Breathing-based Continuous Non-intrusive User Verification Leveraging Commodity WiFi

Abstract

This paper opens up a new pathway of the utility of breathing pattern for user verification. We demonstrate that it is possible to capture people’s breathing pattern leveraging commodity WiFi devices. While prior solutions for biometrics-based user recognition usually require dedicated devices (e.g., video cameras or IR sensors), this paper introduces the first general, low-cost breathing-based user verification system using commodity WiFi devices. The proposed system is based on the fact that the breathing pattern always keeps consistent for the same user but distinct among different people. Our innovative method successfully extracts the breathing pattern of different people based on channel state information of WiFi signal to facilitate user verification. The prototype study using two commodity WiFi devices can differentiate people with an average verification accuracy over 90%, suggesting that our breathing-based user verification system using commerical off-the-shelf (COTS) WiFi is promising to be one of the most critical methods in biometrics.