Coordinated optimization of EVD-based channel estimators in multi-cell massive MIMO networks

This paper studies the eigenvalue decomposition (EVD) based uplink channel estimation in multi-cell massive multi-input multi-output (MIMO) networks. The performance of the conventional EVD-based channel estimators is limited by the interference from other users in the network due to finite number of antennas and finite transmit samples. We propose an iterative optimization procedure with multiple steps in order to adaptively improve the channel estimations by means of jointly processing a subset of the receive sample data from multiple cells at a centralized node via coordination. Furthermore, we analyze the amount of necessary backend signaling among the coordinating cells and describe an efficient signaling scheme for the proposed multi-cell coordination. We compare the performance of this coordinated optimization scheme with the conventional EVD-based estimators via numerical analysis and show that multi-cell coordination helps to achieve lower symbol error rates.