On the Optimal Sum-Capacity Growth of ZFBF with Quality Based Channel Reporting in Multiuser Downlink Channels

Abstract

In multiuser multiple-input-multiple-output (MIMO) downlink channels, zero-forcing beamforming (ZFBF), though it is simpler than dirty-paper coding (DPC), has been shown to achieve an optimal sum-capacity growth as the number of users goes to infinity. However, the result is based on the assumption that perfect channel information at the transmitter from all the users is known, which is prohibitive especially for a large number of users due to feedback congestion. In this case, wisely preventing the users with poor channel quality from sending the feedback might relieve the burdensome feedback load, which is often called quality-based channel state reporting (QBR). In this letter, we prove ZFBF with QBR also can achieve the asymptotic optimal growth. With numerical investigations, the optimality is confirmed. In addition, we numerically show how much feedback burden can be reduced by QBR at a negligible cost of the sum-capacity.