A Passive Localization Scheme Based on Channel State Information in an Indoor
Environment
Hongli Yu
Chuzhou University
Anhui, China
hly @chzu.edu.cn
Guilin Chen
Chuzhou University
Anhui, China
glchen @chzu.edu.cn
Shenghui Zhao
Chuzhou University
Anhui, China
shzhao@ah.edu.cn
Chih-Yung Chang
Tamkang University
New Taipei City, Taiwan
cychang@mail.tku.edu.tw
Abstract
Localization is an important issue and has been
applied in many applications. Many localization mechanisms
have been developed in the last decade, including GPS-based,
Wi
Fi RSSI Signal based as well as BLE Beacon based
localizations. The common disadvantages of the existing
localization mechanisms is that they belong the class of Active
Localization which requires the localization target carrying a
receiver such as mobile phone or navigator. Compared with
the passive localization mechanisms, the active localization
mechanisms cause inconvenient and create hardware cost.
This paper proposed a CSI-based Passive Localization
Mechanism, or called CPLM in short which identifies the
human locations in the indoor environment based on Channel
Impulse Response
(CIR) information. Experiment results
reveal that the proposed CPLM outperforms existing works in
terms of localization accuracy, latency as well as the reliability.
Keywords—LBS; Indoor passive localization; CSI; Wi
Fi
I. I
NTRODUCTION
With the development of the economics, localization
plays a crucial role in the raising demand of various
location-based services (LBS), such as the building
surveillance, health care, smart home, and assisted living
[1-5]. In the past years, the LBS has been widely applied in
the outdoor applications, including locating the typical
buildings, tracing the parcel, providing the current location
information for navigator and so on. The Global Positioning
Satellite (GPS) is one of the most used technology for
outdoor localization[6]. However, the GPS is unable to be
used in indoor localization. It is still a big challenge for
indoor localization.
The indoor localization using wireless signals and
infrastructure has become a hot topic due to the lower cost,
ease of access and applying the common wireless
infrastructure. In literature, a large number of indoor
localization schemes and systems have been proposed based
on wireless signals [7---9]. In the earlier indoor localization
schemes, the infrared, laser and ultrasonic technologies are
used to identify position of the target. However, these
schemes require high cost and use bulky devices. To cope
with the problem existing in those schemes, lots of indoor
localization systems based on various signal parameters
have been proposed, such as the signal-based localization
system named Received Signal Strength Indicator (RSSI),
the time-based localization system called Time of Arrival
(TOA) and the Time Difference of Arrival (TDOA), and the
angle-based localization system called Angle of Arrival
(AOA). However, the localization accuracy decreased with
the distance between the transmitter and receiver. These
localization schemes can be classified into two categories:
active localization and passive localization mechanisms.
The active localization mechanism needs the target
carries the fixed device, such as mobile phone, and actively
participate the localization process. Typically, the active
localization system consists of two parts, the terminal device
and the transmitter. These systems require the device (smart
phone) carried by target to scan the signal from different
access points. Although the accuracy of localization is
enhanced, leading to high cost.
In [10], the authors proposed the Radio Tomographic
Imaging (RTI) based on Received Signal Strength
Information (RSSI). This mechanism achieves the passive
localization by densely deploying the sensors in the
monitoring environment. When the target appears in the
environment, the RSSI in part links will be changed. Then
the distance from the transmitter to the receiver can be
calculated easily and hence the location of the target can be
distinguished. However, the localization accuracy of the
proposed system is reduced due to the multipath effect and
the spatial and temporal characteristics.
To cope with the problem found in active localization
mechanisms, many passive localization mechanisms [11][12]
have been proposed. In these mechanisms, the target need
not to participate the process of localization. Several
transmitter and receiver must be deployed to the localization
scene in advance. A server is used to receive the information
from the receiver, and then execute the passive localization
algorithms to identify the target location.
In [11], the authors proposed an FIFS system based on
the weighted average Channel State Information (CSI) value.
The laptop is equipped with an off-the-shelf WiFi network
card (NIC 5300) to extract the CSI. The stability of the
localization system has been improved. Furthermore, the
authors also proposed another localization in [12] called
PinLoc system. This system also considers CSI as the
feature information, the training data is obtained in a mining
space, thus leading to high
war-driving.
All of the mechanisms mentioned above aim to cope
with the indoor passive localization problem. However, their
accuracies still can be
improved and most of them lack real-
time demand. This paper presents a CSI-based Passive
Localization Mechanism, or called CPLM, which utilizes the
Channel Impulse Response (CIR) information to identify the
presence location of the person.
The remainder of this paper is organized as follows.
Section 2 presents the proposed mechanism. Section 3
presents the performance of the proposed schemes while
2017 31st International Conference on Advanced Information Networking and Applications Workshops
978-1-5090-6231-7/17 $31.00 © 2017 IEEE
DOI 10.1109/WAINA.2017.85
576