A Novel Fully-Connected Beamforming Network Architecture with the QR Factorization Synthesis Method

Li Sun, Jun Yi Gu, Yang Ming Wan, Shi Gang Zhou

Research output: Contribution to journalArticlepeer-review

Abstract

Fully-connected beamforming networks (FCBNs) can provide better performance in a hybrid beamforming (HBF) architecture for massive multiple-input-multiple-output (MIMO) systems. The traditional FCBN adopts the 'splitters + phase shifters + attenuators + combiners' architecture, which has a huge intrinsic loss by the theoretical analysis. The network loss and topology complexity pose challenges for its engineering applications. To address the aforementioned issues, this article proposes a new efficient FCBN architecture, eliminating the intrinsic loss of the network. The proposed new FCBN architecture consists of tunable and fixed directional couplers and phase shifters in cascading, achieving both amplitude and phase adjustments. For synthesizing such a tunable network with arbitrary I/O ports, the complete QR factorization synthesis method with the corresponding tunable complex Givens rotation matrix has been proposed and proved. To demonstrate the proposed synthesis method, a 2×4 FCBN network has been synthesized, fabricated, and measured. Concentrating on the characteristics of the FCBN, an optimization process has been introduced to simplify hardware topology complexity. The results show that the proposed FCBN architecture has advantages in no intrinsic loss, simple topology, and low cost, making it widely applied in wireless communication systems to improve the system efficiency.

Original languageEnglish
Pages (from-to)8772-8785
Number of pages14
JournalIEEE Transactions on Antennas and Propagation
Volume72
Issue number11
DOIs
StatePublished - 2024

Keywords

  • Beamforming network
  • fully-connected
  • optimization
  • QR factorization

Fingerprint

Dive into the research topics of 'A Novel Fully-Connected Beamforming Network Architecture with the QR Factorization Synthesis Method'. Together they form a unique fingerprint.

Cite this