Multiuser CQI prediction based on quantization error feedback for massive MIMO systems

In this paper, multiuser channel quality indicator (MU-CQI) prediction based on channel quantization error feedback is proposed for massive multiple-input multiple-output (MIMO) systems. Basically, CQI reporting in the practical systems is performed based on the single user (SU) MIMO transmission mode to keep low feedback overhead and computational complexity. To operate in MU-MIMO transmission mode, the base station needs to predict the MU-CQI from the reported SU-CQI. The conventional CQI update method is based on the worst-case approach to minimize the outage probability. However, this is not efficient way for the massive MIMO systems, since there is a lot of capacity loss when the number of served users is large. To improve the accuracy of prediction, we propose MU-CQI prediction method based on channel quantization error feedback. Although the proposed scheme requires few feedback overhead like only 1bit, it shows remarkable performance gain compared to the conventional prediction scheme in terms of system goodput.