned user which is prone to interference. Simulation results show that the proposed method has superiority over other PA methods, especially under the scenario that a large number of users need to be served.