WiWho: WiFi-Based Person Identification in Smart Spaces

There has been a growing interest in equipping the objects and environment surrounding the user with sensing capabilities. Smart indoor spaces such as smart homes and offices can implement the sensing and processing functionality, relieving users from the need of wearing or carrying smart devices. Enabling such smart spaces requires device-free effortless sensing of user's identity and activities. Device-free sensing using WiFi has shown great potential in such scenarios, however, fundamental questions such as person identification have remained unsolved. In this paper, we present WiWho, a framework that can identify a person from a small group of people in a device-free manner using WiFi. We show that Channel State Information (CSI) used in recent WiFi can identify a person's steps and walking gait. The walking gait being distinguishing characteristics for different people, WiWho uses CSI-based gait for person identification. We demonstrate how step and walk analysis can be used to identify a person's walking gait from CSI, and how this information can be used to identify a person. WiWho does not require a person to carry any device and is effortless since it only requires the person to walk for a few steps (e.g. entering a home or an office). We evaluate WiWho using experiments at multiple locations with a total of 20 volunteers, and show that it can identify a person with average accuracy of 92% to 80% from a group of 2 to 6 people. We also show that in most cases walking as few as 2-3 meters is sufficient to recognize a person's gait and identify the person. We discuss the potential and challenges of WiFi- based person identification with respect to smart space applications.

[1]  Fadel Adib,et al.  See through walls with WiFi! , 2013, SIGCOMM.

[2]  Shaojie Tang,et al.  Wi-Sleep: Contactless Sleep Monitoring via WiFi Signals , 2014, 2014 IEEE Real-Time Systems Symposium.

[3]  Kaishun Wu,et al.  WiFall: Device-free fall detection by wireless networks , 2017, IEEE INFOCOM 2014 - IEEE Conference on Computer Communications.

[4]  David Wetherall,et al.  Tool release: gathering 802.11n traces with channel state information , 2011, CCRV.

[5]  Hae Young Noh,et al.  Indoor Person Identification through Footstep Induced Structural Vibration , 2015, HotMobile.

[6]  Woo Chaw Seng,et al.  A review of biometric technology along with trends and prospects , 2014, Pattern Recognit..

[7]  Kaishun Wu,et al.  We Can Hear You with Wi-Fi! , 2014, IEEE Transactions on Mobile Computing.

[8]  Zhaohui Wu,et al.  Accelerometer-Based Gait Recognition by Sparse Representation of Signature Points With Clusters , 2015, IEEE Transactions on Cybernetics.

[9]  Shwetak N. Patel,et al.  Whole-home gesture recognition using wireless signals , 2013, MobiCom.

[10]  Andreas Savvides,et al.  PEM-ID: Identifying people by gait-matching using cameras and wearable accelerometers , 2009, 2009 Third ACM/IEEE International Conference on Distributed Smart Cameras (ICDSC).

[11]  Richard P. Martin,et al.  Tracking human queues using single-point signal monitoring , 2014, MobiSys.

[12]  Parameswaran Ramanathan,et al.  Leveraging directional antenna capabilities for fine-grained gesture recognition , 2014, UbiComp.

[13]  Larry S. Davis,et al.  Stride and cadence as a biometric in automatic person identification and verification , 2002, Proceedings of Fifth IEEE International Conference on Automatic Face Gesture Recognition.

[14]  Gora Chand Nandi,et al.  Gait Based Personal Identification System Using Rotation Sensor , 2012 .

[15]  Ian Witten,et al.  Data Mining , 2000 .

[16]  อนิรุธ สืบสิงห์,et al.  Data Mining Practical Machine Learning Tools and Techniques , 2014 .

[17]  Rob Miller,et al.  3D Tracking via Body Radio Reflections , 2014, NSDI.

[18]  Carla Schlatter Ellis,et al.  Using Ground Reaction Forces from Gait Analysis: Body Mass as a Weak Biometric , 2007, Pervasive.

[19]  Imed Bouchrika,et al.  On Using Gait in Forensic Biometrics , 2011, Journal of forensic sciences.

[20]  Shaojie Tang,et al.  Electronic frog eye: Counting crowd using WiFi , 2014, IEEE INFOCOM 2014 - IEEE Conference on Computer Communications.

[21]  Kirsi Helkala,et al.  Biometric Gait Authentication Using Accelerometer Sensor , 2006, J. Comput..

[22]  Emmanuel,et al.  Using machine learning for real-time activity recognition and estimation of energy expenditure , 2008 .

[23]  Parth H. Pathak,et al.  Your AP knows how you move: fine-grained device motion recognition through WiFi , 2014, HotWireless@MobiCom.

[24]  Yasushi Makihara,et al.  The largest inertial sensor-based gait database and performance evaluation of gait-based personal authentication , 2014, Pattern Recognit..

[25]  Parth H. Pathak,et al.  Analyzing Shopper's Behavior through WiFi Signals , 2015, WPA@MobiSys.

[26]  Michael L. Littman,et al.  Activity Recognition from Accelerometer Data , 2005, AAAI.

[27]  Joseph A. Paradiso,et al.  Gait Analysis Using a Shoe-Integrated Wireless Sensor System , 2008, IEEE Transactions on Information Technology in Biomedicine.

[28]  Lawrence Wai-Choong Wong,et al.  Indoor localization with channel impulse response based fingerprint and nonparametric regression , 2010, IEEE Transactions on Wireless Communications.

[29]  Jie Yang,et al.  E-eyes: device-free location-oriented activity identification using fine-grained WiFi signatures , 2014, MobiCom.

[30]  Jie Yang,et al.  Smartphone based user verification leveraging gait recognition for mobile healthcare systems , 2013, 2013 IEEE International Conference on Sensing, Communications and Networking (SECON).