Multimodal-Wireless
收藏资源简介:
Multimodal-Wireless是一个开源的多模态感知数据集,专为无线通信研究而设计。该数据集基于CARLA模拟器和Sionna框架的集成和可定制数据管道生成,包含约160,000帧数据,涵盖四个虚拟城镇、十六种通信场景和三种天气条件,包括多种传感模态:通信信道、激光雷达、RGB和深度相机、惯性测量单元和雷达。数据集的创建过程包括在CARLA中执行场景和数据采集、在Blender中重建场景、以及在Sionna中生成信道。该数据集适用于通信和协同感知的研究,例如使用多模态大型语言模型进行波束预测。
Multimodal-Wireless is an open-source multimodal perception dataset specifically designed for wireless communications research. Generated via an integrated and customizable data pipeline built upon the CARLA simulator and Sionna framework, this dataset contains approximately 160,000 frames spanning four virtual towns, sixteen communication scenarios, and three weather conditions, and encompasses multiple sensing modalities including communication channels, LiDAR, RGB and depth cameras, inertial measurement units (IMUs), and radar. The dataset creation process includes scene execution and data collection in CARLA, scene reconstruction in Blender, as well as channel generation in Sionna. This dataset is applicable to research on communications and collaborative perception, such as beam prediction using multimodal large language models (LLMs).
Multimodal-Wireless 数据集概述
数据集简介
Multimodal-Wireless 是一个开源的、面向通信与协同感知研究的多模态感知数据集,通过基于 CARLA 模拟器和 Sionna 框架构建的集成可定制数据流水线生成。
数据集规模
- 总帧数:161,400 帧
- 覆盖范围:4 个虚拟城镇、16 个通信场景、3 种天气条件
数据模态
- 通信信道
- LiDAR
- RGB 和深度相机
- IMU
- 雷达
- 边界框
技术规格
- 采样频率:100 Hz
- 信道采样:100 Hz 采样,适用于多样化应用
- 传感器数据:100 Hz 过采样,时间对齐,可轻松降采样供实际使用
特性优势
- 全天气适应能力
- 完全灵活性和可扩展性,仅需修改配置文件
- 车联网通信与感知的统一资源
引用方式
bibtex @article{mao2025mmw, title = {Multimodal-Wireless: A Large-Scale Dataset for Sensing and Communication}, author={Tianhao Mao and Le Liang and Jie Yang and Hao Ye and Shi Jin and Geoffrey Ye Li}, journal={arXiv preprint arXiv:2511.03220}, year={2025} }
资源链接
- 论文:https://le-liang.github.io/mmw/
- 代码/资源:https://le-liang.github.io/mmw/
- 数据集下载:
- 全球(OneDrive):https://le-liang.github.io/mmw/
- 中国大陆镜像:https://le-liang.github.io/mmw/
- 复现/扩展:
- 教程视频:https://le-liang.github.io/mmw/
- 中国大陆镜像:https://le-liang.github.io/mmw/

- 1Multimodal-Wireless: A Large-Scale Dataset for Sensing and Communication东南大学信息科学与工程学院, 南京 210096, 中国 · 2025年



