Hybrid Beamforming for Multi-User Millimeter-Wave Networks

Abstract

This paper considers hybrid beamforming by combining an analog beamformer with a new regularized zero forcing baseband one, for multiuser millimeter-wave networks under a limited number of radio frequency (RF) chains. Three popular scenarios are examined: i) the number of users is up to the number of RF chains in a single-cell network, ii) the number of users is up to twice the number of RF chains in a singlecell network, and iii) the number of users is up to twice the number of RF chains in each cell of a two-cell network. In the second and third scenarios, we group the users into two categories of cell-center users as well as cell-edge users and serve them in two different time fractions. In the third scenario, we propose to suppress the inter-cell interference by serving the cell-center and cell-edge users in alternate fractional-time slots. In all the three scenarios, we determine the optimal power allocation maximizing the users' minimum rate. Finally, lowcomplexity path-following algorithms having rapid convergence are developed for the computation of the optimal power. Our simulation results show that the proposed algorithms achieve a clear performance gain over the existing benchmarkers.

Finite-Resolution Digital Beamforming for Multi-User Millimeter-Wave Networks

“…Therefore, using (15), we can find the equivalent of the optimization problem ( 14) at the κ-th iteration, which is given by:…”

“…Moreover, there is no systematic design for analog beamformers due to the unit-modulus constraint imposed on the entries of the analog precoder matrices. Additionally, analog or hybrid beamforming architectures prevent the employment of computationally efficient digital beamformers such as zero-forcing beamforming (ZFB), which can only be used if the number of users is not more than the number of RF chains [15].…”

Maximizing the Geometric Mean of User-Rates to Improve Rate-Fairness: Proper vs. Improper Gaussian Signaling

“…Furthermore, the SR performance is typically improved with more users involved because there are more flexible choices for the users' channels [3]. The spectral efficiency is thus addressed more appropriately via either SR maximization under specific quality-of-service (QoS) constraints in terms of the users' minimum rate, or by max-min user-rate optimization, but their computation is quite demanding [1], [2], [4]- [6].…”

Deep Learning‐Based Symbol‐Level Precoding for Large‐Scale Antenna System

“…The sum-rate maximization is addressed in [7], where the analog beamformer is optimized via alternating optimization approach. The maximization of the minimum total transmit rate with optimal power allocation is further developed via low-complexity path-following algorithms [8]. However, the limitation on the number of RF chains makes the designs of hybrid beamforming considerable challenging due to the unit-modulus constraints imposed on the analog beamformers.…”