FreeSense:Indoor Human Identification with WiFi Signals

📝 Abstract

Human identification plays an important role in human-computer interaction. There have been numerous methods proposed for human identification (e.g., face recognition, gait recognition, fingerprint identification, etc.). While these methods could be very useful under different conditions, they also suffer from certain shortcomings (e.g., user privacy, sensing coverage range). In this paper, we propose a novel approach for human identification, which leverages WIFI signals to enable non-intrusive human identification in domestic environments. It is based on the observation that each person has specific influence patterns to the surrounding WIFI signal while moving indoors, regarding their body shape characteristics and motion patterns. The influence can be captured by the Channel State Information (CSI) time series of WIFI. Specifically, a combination of Principal Component Analysis (PCA), Discrete Wavelet Transform (DWT) and Dynamic Time Warping (DTW) techniques is used for CSI waveform-based human identification. We implemented the system in a 6m*5m smart home environment and recruited 9 users for data collection and evaluation. Experimental results indicate that the identification accuracy is about 88.9% to 94.5% when the candidate user set changes from 6 to 2, showing that the proposed human identification method is effective in domestic environments.

💡 Analysis

Human identification plays an important role in human-computer interaction. There have been numerous methods proposed for human identification (e.g., face recognition, gait recognition, fingerprint identification, etc.). While these methods could be very useful under different conditions, they also suffer from certain shortcomings (e.g., user privacy, sensing coverage range). In this paper, we propose a novel approach for human identification, which leverages WIFI signals to enable non-intrusive human identification in domestic environments. It is based on the observation that each person has specific influence patterns to the surrounding WIFI signal while moving indoors, regarding their body shape characteristics and motion patterns. The influence can be captured by the Channel State Information (CSI) time series of WIFI. Specifically, a combination of Principal Component Analysis (PCA), Discrete Wavelet Transform (DWT) and Dynamic Time Warping (DTW) techniques is used for CSI waveform-based human identification. We implemented the system in a 6m*5m smart home environment and recruited 9 users for data collection and evaluation. Experimental results indicate that the identification accuracy is about 88.9% to 94.5% when the candidate user set changes from 6 to 2, showing that the proposed human identification method is effective in domestic environments.

📄 Content

FreeSense:Indoor Human Identification with WiFi Signals

Tong Xin, Bin Guo, Zhu Wang, Mingyang Li, Zhiwen Yu School of Computer Science Northwestern Polytechnical University
Xi’an, P. R. China guob@nwpu.edu.cn

Abstract—Human identification plays an important role in human-computer interaction. There have been numerous methods proposed for human identification (e.g., face recognition, gait recognition, fingerprint identification, etc.). While these methods could be very useful under different conditions, they also suffer from certain shortcomings (e.g., user privacy, sensing coverage range). In this paper, we propose a novel approach for human identification, which leverages WIFI signals to enable non-intrusive human identification in domestic environments. It is based on the observation that each person has specific influence patterns to the surrounding WIFI signal while moving indoors, regarding their body shape characteristics and motion patterns. The influence can be captured by the Channel State Information (CSI) time series of WIFI. Specifically, a combination of Principal Component Analysis (PCA), Discrete Wavelet Transform (DWT) and Dynamic Time Warping (DTW) techniques is used for CSI waveform-based human identification. We implemented the system in a 6m*5m smart home environment and recruited 9 users for data collection and evaluation. Experimental results indicate that the identification accuracy is about 88.9% to 94.5% when the candidate user set changes from 6 to 2, showing that the proposed human identification method is effective in domestic environments. Keywords—Human identification; WIFI sensing; channel state information; smart home; feature extraction. I. INTRODUCTION
Generally, human identification is based on one or more intrinsic physiological [1-4] or behavioral [5-7] distinctions, which either is related to the shape of the body (e.g., fingerprint, iris, palm print, face characters) or certain behavior patterns of a person (e.g., typing, gait, voice rhythms). It plays a significant role in the area of pervasive computing and human-computer interaction. Currently, fingerprint- [2], iris- [4], and vein-based methods [8] have been successfully employed in automatic human identification systems. However, these systems require the user to be close to the sensing device for accurate identification. Researchers also make numerous attempts to develop methods for behavioral biometrics (mainly gait analysis) using cameras, radars, or wearable sensors [7]. However, the vision-based approaches only work with line-of- sight coverage and rich-lighting environments, which also cause privacy concerns. The low cost 60 GHz radar solutions can only offer an operation range of tens of centimeters [9], and the devices are not widely deployed in our daily life. Finally, wearable sensor-based approaches require people to wear some extra sensors.
WIFI techniques have been widely used in our daily life. Due to its popularity and low-deployment-cost, numerous researchers have devoted to pervasive sensing using WIFI devices, such as indoor localization [10], gesture recognition [11], etc. These studies are mainly based on RSSI, i.e., the coarse-grained signal strength information. Quite recently, Channel State Information (CSI), i.e., fine-grained information regarding WIFI communication become available. Specifically, CSI describes how the signal propagates from the transmitter to the receiver and reflects the combined effects of the surrounding objects (e.g., scattering, fading, and power decay with distance). There are many subcarriers in CSI, each of which contains the information of attenuation and phase shift. Therefore, CSI contains rich information and is more sensitive to environmental variances caused by moving objects [12]. Several notable studies on pervasive sensing have been conducted using CSI, such as high-accuracy human localization [13], human activity recognition [14-18], and crowd counting [12]. This paper, however, presents a new application based on CSI sensing — human identification.
Due to the difference of body shapes and motion patterns, each person can have specific influence patterns on surrounding WIFI signals while she moves indoors, generating a unique pattern on the CSI time series of the WIFI device. In this paper, we propose a novel approach called FreeSense, which can identify human indoors based on CSI-enriched WIFI devices. It is supposed to work in home environments (usually 2-6 family members) and deliver personalized services when recognizing the identity of a family member. There are two technical challenges faced by FreeSense, as presented below.  Identification condition settlement and CSI time series segmentation. When a person moves around in the house, the influence level over multi-path communication of WIFI signals can be different when people walk across different path

This content is AI-processed based on ArXiv data.