版权所有:内蒙古大学图书馆 技术提供:维普资讯• 智图
内蒙古自治区呼和浩特市赛罕区大学西街235号 邮编: 010021
作者机构:Tech Univ Munich TUM Sch Computat Informat & Technol D-80333 Munich Germany
出 版 物:《IEEE WIRELESS COMMUNICATIONS LETTERS》 (IEEE Wireless Commun. Lett.)
年 卷 期:2025年第14卷第5期
页 面:1491-1495页
核心收录:
学科分类:0810[工学-信息与通信工程] 0808[工学-电气工程] 08[工学] 0812[工学-计算机科学与技术(可授工学、理学学位)]
基 金:Federal Ministry of Education and Research of Germany in the Program of "Souveraen. Digital. Vernetzt" 16KISK002
主 题:Feature extraction Covariance matrices Training Vectors Stochastic processes Signal to noise ratio Indexes Graph neural networks Discrete Fourier transforms Precoding Statistical precoding graph neural network Gaussian mixture model limited feedback measurement data
摘 要:This letter proposes a graph neural network (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 frequency division duplex (FDD) systems obtained through a Gaussian mixture model (GMM)-based limited feedback scheme in massive multiple-input multiple-output (MIMO) systems with low pilot overhead. Simulations demonstrate the superiority of the proposed framework over baseline methods, including stochastic iterative algorithms and discrete Fourier transform (DFT) codebook-based approaches, particularly in systems with low pilot overhead.