The doi above is for this specific version of this dataset, which is currently the latest. Newer versions may be published in the future. For a link that will always point to the latest version, please use
doi: 10.4121/e1d6a078-9022-4f48-9aa6-e22389980fee

Guendel, Ronny G.; Yarovoy, Alexander; Francesco Fioranelli (2024): Dataset: Continuous Human Activities Utilizing Three Pulsed Radars Exploiting Multipath. Version 2. 4TU.ResearchData. dataset. https://doi.org/10.4121/e1d6a078-9022-4f48-9aa6-e22389980fee.v2

Dataset

x choose version: version 1 - 2024-02-01

• Artificial Intelligence and Image Processing
• Electrical and Electronic Engineering
• Data Format
• Other Engineering
• Health and Support Services
• Applied Mathematics
• Other Health
• Mathematical Physics
• Engineering
• Health
• Mathematical Sciences
• Technology
• Information and Computing Sciences

distributed radar, hierarchical clustering, human activity recognition, Machine learning, radar multipath, Radar signal processing

CC BY 4.0

RefWorks, BibTeX, Reference Manager, Endnote, DataCite, NLM, DC, CFF

by Ronny G. Guendel , Alexander Yarovoy, Francesco Fioranelli

The dataset consists of recordings showcasing multipath reflection for two out of three radars within a radar network. It utilizes the PulsON P410 - Time Domain radar by Humatics, operating as an Ultra Wideband radar. The dataset consists of 45 folders, organized for 14 participants, covering range-time data, micro-Doppler spectrogram data, and range-Doppler data. It includes 12 classes like 'Junk/BRN,' 'Walking,' 'Stationary,' 'Sitting down,' 'Standing up (Sit),' 'Bending (Sit),' 'Bending (Sta),' 'Falling (Wlk),' 'Standing up (Fal),' 'Falling (Sta),' 'Micro Gesture,' and 'Walking (Obj),' each characterizing specific movements or actions during three repetitions of 60 seconds of data recording. Further details can be found in the accompanying readme file.

Dataset parameters:

• Continuous recorded data using a radar network of 60 sec
• 2 radars out of 3 radars provide multipath reflections
• Ready to use radar data of:
• range-time data
• micro-Doppler spectrogram data
• range-Doppler data
• 14 participants (7 male, 7 female)
• 12 classes of translation and in-place activities
• Background noise recording included
• Zenodo source and GitHub to read the files (code also included here)

Radar parameters:

• Center frequency: 4.3 GHz
• Bandwidth: 2.2GHz
• Operation mode: Monostatic
• Pulse integration Index: 10 (1024 coherent integrations)
• PRI/PRF: 8.2ms / 122Hz
• Antennae: Omnidirectional broadband antennas

• 2024-02-01 first online, published, posted

4TU.ResearchData

Matlab data files <.mat> in compressed files <zip> and data read file as Matlab script <.m>

TU Delft, Faculty Electrical Engineering, Mathematics and Computer Science, Department of Microelectronics