Multi-Person Localization via RF Body Reflections

Explore a groundbreaking RF-based indoor localization system capable of tracking multiple users without wearable devices in this 26-minute USENIX conference talk. Discover how WiTrack2.0 leverages wireless signal reflections off human bodies to pinpoint locations, even for static users, by detecting subtle breathing movements. Learn about the system's ability to operate in multipath-rich environments, localize up to five people simultaneously with high accuracy, and provide coarse tracking of body parts. Delve into the technical aspects of FMCW, multi-shift FMCW, and methods for combining multiple vantage points to achieve precise localization. Gain insights into the implementation, experimental results, and potential applications of this innovative technology in various fields.

Intro
Applications
Limitations of Past Work
How do these systems work?
FMCW: Measure time by measuring frequency
Mapping Distance to Location
Dealing with multi-path is easy when there is one moving user
Combining across multiple vantage points Experiment: Two users walking
Achieving 16 vantage points • Naive solution: 1 Transmitter and 16 Receivers
Ideally: Transmit in the same time and in the same frequency band without interfering
Objective: Transmit and Get Reflection
Multi-shift FMCW enables multiple transmissions at the same time and in the same frequency band without interference
Exploit breathing motion for localize static users
User Sitting Still (Breathing)
Implementation
Accuracy of Localizing Static Users 100 experiments where users sit down in different locations in the room
Reflection-based Human Localization