Statistical Precoder Design in Multi-User Systems via Graph Neural Networks and Generative Modeling

Nurettin Turan; Srikar Allaparapu; Donia Ben Amor; Benedikt Bock; Michael Joham; Wolfgang Utschick

doi:10.1109/LWC.2025.3546633

Statistical Precoder Design in Multi-User Systems via Graph Neural Networks and Generative Modeling

Nurettin Turan, Srikar Allaparapu, Donia Ben Amor, Benedikt Bock, Michael Joham, Wolfgang Utschick

Chair of Signal Processing

Technical University of Munich

Research output: Contribution to journal › Article › peer-review

Abstract

This letter proposes a GNN-based framework for statistical precoder design that leverages model-based insights to compactly represent statistical knowledge, resulting in efficient, lightweight architectures. The framework also supports approximate statistical information in FDD systems obtained through a GMM-based limited feedback scheme in massive MIMO systems with low pilot overhead. Simulations demonstrate the superiority of the proposed framework over baseline methods, including stochastic iterative algorithms and DFT codebook-based approaches, particularly in systems with low pilot overhead.

Original language	English
Journal	IEEE Wireless Communications Letters
DOIs	https://doi.org/10.1109/LWC.2025.3546633
State	Accepted/In press - 2025

Keywords

Gaussian mixture model
graph neural network
limited feedback
measurement data
Statistical precoding

Access to Document

10.1109/LWC.2025.3546633

Cite this

@article{cc4c4f1441b04284b7920ae87f4c3f36,

title = "Statistical Precoder Design in Multi-User Systems via Graph Neural Networks and Generative Modeling",

abstract = "This letter proposes a GNN-based framework for statistical precoder design that leverages model-based insights to compactly represent statistical knowledge, resulting in efficient, lightweight architectures. The framework also supports approximate statistical information in FDD systems obtained through a GMM-based limited feedback scheme in massive MIMO systems with low pilot overhead. Simulations demonstrate the superiority of the proposed framework over baseline methods, including stochastic iterative algorithms and DFT codebook-based approaches, particularly in systems with low pilot overhead.",

keywords = "Gaussian mixture model, graph neural network, limited feedback, measurement data, Statistical precoding",

author = "Nurettin Turan and Srikar Allaparapu and Amor, {Donia Ben} and Benedikt Bock and Michael Joham and Wolfgang Utschick",

year = "2025",

doi = "10.1109/LWC.2025.3546633",

language = "English",

journal = "IEEE Wireless Communications Letters",

issn = "2162-2337",

publisher = "IEEE Communications Society",

}

TY - JOUR

T1 - Statistical Precoder Design in Multi-User Systems via Graph Neural Networks and Generative Modeling

AU - Turan, Nurettin

AU - Allaparapu, Srikar

AU - Amor, Donia Ben

AU - Bock, Benedikt

AU - Joham, Michael

AU - Utschick, Wolfgang

PY - 2025

Y1 - 2025

N2 - This letter proposes a GNN-based framework for statistical precoder design that leverages model-based insights to compactly represent statistical knowledge, resulting in efficient, lightweight architectures. The framework also supports approximate statistical information in FDD systems obtained through a GMM-based limited feedback scheme in massive MIMO systems with low pilot overhead. Simulations demonstrate the superiority of the proposed framework over baseline methods, including stochastic iterative algorithms and DFT codebook-based approaches, particularly in systems with low pilot overhead.

AB - This letter proposes a GNN-based framework for statistical precoder design that leverages model-based insights to compactly represent statistical knowledge, resulting in efficient, lightweight architectures. The framework also supports approximate statistical information in FDD systems obtained through a GMM-based limited feedback scheme in massive MIMO systems with low pilot overhead. Simulations demonstrate the superiority of the proposed framework over baseline methods, including stochastic iterative algorithms and DFT codebook-based approaches, particularly in systems with low pilot overhead.

KW - Gaussian mixture model

KW - graph neural network

KW - limited feedback

KW - measurement data

KW - Statistical precoding

UR - http://www.scopus.com/inward/record.url?scp=85219574522&partnerID=8YFLogxK

U2 - 10.1109/LWC.2025.3546633

DO - 10.1109/LWC.2025.3546633

M3 - Article

AN - SCOPUS:85219574522

SN - 2162-2337

JO - IEEE Wireless Communications Letters

JF - IEEE Wireless Communications Letters

ER -

Statistical Precoder Design in Multi-User Systems via Graph Neural Networks and Generative Modeling

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this