Journal of Computer Science 6 (10): 1206-1212, 2010
ISSN 1549-3636
© 2010 Science Publications
Corresponding Author: Hamid Mehmood, Department of Remote Sensing and Geographic Information Systems,
School of Engineering and Technology, Asian Institute of Technology, P.O. Box 4, Klong Luang,
Pathumthani 12120, Thailand
1206
Indoor Positioning System Using Artificial Neural Network
Hamid Mehmood, Nitin K. Tripathi and Taravudh Tipdecho
Department of Remote Sensing and Geographic Information Systems,
School of Engineering and Technology, Asian Institute of Technology,
P.O. Box 4, Klong Luang, Pathumthani 12120, Thailand
Abstract: Problem statement: Location knowledge in indoor environment using Indoor Positioning
Systems (IPS) has become very useful and popular in recent years. A number of Location Based
Services (LBS) have been developed, which are based on IPS, these LBS include asset tracking,
inventory management and security based applications. Many next-generation LBS applications such
as social networking, local search, advertising and geo-tagging are expected to be used in urban and
indoor environments where GNSS either underperforms in terms of fix times or accuracy, or fails
altogether. To develop an IPS based on Wi-Fi Received Signal Strength (RSS) using Artificial Neural
Networks (ANN), which should use already available Wi-Fi infrastructure in a heterogeneous
environment. Approach: This study discussed the use of ANN for IPS using RSS in an indoor wireless
facility which has varying human activity, material of walls and type of Wireless Access Points
(WAP), hence simulating a heterogeneous environment. The proposed system used backpropogation
method with 4 input neurons, 2 output neurons and 4 hidden layers. The model was trained with three
different types of training data. The accuracy assessment for each training data was performed by
computing the distance error and average distance error. Results: The results of the experiments
showed that using ANN with the proposed method of collecting training data, maximum accuracy of
0.7 m can be achieved, with 30% of the distance error less than 1 m and 60% of the distance error
within the range of 1-2 m. Whereas maximum accuracy of 1.01 can be achieved with the commonly
used method of collecting training data. The proposed model also showed 67% more accuracy as
compared to a probabilistic model. Conclusion: The results indicated that ANN based IPS can provide
accuracy and precision which is quite adequate for the development of indoor LBS while using the
already available Wi-Fi infrastructure, also the proposed method for collecting the training data can
help in addressing the noise and interference, which are one of the major factors affecting the accuracy
of IPS.
Key words: Artificial neural networks, indoor positioning systems, Wi-Fi, backpropogation method
INTRODUCTION
Wi-Fi Received Signal Strength (RSS) has drawn
great attention in recent years for the development of
Indoor Positioning Systems (IPS), because of its
distinct advantages like already existing infrastructure
and wide coverage area. With numerous new Wi-Fi
nodes being added daily, the global Wi-Fi coverage is
continuously growing. Recent tests have shown that
indoor positioning with Wi-Fi RSS based Positioning
Systems (WPS) can achieve positioning accuracy of 1 to
4 meter indoors and 10-40 m in the outdoor environment
(Prasithsangaree et al., 2002; Doherty et al., 2001;
Hightower and Borriello, 2001).
A typical WPS consists of a Mobile User (MU),
wireless Access Point (AP) and a positioning algorithm.
There are three basic methods to compute the location
of a MU in indoor envrionment; (a) Angle of Arrival
(AoA), (b) Time of Arrival (ToA)/Time Difference of
Arrival (TDoA) (Yamasaki et al., 2005) and (c)
Received Signal Strength (RSS) (Kolodziej and Hjelm,
2006).
AoA is based on the time of flight of
electromagnetical waves, which requires very accurate
timming measurements and thus implementation of
AoA requires additional hardware. ToA and TDoA are
based on the measurement of the time flight of the
signal, the implementation of this technique requires