合成无线信道-传感器数据配对数据集
收藏资源简介:
该数据集由首尔国立大学和韩国原子能研究所的研究团队构建,旨在通过3D建模和射线追踪技术生成合成数据,以模拟真实的无线通信场景。数据集包含两个部分,分别将下行信道与RGB图像和上行信道配对,用于训练和评估多模态信道重建框架。数据集通过高质量3D对象建模和射线追踪仿真生成,符合5G NR标准,适用于5G通信系统中的信道状态信息(CSI)重建研究,旨在提高CSI重建的准确性,从而优化波束成形增益。
This dataset was constructed by a research team from Seoul National University and the Korea Atomic Energy Research Institute. It aims to generate synthetic data via 3D modeling and ray tracing techniques to simulate realistic wireless communication scenarios. The dataset comprises two parts, which pair downlink channels with RGB images and uplink channels respectively for training and evaluating multimodal channel reconstruction frameworks. Generated via high-fidelity 3D object modeling and ray tracing simulations, the dataset complies with 5G NR standards and is applicable to channel state information (CSI) reconstruction research in 5G communication systems, with the goal of improving the accuracy of CSI reconstruction and thereby optimizing beamforming gain.

- 1Multi-Modal Variable-Rate CSI Reconstruction for FDD Massive MIMO Systems首尔国立大学, 韩国原子能研究所 · 2025年



