A WiFi-Based Passive Sensing System for Human Presence and Activity Event Classification

Wenda Li, Bo Tan, Robert Piechocki

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

Human presence detection and activity event classification are of importance to a variety of context-awareness applications such as e-Healthcare, security and low impact building. However, existing Radio Frequency IDentification (RFID) tags, wearables and passive infrared (PIR) approaches require the user to carry dedicated electronic devices, suffer low detection accuracy and false alarm problems. This paper proposes a novel system for non-invasive human sensing by analyzing the Doppler information contained in the human reflections of WiFi signal. Doppler information is insensitive to the stationary objects, thus no requirement on scenario-specific calibration that makes it ideal for human sensing. We also introduce the time-frequency domain feature vectors of WiFi Doppler information for the Support Vector Machine (SVM) classifier towards activity event recognition. The proposed methodology is evaluated on a Software Defined Radio (SDR) system together with the experiment of five different events. The results indicate the proposed system is sufficient for indoor context-awareness with 95% overall accuracy in event classification and 84.8% PPV & 93.3% TPR in human presence detection which outperforms the traditional Received Signal Strength (RSS) approach (62.5% PPV and 83.3% TPR).
Original languageEnglish
Pages (from-to)276-283
JournalIET Wireless Sensor Systems
Volume8
Issue number6
Early online date23 Aug 2018
DOIs
Publication statusPublished - Dec 2018

Fingerprint

Radio systems
Radio frequency identification (RFID)
Support vector machines
Classifiers
Calibration
Infrared radiation
Experiments

Keywords

  • Passive Radar
  • Healthcare
  • Activity Recognition
  • Sensing

ASJC Scopus subject areas

  • Engineering(all)

Cite this

A WiFi-Based Passive Sensing System for Human Presence and Activity Event Classification. / Li, Wenda; Tan, Bo; Piechocki, Robert.

In: IET Wireless Sensor Systems, Vol. 8, No. 6, 12.2018, p. 276-283.

Research output: Contribution to journalArticle

Li, Wenda ; Tan, Bo ; Piechocki, Robert. / A WiFi-Based Passive Sensing System for Human Presence and Activity Event Classification. In: IET Wireless Sensor Systems. 2018 ; Vol. 8, No. 6. pp. 276-283.
@article{4472a2debf6e433f9d8c4346c51d4b8c,
title = "A WiFi-Based Passive Sensing System for Human Presence and Activity Event Classification",
abstract = "Human presence detection and activity event classification are of importance to a variety of context-awareness applications such as e-Healthcare, security and low impact building. However, existing Radio Frequency IDentification (RFID) tags, wearables and passive infrared (PIR) approaches require the user to carry dedicated electronic devices, suffer low detection accuracy and false alarm problems. This paper proposes a novel system for non-invasive human sensing by analyzing the Doppler information contained in the human reflections of WiFi signal. Doppler information is insensitive to the stationary objects, thus no requirement on scenario-specific calibration that makes it ideal for human sensing. We also introduce the time-frequency domain feature vectors of WiFi Doppler information for the Support Vector Machine (SVM) classifier towards activity event recognition. The proposed methodology is evaluated on a Software Defined Radio (SDR) system together with the experiment of five different events. The results indicate the proposed system is sufficient for indoor context-awareness with 95{\%} overall accuracy in event classification and 84.8{\%} PPV & 93.3{\%} TPR in human presence detection which outperforms the traditional Received Signal Strength (RSS) approach (62.5{\%} PPV and 83.3{\%} TPR).",
keywords = "Passive Radar, Healthcare, Activity Recognition, Sensing",
author = "Wenda Li and Bo Tan and Robert Piechocki",
year = "2018",
month = "12",
doi = "10.1049/iet-wss.2018.5113",
language = "English",
volume = "8",
pages = "276--283",
journal = "IET Wireless Sensor Systems",
issn = "2043-6386",
publisher = "Institution of Engineering and Technology",
number = "6",

}

TY - JOUR

T1 - A WiFi-Based Passive Sensing System for Human Presence and Activity Event Classification

AU - Li, Wenda

AU - Tan, Bo

AU - Piechocki, Robert

PY - 2018/12

Y1 - 2018/12

N2 - Human presence detection and activity event classification are of importance to a variety of context-awareness applications such as e-Healthcare, security and low impact building. However, existing Radio Frequency IDentification (RFID) tags, wearables and passive infrared (PIR) approaches require the user to carry dedicated electronic devices, suffer low detection accuracy and false alarm problems. This paper proposes a novel system for non-invasive human sensing by analyzing the Doppler information contained in the human reflections of WiFi signal. Doppler information is insensitive to the stationary objects, thus no requirement on scenario-specific calibration that makes it ideal for human sensing. We also introduce the time-frequency domain feature vectors of WiFi Doppler information for the Support Vector Machine (SVM) classifier towards activity event recognition. The proposed methodology is evaluated on a Software Defined Radio (SDR) system together with the experiment of five different events. The results indicate the proposed system is sufficient for indoor context-awareness with 95% overall accuracy in event classification and 84.8% PPV & 93.3% TPR in human presence detection which outperforms the traditional Received Signal Strength (RSS) approach (62.5% PPV and 83.3% TPR).

AB - Human presence detection and activity event classification are of importance to a variety of context-awareness applications such as e-Healthcare, security and low impact building. However, existing Radio Frequency IDentification (RFID) tags, wearables and passive infrared (PIR) approaches require the user to carry dedicated electronic devices, suffer low detection accuracy and false alarm problems. This paper proposes a novel system for non-invasive human sensing by analyzing the Doppler information contained in the human reflections of WiFi signal. Doppler information is insensitive to the stationary objects, thus no requirement on scenario-specific calibration that makes it ideal for human sensing. We also introduce the time-frequency domain feature vectors of WiFi Doppler information for the Support Vector Machine (SVM) classifier towards activity event recognition. The proposed methodology is evaluated on a Software Defined Radio (SDR) system together with the experiment of five different events. The results indicate the proposed system is sufficient for indoor context-awareness with 95% overall accuracy in event classification and 84.8% PPV & 93.3% TPR in human presence detection which outperforms the traditional Received Signal Strength (RSS) approach (62.5% PPV and 83.3% TPR).

KW - Passive Radar

KW - Healthcare

KW - Activity Recognition

KW - Sensing

U2 - 10.1049/iet-wss.2018.5113

DO - 10.1049/iet-wss.2018.5113

M3 - Article

VL - 8

SP - 276

EP - 283

JO - IET Wireless Sensor Systems

JF - IET Wireless Sensor Systems

SN - 2043-6386

IS - 6

ER -