Evaluation of Massive MU-MIMO Channel Estimation Based on Uplink Achievable-Sum Rate Criteria

In previous work, the nuclear norm (NN) and iterative weighted nuclear norm (IWNN) estimation methods have been previously proposed for single and multi-cell time division duplex (TDD) massive multiuser multi-input multi-output (MU-MIMO) systems. In this paper, the uplink achievable-sum rate (ASR) performance metric is used to evaluate the effectiveness of NN and IWNN proposed estimation methods for these systems. To investigate the above, a minimum mean square error (MMSE) detector is used to detect the uplink data received at each base station (BS). The simulation results in both single and multi-cell systems show that the uplink ASR performances obtained by proposed estimation methods are improved compared to the conventional least square (LS) method as the number of antennas increase. Also, the impact of the pilot contamination on the uplink ASR performance in multi-cell setting is studied. The simulation results show that the uplink ASR performance obtained by IWNN method outperforms both NN and LS estimation methods even in the strong pilot contamination.