Prediction of SINR using BER and EVM for Massive MIMO Applications

TL;DR

This paper demonstrates how BER and EVM measurements can predict SINR in massive MIMO systems without channel state information, aiding in evaluating precoder performance in dynamic environments.

Contribution

It introduces a method to estimate SINR from BER and EVM measurements in massive MIMO systems, validated through real-world channel sounder data.

Findings

BER and EVM can accurately predict SINR levels.

The method works for both average and dynamic SINR conditions.

Validation conducted with 32x3 MIMO measurements at 2.4 GHz.

Abstract

Future communication systems employing massive multiple input multiple output will not have the ability to use channel state information at the mobile user terminals. Instead, it will be necessary for such devices to evaluate the downlink signal to interference and noise ratio (SINR) with interference both from the base station serving other users within the same cell and other base stations from adjacent cells. The SINR will act as an indicator of how well the precoders have been applied at the base station. The results presented in this paper from a 32 x 3 massive MIMO channel sounder measurement campaign at 2.4 GHz show how the received bit error rate and error vector magnitudes can be used to obtain a prediction of both the average and dynamically changing SINR.