Sparse Approximation of an Outdoor-to-Indoor Massive MIMO Channel Measurement

Models for the wireless channel for massive multiple-input multiple-output (MIMO) communication in the sub-6 GHz frequencies often assume a rich scattering environment. However, measurements in outdoor-to-indoor scenarios indicate a low rank of the wireless channel coefficients matrix. We approximate a measured 70×169 massive MIMO outdoor-to-indoor channel with its sparse representation by applying atomic norm minimization. We further investigate the enhanced resolution in direction of departure and direction of arrival for closely located multipath components compared to conventional beamforming. Finally, we compare the achievable rate for the measured channel with sparse approximated versions.