Exploiting correlation for MMSE channel estimation in massive MIMO systems with RF-impairments

The Large Scale Antenna System is expected to bring significant improvements to overall Communication System Performance. However the use of so many antennas in Massive Multiple Inputs Multiple Outputs (Massive MIMO) architectures highlights the cost implication of the technology. It is therefore possible that Massive MIMO may be deployed using low cost equipment to avoid HUGE implementation costs. The use of Low Cost Equipment will results into significant Radio Frequency Impairments (RF-Impairments) such as Phase Noise, IQ Imbalance and Nonlinearities, which will affect the overall system performance. Among other aspects channel estimations will also be affected since the signal used for these estimations will be affected by the RF-Impairments. This paper analyzes a way to exploit correlation during Minimum Mean Square Error (MMSE) Channel Estimation of a Massive MIMO System in presence of aggregated RF-Impairments (i.e. combined effects of the impairments). Specifically it derived the statistical distribution of the equivalent aggregated error/noise term using statistical identities and uses the derived term to obtain temporal channel estimation error and consequently the correlation matrix necessary to minimizes the channel Mean Square Error (MSE) of the channel estimate.