To obtain access to full text of journal and articles you must register!
- Article name
- RESEARCH OF TEMPORAL-DRIFT ROBUSTNESS OF WI-FI CSI FINGERPRINTING FOR INDOOR HUMAN LOCALIZATION
- Authors
- Dementiev V. D., , vadim23dem@yandex.ru, Kaluga Branch of Bauman Moscow State Technical University, Kaluga, Russia
Borsuk N. A., , borsuk.65@yandex.ru, Kaluga Branch of Bauman Moscow State Technical University, Kaluga, Russia
- Keywords
- Channel State Information / Wi-Fi sensing / CSI fingerprinting / indoor localization / ESP32 / machine learning / temporal drift / random forest
- Year
- 2026 Issue 3 Pages 46 - 50
- Code EDN
- TSUZBA
- Code DOI
- 10.52190/2073-2597_2026_3_46
- Abstract
- An experimental indoor human localization system based on a single pair of ESP32-WROOM-32U transceiver modules is presented. The classifier's robustness against temporal drift in the CSI signal between measurement sessions is investigated. It is shown that standard accuracy estimation using a random split of data into training and test sets yields an inflated result compared to a rigorous evaluation based on independent measurement sessions. To improve accuracy, a feature based on subtracting the reference spectrum of an empty room-measured during the same session-is proposed. Practical aspects of collecting training data and implementing a real-time localization system are discussed.
- Text
- BUY for read the full text of article
- Buy
- 800.00 rub
