Open AccessSecuriFi: Highly Robust Person Intrusion Sensing and Localization System Based on Wi-Fi Signals
Author(s) -
Daiyang Zhang,
Zhanjun Hao,
Xiaochao Dang,
Gaoyuan Liu
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1972/1/012022
Subject(s) - computer science , intrusion detection system , intrusion , classifier (uml) , channel state information , process (computing) , filter (signal processing) , computer security , real time computing , fingerprint (computing) , artificial intelligence , wireless , computer vision , telecommunications , geochemistry , geology , operating system
Home safety has always been a major concern for every family member. How to quickly sense intrusions while keeping costs low and provide users and police with accurate information about the intruder after sensing the intrusion has become an important challenge for every home security system. So we propose SecuriFi, a highly robust indoor person intrusion sensing and localization system based on Wi-Fi signals. SecuriFi uses Channel State Information (CSI) extracted from Wi-Fi signals as a medium for sensing intrusion and locating people, ensuring high system performance while effectively reducing costs. SecuriFi consists of two modules: intrusion sensing and localization. The intrusion sensing module senses the intrusion behavior by judging the change of signal energy. The localization module effectively removes the environmental noise by constructing a combined filter, and then uses an Extreme Learning Machine (ELM) as a classifier to process and form an offline fingerprint database, which maps the CSI data to the human location. In this paper, SecuriFi is verified in two different real-world environments, and the experimental results prove that SecuriFi has stronger sensitivity to intruders and high localization accuracy at the same time.