SCF: Sparse channel-state-information feedback using Karhunen-Loève transform

In the paper, we propose a channel-state-information (CSI) feedback method, especially, for large-scale MIMO (LMIMO) systems. From the enormous CSI, to extract essential information without redundant information, which arises from the highly correlated antennas, and to make the CSI to be sparse with its essential information, we propose to use a primary component analysis using a Karhunen-Loève transform (KLT) matrix. To obtain the KLT matrix, we express a channel covariance matrix with its statistics, namely transmit antennas’ and receive antennas’ correlation matrices, channel variance, and channel delay profile. Numerical results verify that the proposed sparse CSI feedback (SCF) method is a promising method to feed back the highly correlated huge CSIs for L-MIMO systems.