Sensing Without Seeing: Wi-Fi CSI for Human Presence Detection
Problem statement
To enhance the quality of life of older adults living independently at home, various sensing technologies have been explored, including wearable devices (e.g., fall detectors) and camera-based systems. However, many of these solutions are often considered unsuitable for older adults due to concerns related to privacy and usability. To address these issues, researchers are increasingly investigating unobtrusive sensing approaches, such as using Wi-Fi Channel State Information (CSI) signals to classify daily activities of older adults. To continue this exploration, this assignment aims to classify the presence versus absence of a person in a simulated real-world environment using Wi-Fi CSI data.
Task
To collect Wi-Fi Channel State Information (CSI) data corresponding to a range of Activities of Daily Living (ADLs) performed within the eHealth house. Participants will simulate everyday activities of older adults, while multiple Wi-Fi devices will be deployed to record the CSI signals associated with these activities.
Expected outcomes
- 20% Theory - Survey of available literature, specifically looking at ways of data collection and data analysis techniques using CSI data.
- 60% Simulations - Data collection in ehealth house (20 participants)
- 20 %Writing - Report
Contact
Alex Chiumento (a.chiumento@utwente.nl)
Nikita Sharma (n.sharma@utwente.nl)