arXiv · 2611.15532
A Proof-of-Concept Study of Wi-Fi Channel State Information-Based Fall Detection Using ESP32 Devices
Abstract
This paper reports a small-scale proof-of-concept study of contactless fall detection using Wi-Fi channel state information (CSI) collected with two ESP32 devices. One device acts as the transmitter and the other as the receiver, with a separation of 3.5~m. The evaluated signal-processing pipeline uses CSI amplitude, second-level discrete wavelet transform (DWT) denoising with a Symlet-10 wavelet, z-score normalization, and fixed-length segmentation. Wavelet-scattering coefficients are used as features, and MATLAB's automated classifier-selection workflow identifies a LogitBoost ensemble with a minimum leaf size of four and a learning rate of 0.022052. The proposed classifier achieved an accuracy of 93.10%, with a precision of 100%, recall of 87.50%, and F1-score of 93.33%. These results demonstrate the potential of Wi-Fi CSI-based sensing for contactless fall detection using low-cost ESP32 hardware.
Explore related subjects
Keep this discovery
Explore connections, maps & timelines
Amritha S. S., Anhajul S. Rahman, Mohammed Bin Saleem, Vaishnav Sankar, Jerrin Thomas Panachakel. 2026-11-11. A Proof-of-Concept Study of Wi-Fi Channel State Information-Based Fall Detection Using ESP32 Devices. https://searcharxiv.com/abs/2611.15532
Citations retain the original authors and DOI and include this site’s record URL. If this discovery service contributed to your research, you can also cite Searcharxiv by Indiciae: BibTeX · RIS.