Kalman Filter based Tracking for Channel Aging in Massive MIMO Systems

We consider the problem of channel aging in massive multi-user multiple-input-multiple-output (MIMO) wireless systems using the Kalman filters without knowledge of channel correlation statistics. Although Wiener filtering can be used to estimate the channel state information to construct accurate precoders and decoders, their use is limited in the case where regular adaptation to channel changes is needed. In addition, since different users exhibit different mobility conditions, use of an adaptive Kalman filter is advantageous when compared to conventional estimation approaches. We build a general model that allows a wide range of user mobilities in each cell. Simulations using the model reveal that the Kalman filter is effective in predicting channel coefficients to build an effective precoder. In addition, we perform both uplink and downlink analysis and show that with poorer channel tracking, the achievable rates in massive MIMO diminish unless more accurate channel estimates are present.