Scholarlink

<p>Accurate reconstruction of wireless communication channels is essential for simulating and optimizing real-world network environments. This paper introduces a novel machine learning (ML) framework for generating Clustered Delay Line (CDL) channel profiles from limited Channel State Information (CSI). We adopted the supervised learning paradigm and trained ML models on synthetic data generated in accordance with the CDL profile in 3GPP TS 38.901. The proposed models effectively reconstruct the wireless channel by estimating key Large-Scale Parameters (LSP) and Small-Scale Parameters (SSP) solely from CSI feedback from the User Equipment (UE). We present a detailed evaluation of our architecture and models, demonstrating the feasibility and potential applications of a limited CSI-based channel reconstruction.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Machine learning framework for CDL channel profile estimation and channel reconstruction from limited CSI feedback

  • 期刊论文

  • 发表日期：

    2026年4月20日

查看全文
  • Ben Earle,
  • Ala’a Al-Habashna,
  • Gabriel Wainer,
  • Xingliang Li,
  • Guoqiang Xue

摘要

Accurate reconstruction of wireless communication channels is essential for simulating and optimizing real-world network environments. This paper introduces a novel machine learning (ML) framework for generating Clustered Delay Line (CDL) channel profiles from limited Channel State Information (CSI). We adopted the supervised learning paradigm and trained ML models on synthetic data generated in accordance with the CDL profile in 3GPP TS 38.901. The proposed models effectively reconstruct the wireless channel by estimating key Large-Scale Parameters (LSP) and Small-Scale Parameters (SSP) solely from CSI feedback from the User Equipment (UE). We present a detailed evaluation of our architecture and models, demonstrating the feasibility and potential applications of a limited CSI-based channel reconstruction.