The Fusion of Sensors and Temporal Integration

Using Smartphone

Hani RAMADHAN

2015-06-05
Contents

1 Introduction 1
1.1 Motivational Background . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Aim . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

2 Materials and Methods 3

2.1 Three dimensional position and orientation . . . . . . . . . . . . 3
2.2 Smartphone sensors . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.2.1 Accelerometer . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2.2 Gyroscope . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2.3 Magnetometer . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2.4 Sensor errors . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.3 Radio signal strength indication of Wi-Fi/802.11 . . . . . . . . . 4
2.4 Kalman Filter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.5 Complementary Filter . . . . . . . . . . . . . . . . . . . . . . . . 4
2.6 Wi-Fi Fingerprinting . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.7 Weighted k-Nearest Neighbor . . . . . . . . . . . . . . . . . . . . 4

i
Chapter 1

Introduction

1.1 Motivational Background

During cultural or tourism visits, a guide of the interesting objects is necessary
to enhance the knowledge and the experience of the visitors. In these times,
the mobile application using smartphones and tablets are possible to serve as an
autonomous guide personally, in the case of context-aware system. The elements
of the contexts of the users which mainly used in the visits are the positions,
which will change in timely manner.
The guidance approach using the smartphone is composed in 3 dimensions,
which are positions in X, Y, and Z, and the orientation in X, Y, and Z. Those
dimensions are used in order to explore the possibility of smartphone of being
pointer to interesting object, not for the position estimator only. Thus, the user
experience should be possible to be increased, as they dont get vision-blocked
(by always looking at the screen) or audio-blocked (by using an audio guidance
headphones).
Unfortunately, determining the position and the orientation of the smart-
phone as the pointing guide, which is held by users, is probably not so accurate,
especially in the indoor environment of a building without the aid of GPS [1].
The inertial navigation system [2] sensors included in a smartphone have many
disadvantages. For example, accelerometer [3] to measure the displacement of
the smartphone, but it tends to be highly noisy. Yet, the gyroscope [4], in mea-
suring orientation, has some drawbacks in estimating the orientation since it has
integration error called drifts. While the magnetometer [5] is easily disturbed
by the magnetic field that produced by metal materials around the smartphone.
Thus, an accurate estimation approach for the position and the orientation of
the smartphone is important.
In the other hand, the approach in doing a combination or sensor fusion
towards the smartphone sensors yields in a fine result [6–8] using some filters,
namely Complementary Filter [7] and Kalman Filter [9]. However, these internal
sensors are not very adequate to deal with precision of the trajectory. Hence, the

1
external environment, such as WiFi, is used to improve the localization quality
of the smartphone location and orientation [10, 11].
Some research results have considered the orientation estimation using 3D
Accelerometer and 3D magnetometer sensor fusion using Kalman Filter [4],
combination of accelerometer, magnetometer, and gyroscope sensors in smart-
phone [1], and combination of heading (orientation in Z axis) plus the position
tracking to do Indoor Positioning [5, 6, 7]. While the 3D position and orien-
tation estimation by Kalman Filter have been provided by [8], it uses camera
sensor, which in the purposed study it is not considered in this study because it
blocks the vision. The more sophisticated heading and position estimation us-
ing Particle Filter and Dead Reckoning using Android Phone sensors and WiFi
Fingerprinting has been conducted by [9], which positioning error reached up
to 6 meters. Then, a newer approach of combining the WiFi Weighted Path
Loss, Pedestrian Dead Reckoning using Smartphone sensors, and Landmarks
[10] also give a good positioning error result about 1 meter. Without WiFi aid,
the native sensors of smartphone [7] provided small error about 0.3
Then, the purpose of this study is to explore the possibility to gain the 3D
position and orientation estimation by the smartphone and WiFi fingerprinting.
Thus, the smartphone can be a decent yet affordable tool to be a guide in a
visit, yet its possibility to be an aid for the visually impaired also.

1.2 Aim
The aim of this project is to study the possibility of data fusion of smartphone
sensors and Wi-Fi fingerprinting to make a smartphone as a precise pointing
object.

2
Chapter 2

Materials and Methods

In this section, the materials and method related to this study will be explained.
The related material studied are smartphone 3D orientation and position; sensor
aspects related to smartphone, which are accelerometer, gyroscope, and mag-
netometer; and Wi-Fi Radio Signal Strength Indication (RSSI). The methods
that will be explained are the Kalman Filter; Complementary Filter; and Wi-
Fi fingerprinting, where the deterministic positioning algorithm, the weighted
k-Nearest Neighbor, was applied.

2.1 Three dimensional position and orientation

The three dimensional position and orientation in this study is represented in 3-
axis coordinates X, Y, Z and their representation in the smartphone local frame.
This representation of the position and orientation is important because the
sensors only know their position and orientation locally. In further application,
this local reference can be transformed into global reference. This representation
is depicted in ....
The orientation angles in 3D are called roll, pitch, and yaw (or azimuth, in
several references, for example [?]) in X, Y, Z axes respectively. Even though
they are measured locally by the smartphone, if the help of digital compass (will
be explained later) is being used, the direction towards north global reference
can be approximated.

2.2 Smartphone sensors

The smartphone sensors which will be studied in this project are the accelerome-
ter, gyroscope, and the magnetometer. The accelerometer measures the acceler-
ation of the smartphone movement, the gyroscope measures the rotation of the
orientation of the smartphone, and the magnetometer measures the magnetic
field around the smartphone.

3
2.2.1 Accelerometer
2.2.2 Gyroscope
2.2.3 Magnetometer
2.2.4 Sensor errors

2.3 Radio signal strength indication of Wi-Fi/802.11

2.4 Kalman Filter
2.5 Complementary Filter
2.6 Wi-Fi Fingerprinting
2.7 Weighted k-Nearest Neighbor

4
Bibliography

[1] U. Shala and A. Rodriguez, “Indoor positioning using sensor-fusion in an-

droid devices,” 2011.

[2] O. J. Woodman, “An introduction to inertial navigation,” University of

Cambridge, Computer Laboratory, Tech. Rep. UCAMCL-TR-696, vol. 14,
p. 15, 2007.
[3] J. Doscher and M. Evangelist, “Accelerometer design and applications,”
Analog Devices, vol. 3, 1998.

[4] “Ieee standard specification format guide and test procedure for coriolis
vibratory gyros,” IEEE Std 1431-2004, pp. 1–78, Dec 2004.
[5] D. Hovde, M. Prouty, I. Hrvoic, and R. Slocum, “Commercial magnetome-
ters and their application,” Optical Magnetometry, p. 387, 2013.

[6] T. Ozyagcilar, “Implementing a tilt-compensated ecompass using ac-

celerometer and magnetometer sensors,” Freescale semiconductor, AN,
vol. 4248, 2012.
[7] S. Ayub, A. Bahraminisaab, and B. Honary, “A sensor fusion method for
smart phone orientation estimation,” in Proceedings of the 13th Annual
Post Graduate Symposium on the Convergence of Telecommunications,
Networking and Broadcasting, Liverpool, 2012.
[8] B. Ando, S. Baglio, C. O. Lombardo, and V. Marletta, “An advanced
tracking solution fully based on native sensing features of smartphone,” in
Sensors Applications Symposium (SAS), 2014 IEEE, pp. 141–144, IEEE,
2014.
[9] G. Welch and G. Bishop, “An introduction to the kalman filter,” tech. rep.,
Chapel Hill, NC, USA, 1995.
[10] O. Costilla-Reyes and K. Namuduri, “Dynamic wi-fi fingerprinting indoor
positioning system,” in International Conference on Indoor Positioning and
Indoor Navigation, vol. 27, p. 30th, 2014.

5
[11] Z. Chen, H. Zou, H. Jiang, Q. Zhu, Y. C. Soh, and L. Xie, “Fusion of
wifi, smartphone sensors and landmarks using the kalman filter for indoor
localization,” Sensors, vol. 15, no. 1, pp. 715–732, 2015.