The massive MU-MIMO system with optimal number of users through linear precoders

In this paper, the downlink (DL) of the multiuser multiple-input multiple-output (MU-MIMO) system is investigated, where channel estimation and large-scale fading are taken into consideration and the number of antennas M at the base station (BS) is extremely large. We calculate the asymptotic sum rate of singular value decomposition (SVD)-based precoder and regularized zero-forcing (RZF) precoder. The optimal number of served users is taken into account. The result shows that M should be more than the product of the number of users and the number of antennas at each user. Based on the optimal number, we propose a convenient data stream allocation strategy aiming to improve the system sum rate, meanwhile maintaining the number of simultaneously served users. Numerical simulations reveal that our approximations are appropriate for the large-scale MU-MIMO system and the proposed data stream allocation can provide significant improvement in sum-rate performance.