CQI Estimation in 5G LEO Satellite Communications Under Ultra-Low SNR Conditions

Due to large path losses between low earth orbit (LEO) satellites and the ground User Equipment (UE), satellite networks operate mainly in low signal-to-noise ratio (SNR) regimes. Assuming a multi-antenna receiver, the maximum ratio combining (MRC) technique is a well-known method to maximize the receiver output SNR. We consider a UE equipped with an MRC receiver where the selection of channel quality indicator (CQI) depends upon post-MRC SNR estimates. However, these estimates are not accurate in practice due to combining losses and SNR estimates at the combiner input. It has been shown that if the channel estimation is imperfect, mean and variance of combiner output SNR probability distribution function (PDF) depend upon the channel estimation gain and combiner input SNR estimates. Therefore, the choice of SNR estimator for combiner input SNR estimates is fundamentally important. For this purpose, a detailed investigation of existing SNR estimators and the associated bias issue is carried out to analyze the robustness of combiner input SNR estimates. In the presence of combining losses, proposed SNR estimation demonstrates negligible bias to the actual reference post-MRC SNR. Consequently, the CQI estimation and its reporting to the network is improved.