Antenna selection in measured massive MIMO channels using convex optimization

Massive MIMO, also known as very-large MIMO or large-scale antenna systems, is a new technique that potentially can offer large network capacities in multi-user scenarios, where the base stations are equipped with a large number of antennas simultaneously serving multiple single-antenna users on the same frequency. However, the radio-frequency (RF) chains associated with the antennas increase the system complexity and hardware cost. Antenna selection is a signal processing technique that can help reduce the number of RF chains, while preserving the system performance at a certain required level. We study the transmit antenna selection in measured massive MIMO channels from several measurement campaigns in the 2.6 GHz frequency range. Convex optimization is used to select the antenna subset that maximizes the dirty-paper coding (DPC) capacity in the downlink. With a certain number of RF chains, we increase the number of base station antennas from the same as the RF chains to a large number, from which we perform the antenna selection. The investigation shows that with more available antennas than RF chains, the antenna selection can significantly improve the system performance, especially for a compact cylindrical array, which without this antenna selection shows lower performance than a physically large linear array with the same number of elements, in the studied scenarios.