Millimeter Wave Massive MIMO Downlink Per-Group Communications With Hybrid Linear Precoding

We address the problem of analyzing and classifying into groups downlink channel environment in a dense millimeter-wavelength cell, accounting for path loss, multipath fading, User Equipment (UE) blocking, by employing hybrid propagation fading model, thus using accurate inter-group interference modeling. The base station (BS) employs large Uniform Planar Array (UPA) to facilitate massive Multiple-Input, Multiple-Output (MIMO) communications with high efficiency. UEs are equipped single antenna distributed uniformly within cell. key is defining toward efficient precoding. Because balanced throughput desired between groups, up four separate frequency sub-carriers employed an Orthogonal Frequency Division Multiple Access (OFDMA) system. show that three or OFDMA subcarriers, depending on number cell range, can be efficiently separated high-throughput each group Virtual Channel Model Beams (VCMB) based inner precoding, followed Multi-User Multiple-Input (MU-MIMO) outer precoders. For group, we study two different sub-grouping methods offering advantages. improvement offered Zero-Forcing Per-Group Precoding (ZF-PGP) over Matched Filter (MFP) (ZFP) very when finite-alphabet inputs used. also ZF-PGP significantly outperforms MFP Gaussian certain realistic setups. Finally, new technique power allocation among proposed, called Optimized Power Allocation (OPGPA), which achieves much higher efficiency jointly equal all group.