Performance Evaluation of the SRS Based MIMO Channel Estimation on 5G NR Open Source Channel Model

Abstract

In 5G NR sounding reference signals are used for pilot-aided uplink MIMO channel estimation. Obtained channel responses due to channel reciprocity are processed further to compute downlink MU-MIMO beamforming vectors. The resulting performance of beamforming depends of accuracy obtained in channel response estimation. In this work typical scenario of massive MIMO deployment in the open source 5G channel model QUADRIGA is configured. Least square channel estimation based on SRS under pilot contamination from neighbouring cell are performed and MSE compared.