BeamSense: MU-MIMO Beamforming Feedback Dataset for Wi-Fi Sensing

Dataset Summary

BeamSense is a large-scale MU-MIMO Wi-Fi sensing dataset built upon Beamforming Feedback Information (BFI) and Channel State Information (CSI) collected from commodity 802.11ac devices. The dataset corresponds to the research presented in:

• BeamSense: Rethinking Wireless Sensing with MU-MIMO Wi-Fi Beamforming Feedback (Computer Networks, 2025)

• BFA-Sense: Learning Beamforming Feedback Angles for Wi-Fi Sensing (PerCom Workshops, 2024)

BeamSense contains synchronized BFI, BF angles, and CSI samples from multiple Wi-Fi stations (STAs) under different environments and orientations. It is designed for:

• Human activity recognition with Beamforming Feedback Information (BFI) and CSI
• Gesture recognition with Beamforming Feedback Information (BFI) and CSI
• Device-free sensing with BFI
• Cross-domain generalization
• Beamforming Feedback analysis
• Beamforming Feedback-based attacks
• Wi-Fi sensing algorithm development

Dataset Contents

• Raw Beamforming Feedback Packet Captures: Contains BF report packets (BFI), and PHY metadata
• Synchronized CSI: Extracted from PCAPs using Nexmon-modified pipelines
• Synchronized BFI and CSI

Multi-Environment, Multi-Day Collection:

• Includes data from: Office, Classroom, and Kitchen environments
• LOS and NLOS cases
• Multiple days of data collection
• 3 Different human subjects perform 20 different activities in multiple environments and orientations
• Spatially diverse data from multiple STAs

How To Use

Download the Dataset

Download the dataset with

hf download foysalhaque/BeamSense \
  --repo-type dataset \
  --local-dir your_directory/BeamSense_HF

You will have a directory structure like

your_directory/BeamSense_HF/
    BeamSense_Dataset_aa
    BeamSense_Dataset_ab
    BeamSense_Dataset_ac
    ...
    README.md
    data.json

These BeamSense_Dataset_aa, BeamSense_Dataset_ab, … files are split parts of one large ZIP file.

Move into the download directory with cd your_directory/BeamSense_HF/ and Combine all pieces into one ZIP with cat BeamSense_Dataset_* > BeamSense_Dataset.zip. This produces a large file which is around 180 GB.

Extract the combined ZIP with unzip BeamSense_Dataset.zip -d BeamSense_Data

MATLAB BFI Extraction

./Feedback_split_STAs.sh
cd BFI_Extraction
matlab -nojvm -nosplash -r "pcap_to_bfa; exit"
matlab -nojvm -nosplash -r "bfa_to_batches; exit"

MATLAB CSI Extraction

cd CSI_Extraction
matlab -nojvm -nosplash -r "Extract_CSI; exit"
matlab -nojvm -nosplash -r "CSI_to_batches; exit"

Citation

If you find our work and the dataset useful, please cite the following works:

@article{haque2025beamsense,
  title={BeamSense: Rethinking wireless sensing with MU-MIMO Wi-Fi beamforming feedback},
  author={Haque, Khandaker Foysal and Zhang, Milin and Meneghello, Francesca and Restuccia, Francesco},
  journal={Computer Networks},
  pages={111020},
  year={2025},
  publisher={Elsevier}
}

@inproceedings{haque2024bfa,
  title={BFA-Sense: Learning Beamforming Feedback Angles for Wi-Fi Sensing},
  author={Haque, Khandaker Foysal and Meneghello, Francesca and Restuccia, Francesco},
  booktitle={2024 IEEE International Conference on Pervasive Computing and Communications Workshops},
  pages={575--580},
  year={2024},
  organization={IEEE}
}

Contact

For dataset questions or access issues:

K. Foysal Haque
Northeastern University
Email: [email protected]