Spatially-Correlated Massive MIMO Channel: An Iterative Estimation Algorithm

In this letter, we propose an iterative algorithm for the joint estimation of the channel covariance matrices and the channel vector for the spatially-correlated massive MIMO system with the following main advantages: low pilot overhead, improved mean square error (MSE), and low complexity. The results show that this scheme outperforms the conventional estimation schemes in terms of the required pilot overhead and the resulting MSE. The results also show that the achievable sum rate of the proposed scheme tends to that of the ideal case that the channel covariance matrices are assumed to be perfectly known at the base station.