Constant modulus hybrid beamforming for multi-user systems in the presence of blockers

Abstract

Hybrid beamforming (HBF) solutions, involving both analog and digital stages, are considered as key enablers for next generation of small cell networks wireless communications. They provide interference rejection and spatial multiplexing features for the base stations with fewer radio frequency (RF) chains and analog-to-digital converters (ADCs) compared to the optimal full digital beamforming. However jointly designing the digital beamforming matrix and the analog beamforming matrix with constant modulus entries makes the optimal signal-to-interferenceplus-noise ratio (SINR) solution complex to find. In this paper we propose a simple solution and investigate its SINR loss as a function of angle of arrivals of blockers. The solution consists of setting the analog weights phases equal to those of a full digital beamformer and keeping the modulus of the weights constant, while the complementary digital part is designed with respect to the effective channel matrix. This algorithm is applied over different full digital algorithms, in a medium signal-to-noise ratio (SNR) and high blocker levels. We show that despite that the analog stage introduces losses, the digital stage is still able to compensate them almost to the performance of the optimal full digital beamforming, and so with minor difference between algorithms used in analog stage.

Multi‐user interference characterisation in antenna arrays: an algebraic angle approach

“…Small-cell network case: In this part, the consideration is moved to the multi-user model in a small-cell network. This scenario is typically encountered in uncoordinated dense small networks, that are duplexed in time division (TDD) [11,12]. A simple geometry-based stochastic channel is used.…”

On the use of semidefinite relaxation for uplink phase-only hybrid beamforming with blockers

“…In the sequel, we compare our proposed solution with the state-of-the-art solution of Zhang et al [1] and demonstrate the relevance of the method. In our preceding work [7], we analyzed the SINR loss in existence of a blocker, as a function of the angle of arrival of the blocker for suboptimal approaches, designed on simple phases extraction algorithms. In this paper, we propose to use a semidefinite relaxation (SDR) approach to convert the non-convex problem into a semidefinite programming (SDP) problem, making it easily solvable by off-the-shelf optimization toolboxes.…”