Beamforming design with fully connected analog beamformer using deep learning (Record no. 27752)

MARC details
000 -LEADER
fixed length control field 02155nam a2200169Ia 4500
008 - FIXED-LENGTH DATA ELEMENTS--GENERAL INFORMATION
fixed length control field 230619s9999 xx 000 0 und d
022 ## - INTERNATIONAL STANDARD SERIAL NUMBER
International Standard Serial Number International Journal of Communication Systems
082 ## - DEWEY DECIMAL CLASSIFICATION NUMBER
Classification number 621.38
100 ## - MAIN ENTRY--PERSONAL NAME
Personal name Jeyakumar P., Tharanitaran N.M., Malar E., Muthuchidambaranathan P.
260 ## - PUBLICATION, DISTRIBUTION, ETC.
Name of publisher, distributor, etc. Wiley online library
Date of publication, distribution, etc. 2022
300 ## - PHYSICAL DESCRIPTION
Extent 1-12
520 ## - SUMMARY, ETC.
Summary, etc. Beamforming (BF) architecture, an emerging approach for large‐scale antenna arrays, is a key task for the next decade's communication systems. The proposed approach in millimeter wave transmission is a fully connected analog phase‐shifter‐based BF architecture with limited radio frequency chains and imperfect channel state information (CSI). Deep learning (DL) is a powerful method for channel estimation and signal identification in wireless communications. Hence, this research proposes a DL‐enabled beamforming neural network (BFNN) which can be programmed to optimize the beamformer to attain better spectral efficiency. Simulation findings reveal that the proposed BFNN achieves significant performance gain and high robustness to imperfect CSI. The proposed BFNN greatly decreases the computational complexity by 0.16 million floating point operations (FLOPs) over 0.26 million FLOPs by conventional BF algorithms. This paper presents a deep learning‐enabled beamforming architecture with fully connected analog beamformer. There are two design phases, namely, offline training phase and online training phase. Simulation samples are created via system model in the offline training phase, and then updated weights obtained from offline training are fed into beamforming neural network (BFNN) with new loss function. The performance analysis of the proposed design is provided with respect to computational complexity and spectral efficiency (bits/s/Hz) and is also contrasted with current literature work.
650 ## - SUBJECT ADDED ENTRY--TOPICAL TERM
Topical term or geographic name entry element beamforming, beamforming neural network, deep learning, FLOPs, millimete rwave communication
856 ## - ELECTRONIC LOCATION AND ACCESS
Uniform Resource Identifier <a href="https://doi.org/10.1002/dac.5109">https://doi.org/10.1002/dac.5109</a>
Holdings
Title Koha item type
Beamforming design with fully connected analog beamformer using deep learning  
  iTech Publications