Device-Free Human Activity Recognition Based on GMM-HMM Using Channel State Information

Xiaoyan Cheng,Binke Huang,Jing Zong

doi:10.1109/access.2021.3082627

Xiaoyan Cheng, Binke Huang + Show 1 more

Open Access

https://doi.org/10.1109/access.2021.3082627

Copy DOI

Journal: IEEE Access	Publication Date: Jan 1, 2021
Citations: 17	License type: CC BY 4.0

Affiliation: Xi'an Jiaotong University

Abstract

This paper presents a machine learning method, Gaussian Mixture Hidden Markov Model (GMM-HMM), for device-free activity recognition using WiFi channel state information (CSI). The basic concept of CSI is introduced and signal changes caused by human activity are described, which demonstrates that human activity can be identified using a unique mapping between action and signal variations. The phase difference expanded matrix is built by the mean and standard deviation of phase difference as feature matrix after linear correction and Savitzky-Golay filter is performed on the CSI raw phase information. The GMM-HMM is used for classification as the human activity can be modeled as the Markov process and the complex activity patterns can be fitted by multiple Gaussian density functions, respectively. The proposed system is verified on the self-collected datasets and several factors affecting the recognition accuracy are analyzed. Furthermore, the system has compared with the previous work. High accuracy and robustness in universal scenarios are realized. Experimental results show that the average recognition accuracy of the proposed system is over 97%.

Highlights

Recent years have witnessed increasing research interest in human activity recognition as it benefits multiple applications, such as intrusion detection [1], [2], smart homes [3], and health care services [4]
With the popularity of WiFi devices and the rich channel characteristics of channel state information (CSI), human activity recognition based on WiFi CSI has attracted widespread attention
BASIC CONCEPT OF CSI CSI is a fine-grained signal feature captured from the physical layer (PHY) of the WiFi communication via Orthogonal Frequency Division Multiplexing (OFDM) technology

Summary

INTRODUCTION

Recent years have witnessed increasing research interest in human activity recognition as it benefits multiple applications, such as intrusion detection [1], [2], smart homes [3], and health care services [4]. With the popularity of WiFi devices and the rich channel characteristics of channel state information (CSI), human activity recognition based on WiFi CSI has attracted widespread attention. Traditional recognition methods, such as sensor-based applications, usually require users to wear or attach smart devices, which increases inconvenience and obstruction for users [5], [6]. Vision-based human activity recognition methods have the problems of privacy security invasion and susceptibility to environmental influences such as light interference [7]. 2) HMM is selected as a WiFi signal-based recognition methodology to enhance robustness and adaptability in the work.

DATA PREPARATION AND FEATURE EXTRACTION

GMM BASICS

HMM BASICS

EXPERIMENTS AND EVALUATION

Findings

CONCLUSION