See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/329396426

TELIL: A Trilateration and Edge Learning based Indoor Localization Technique

for Emergency Scenarios

Conference Paper · September 2018

DOI: 10.1109/ICACCI.2018.8554587

CITATIONS READS
2 506

2 authors:

Adwitiya Mukhopadhyay Aswin Mallissery

Amrita School of Arts and Sciences Mysuru Amrita Vishwa Vidyapeetham
27 PUBLICATIONS   41 CITATIONS    1 PUBLICATION   2 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Telemedicine View project

ICT in Education View project

All content following this page was uploaded by Adwitiya Mukhopadhyay on 17 January 2019.

The user has requested enhancement of the downloaded file.

 TELIL: A Trilateration and Edge Learning based
Indoor Localization Technique for Emergency
Scenarios
Adwitiya Mukhopadhyay1, Aswin Mallissery2
Department of Computer Science1,2
Amrita School of Arts and Sciences, Mysuru
Amrita Vishwa Vidyapeetham, India
1
[email protected], [email protected]

Abstract—When deploying localization techniques using the approximate position of the WiFi-enabled mobile device
wireless devices with only less amount of system resources, along with WiFi access points(APs). When it comes to mobile
indoor WiFi localization is a difficult task. GPS (Global positioning system there are many problems faced, such as the
Positioning System) can provide a fairly accurate position of the ones listed in [2], including radio propagation modelling and
user, but it is usually not used indoors because of the degradation
the infrastructure that has been deployed.
of signals due to the structure of the building. Due to the
unsteady nature of Received Signal Strength Indicator (RSSI), The efficiency of a localization algorithm also depends on
accurate results are hard to generate, in spite of the usage of which signalling technique is used and the system architecture.
various alternative WiFi localization techniques. We have There are various indoor localization techniques that help the
attempted to improve the accuracy of localizing the user by
implementing wall detection and pixel calculation to trilateration
mobile users to find his/her approximate relative position from
technique in smartphones. The users having WiFi enabled the reference points. The techniques that are used to estimate
devices like a smartphone can locate their position and request the positions are the angle of arrival (AOA), Received Signal
for help in case of emergency with the implementation of WiFi Strength Indicator (RSSI), time of arrival (TOA), etc. One
localization on mobile devices. This study proves that the more technique called fingerprinting was introduced [3] which
selection of access points used for generating the position of the stores a set of unique access points and their locations in the
user substantially affects the accuracy. database. The drawback of fingerprinting is that as the scaling
unit of the observed area increases, the size of the database
Keywords—WiFi Localization, Trilateration, Edge Learning,
increases. A more advanced fingerprinting technique is
Indoor Localization, TELIL
presented in [4], [5] where finding the closest match is
I. INTRODUCTION calculated using a probabilistic approach that improved the old
approach of finding the closest match.
In the recent years, mobile phones started to play a
significant role in our day to day life because of the rapid For outdoor localization, to estimate the user location a
development of mobile smartphone technologies. The total common approach is used, GPS [6]. But in case of indoor
amount of smartphone users grows dramatically every year as localization, there are no standard technologies. The various
smartphones have become affordable and easily accessible to technologies that have been presented to locate the user when
the population. indoors are radio frequency [1], computer vision [7], laser [8],
cellular communication [9], ultrasound [10], etc.
In this paper, we are using a smartphone to locate the user.
Localization can be used in many scenarios like navigating a Google localization service is one of the most famous
blind person, keeping track of portable devices and even techniques which combines tower cell ID, WiFi and GPS on
surveilling patients in hospitals. In this paper, we are focusing indoor Google Maps [11] that provides navigation guidance
on localizing a victim in an emergency scenario. Most of the inside the building, but the accuracy of this system is not
smartphones which are currently in the market are generally enough to provide this service yet. There is another alternative
integrated with GPS (Global Position System) module which method Ekahau [12], that introduced software and hardware
will provide fairly accurate geographic coordinates with the for tracking users and portable object, health monitoring and
help of satellite signals as reference points. But when security among other applications. In this paper the measuring
deployed inside a building the signals from the satellite cannot metric which has been used to generate the user’s location is
propagate properly through building structure and hence this RSSI. There are many works which have already implemented
limits its functionality. Thus, an improved indoor localization triangulation technique. In this paper we are introducing a new
technique with more accuracy is required. method, Trilateration and Edge Learning based Indoor
Localization (TELIL), in order to generate a more accurate
WiFi localization technique using RADAR to locate users
result which helps in finding the victims in emergency
inside a building was introduced by [1] where the authors use
scenarios more easily.

978-1-5386-5314-2/18/$31.00 ©2018 IEEE 6

 II. RELATED WORKS deliver an accuracy of 1.5 meters. In [21] as the author has
proposed a model to navigate elderly and physically
A. Works Related to Triangulation Technique
challenged people in an automated powered wheelchair. The
The WiFi localization technique using RADAR was system first maps the floor plan by dividing it into small grids
introduced in [1] where the system uses APs RSSI value and and address each grid uniquely. Using Location-Aware and
their positions to find the user location by a triangulation Remembering Navigation (LARN) algorithm the system finds
technique. In [13] as the author proposed a method that first the patch to the destination. In [22] as the author has proposed
identifies the neighbouring landmarks in the environment and a method that constructs the floor plan by user motions.
it checks with the corresponding landmarks with the map to According to the construction the system design Location in
measure the bearings of landmarks each other and find the Fingerprint Space (LiFS), a localization technique based on
position efficiently. In [14] as the author proposes a method mobile phones and WiFi infrastructure.
for all nodes to find its position and orientation in an ad-hoc
network where only some nodes have the capability of D. Works Related to Localization in Smartphones
positioning in an assumption that each node in the network has In [24] as the author has proposed a peer-assisted
AOA capability. In [15] as the author proposed a method that localization technique that eliminates large errors in
uses Extended Kalman Filter (EKF) with state-vector that has localization technique. The system fetched the distance from
the required information of the external angular measurement. peer phones and then plot the locations against WiFi signature
The system also keeps track of the robot odometric map. As per the experiment, the system was able to reduce the
information with its angle the robot has moved to the actual error rate up to 80%. In [23] as the author has proposed
measurement. This helps inconsistent use of triangulation CUPID, that free from relying on the deployment of additional
method the keep track of the position of the robot at any time infrastructure, or burdensome fingerprinting restrictions.
during its motion. In [16] as the author used a triangulation CUPID used physical layer (PHY) information that only
technique to analyze the relevant signals. The system was fetches the direct path and signal strength. Using CUPID, the
introduced to predict the landslides during rainfalls. The system was able to produce a more accurate result and was
system was able to successfully detect the landslides. able to achieve 2.7 meters’ median localization error. In [25]
as the author has proposed a method for position estimation
B. Works Related to Trilateration Technique
technique that reduces the network traffic by sharing the
In [17] as the author proposed a method to localize the distance between each node to other nodes. After distance
position of a node and its neighbouring nodes by calculating estimation of unknown location nodes is done, then the
the distance between all the neighbouring nodes and it shares position can be computed using the coordinates of the known
the distance with the neighbouring node which forms as a nodes. In [26] as the author has implemented a method which
cluster. Using this cluster with the help of 3 nodes it is records the peer phones in the cluster by COTS sound
possible to find the 4th node position using trilateration hardware to reduce the error rate to overcome the limitations
technique. This helps in tolerating noise and more accurate of WiFi approaches. The system was able to reduce the error
position can be produced. In [18] as the author proposed a rate from 6-8 meter to 2 meters. In [27] as the author has
method which can estimate the location of any sensor in a proposed a method that uses consumer’s smartphones and
Wireless Sensor Networks (WSN). The distance between any existing WiFi access points to reduce the cost and installation
sensor can be obtained accurately by expected hops from a time. The system uses fingerprinting technique to identify the
network model. This method gives less communication location dependency and calculate the distance based on RSSI
overhead and better performance when compared to other value to find the location of the users. With the usage of the
existing schemes. In [19] as the author addresses the problems autonomous mobile robot, the system was able to calibrate the
that occur in localization from the viewpoint of WSN. The signal strength fastly. In [28] as the author has proposed a
author has mentioned the most commonly used localization model to find the location of a user in an indoor environment
methods and the scenarios which can be used. All the study by using GPS, gravity sensors, temperature sensors,
was done based on calculating distance using RSSI. photosensors and compass to extract the information required
C. Works Related to Fingerprinting for localization. The system was able to generate an accurate
logical conclusion about the location and also evaluate in both
In [20] as the author has implemented a real-time
stationary and moving condition. In [29] as the author has
implementation in the Android device by training the system
implemented a new technique called Guoguo, that will fill the
to construct a radio map that uses RSSI values at different
long-lasting gap of mobile-based indoor localization. Guoguo
points in the building and stored in the database to generate
used anchor networking and coordinate protocol in order to
the position of the user. The drawback is that the size of the
transmit modulated localization beacons with the help of high-
database increases when the area of the building and the point
band acoustic signals.
which need to get recorded increases. The system was able to

7
 III. METHODOLOGY
TELIL is a new methodology introduced in our paper Localization Phase
which combines both trilateration and edge learning methods (online)
to provide a more accurate result.
The system will record all the reference access points that
User sends the query
are installed in the building and its location is fetched using
the floor architecture. After recording all the reference access
points, the system will record all the lost RSSI values that
have occurred when the signal passes through the wall. In this Calculate the position
paper, we are only considering the walls between the reference
access points and the user. This is a one-time process until if
there are any changes made to the architecture. Once the
system learning is over, using the android app user can request Plots on the floor architecture
for the location by sending a query which will calculate the
position. After processing the request, the location of the user
will be plotted in the floor architecture and the coordinates of Saves the location in the server
the user location will be saved on the server. Rescue team can
access this coordinates which will help in finding the victim
easier.
A. Framework Fig. 2 Localization phase
The framework of the research implements the standard
structure of android which communicates with the database to i. Learning Phase
store and retrieve all the information required. To make the
application lightweight all the data are stored in the database. In learning phase first, we need to set the IP address and
The application is connected with MySQL database and the connect it to the server. Once the connection is established, a
floor architecture image is accessed with the help of encoding list of all the floor architecture present in the server will be
the image to JSON format for faster retrieval. The system uses fetched. From the list, we need to select to which floor
RSSI value to find the distance between each reference access architecture we need to add reference access points and select
points and using trilateration technique the position of the user add WiFi option. The system will fetch the floor architecture
is predicted. The system is divided into two phases, Learning and we need to select the position of the reference access
phase and Localization phase. points and add the access point which is at that position by
selecting add friendly WiFi option. The system will generate
all the access points available and the required access points
Learning Phase should be selected. Authentication is done at this stage to
(offline) make sure that there won’t be same access points registered at
two different positions. For this, the system makes use of
Select the room architecture MAC address which is unique for every access points. After
authentication system will record all the information required
like SSID, MAC address, the coordinated and the RSSI value
at that position to the database. Once all the reference access
Add all the reference access points
points are stored the next task is to record the lost signals
between the walls by selecting learn edge option which starts
tracking the signals near to the wall and finding the
Record three best access points RSSI combination of all the possible lost signals with the best three
value near to the wall reference access points. The system will fetch the data of only
the top three reference access points with high RSSI value by
comparing with the access points registered in the database.
Add all the lost RSSI value between the
walls to the database

Fig. 1 Learning Phase

8
 i. Localization Phase B. Noise Reduction
Due to the instability of WiFi signals, noise reduction is
Start achieved according to [30]. It iterates the scanning process 3
times and records the RSSI value and find the mean to reduce
the noise. In our system to increase the accuracy, we are
Scan for all the available iterating 10 times and calculate the mean value. When there
Wi-Fi signals are more obstacles as per the study it is always good to
increase the number of iteration so that more accurate result
can be obtained. In [31] the author has used Cayley-Menger
Select all Wi-Fi success points registered distance concept to make corrections to the noisy distance
to the system measurements done in [32]. The method only uses quadratic
solver to satisfy L2 norm optimization problem. By reducing
the computational complexity, the system was able to reduce
the usage of energy.
If(Wi-Fi Signals < 3)
YES C. Edge Learning
NO The system keeps on reading all the RSSI value near to the
Select the highest 3 RSSI value walls and saves it in the database as the user moves from one
end to the other end of the wall. Likewise, all the reading is
taken and saved in the server. At the time of processing the
Calculate RSSI all signals signals, the system will calculate the mean value of the signal
strength by 10 readings to improve the accuracy. The system
saves all the details like the distance from the reference points
to the wall and its RSSI value of all the three reference points
If (signals in Localize at that position. At the time of distance calculation, the system
YES database) NO will look into this data to check whether there are any access
points that are not in the line of sight of the user.
D. Distance Estimation
Calculate the new distance by
considering the obstacles in Location One of the main challenges which are faced in WiFi
middle and localize localizations is its frequency changes with a change in
environment. The accuracy of estimating the location is
Fig. 3 Localization algorithm
limited because of the fluctuating nature of RSSI
measurements. Even if the user is steady at a position the
In localization phase, the user needs to select find me frequency of signal changes frequently. So before calculating
option which will load all the required data to the device and the distance, the system will reduce the noise by iterating 10
user need to select generate query option that fetches the best times and find the mean value. With this new RSSI value
three reference access points at the user’s location and checks distance will be calculated using Free Space Path Loss (FSPL)
with the lost RSSI values in the dataset. If the system finds using the following formula:
any access points outside the line of sight system finds which
access point it is and calculate the new RSSI value by adding FSPL (dB) = 20log10(d) + 20log10(f) + K (1)
up the lost signal. After correcting the RSSI value of all the 3
reference access points the distance from each point to the Where,
user will be calculated. Then the system will calculate the d = distance
location of the user by implementing trilateration technique f = frequency
and the coordinated of the user will be fetched. Once all these K = constant that depends on the units used for d and
tasks are completed, user floor architecture will be loaded and f
the user position will be plotted in architecture. The system For d, f in meters and kilohertz, respectively, the
will also save the coordinate in the database for external use constant becomes -87.55.
by the rescue team at emergency scenarios. For d, f in meters and megahertz, respectively, the
constant becomes -27.55.7
For d, f in kilometres and megahertz, respectively,
the constant becomes 32.45.

9
 The user coordinate can be obtained by rearranging the
To calculate the distance, we can derive the equation from (1) equation in (4), (5) and (6) to solve x and y as shown is the
as, following formula:

Distance = 10(( . ( ∗ ( )) )/ )
(2)

Where, = (7)
(
s = signal level

Consider a frequency of 2412MHz and signal level= - = (8)
(
57dbm. The distance can be calculated as,
Where,
Distance= 10(( . ( ∗ ( )) )/ )

= + −
Distance=7.0004 meter (3)
= + − (9)
E. Position Estimation
= + −
To find the location of the user, at least 3 reference access
points are required and the system finds the distance from all And,
the three points. The system uses trilateration technique to
= ( − )
estimate the user position by calculating the distance to each
access points at know position. The algorithm requires the
= ( − ) (10)
coordinates of all the three reference access points with higher
RSSI value ( , ) and the distance to each of the access
= ( − )
points from the user is calculated to predict the location. The
system also checks for the access points that are not in the line
= ( − )
of sight and if it detects, the distance calculation will be done
by adding up the lost signal strength to that access points.
= ( − ) (11)
The Equation can be achieved using Pythagorean
theorem as shown in the following formula:
= ( − )

= ( − ) + ( − ) (4) Localizing the user is very convenient using (7) and (8)
because the coordinates of all the reference points are stored in
= ( − ) + ( − ) (5) the database and the distance (d1, d2, d3) can be obtained from
ranging.
= ( − ) + ( − ) (6)
F. Architecture Design
The architecture structure of the implementation is shown
in Fig. 5 that can be used to locate a user in emergency
scenarios. When the user requests for the location the system
will generate and plot it on user device and the coordinates
will be stored in the server for external use so that rescue
teams can access the data and find out where the user is and
which will help is rescuing peoples easily and the rescue team
does not need to search for the victims in emergency
scenarios.

Fig. 4 Trilateration Estimation

10
 In scenario 3 there are two walls in between the user and
the access points, that is two of the references access points
are not in the line of sight with the user. Same as in scenario 4
all the three reference access points are not in the line of sight
with the user.
While comparing in all the scenario except scenario 1,
TELIL generates more accurate position compared to that of
Trilateration technique.

Fig. 5 Architecture design

Fig. 7 Trilateration vs TELIL

IV. EXPERIMENTAL RESULTS
The screenshot of the system that have been implemented In Fig. 8 we have compared average error rate between
on an Android device which is Moto X Play manufactured by Trilateration and TELIL. We have considered all the four
Motorola has been shown in Fig. 6. The access point which we scenarios while calculating the average error rate. From the
used is equipped with 2.4 GHz wireless frequency and 20 graph, it is clear that the proposed method provides a better
MHz narrow channel. relation compared to that of localizing a victim just by using
Trilateration technique.

Fig. 6 Working Screenshot of Application

Fig. 8 Average Error Rate between Trilateration and TELIL
In Fig. 7 we have made a comparison between Trilateration
and TELIL with four scenarios. Scenario 1 is when there are
no walls between the user and the reference access points. V. CONCLUSIONS AND FUTURE WORK
Since there are no walls both the methods show the same In an indoor emergency scenario such as fire accidents,
result. When it comes to scenario 2 with one wall in between localizing a victim using a well-defined system can help in
the error rate produced by TELIL is less compared to that of rescuing the victim on time. Most of the methods that are
Trilateration method. currently in use have a need of more nodes to find the accurate
result and in fingerprinting method more data space is required

11
to store the node coordinates. At present, most of the buildings estimation based on receiving signal strength indicator,” in
Proceedings - 5th International Conference on Wireless
are equipped with WiFi access points which provide internet
Communications, Networking and Mobile Computing, WiCOM
access and data sharing. These access points can be utilized 2009, 2009.
for localization by installing it in a well-defined manner. We [6] P. Enge and P. Misra, “Special issue on global positioning system,”
have proposed a system based on trilateration and edge Proc. IEEE, vol. 87, no. 1, pp. 3–15, 1999.
learning which is able to estimate the position of the user with [7] R. Want, A. Hopper, V. Falcão, and J. Gibbons, “The active badge
an error rate of 1.8 meters when there are no walls within the location system,” ACM Trans. Inf. Syst., vol. 10, no. 1, pp. 91–102,
1992.
line of sight of the user and the reference access points. When
[8] R. Barber, M. Mata, M. J. L. Boada, J. M. Armingol, and M. A.
there is one wall between the line of sight the error rate Salichs, “A perception system based on laser information for mobile
increased to 2.14 meters. From the study, it is clear that when robot topologic navigation,” in IECON Proceedings (Industrial
there are obstacles the error rate keeps on increasing. By Electronics Conference), 2002, vol. 4.
implementing our system, we were able to reduce the error [9] B. B. Parodi et al., “Initialization and online-learning of RSS maps
rate caused by obstacles such as walls to an average of 3.0725 for indoor / campus localization,” in Record - IEEE PLANS,
Position Location and Navigation Symposium, 2006, vol. 2006, pp.
meters to 2.73 meters. 164–172.
Future work includes detecting other obstacles and finds [10] N. B. Priyantha, A. Chakraborty, and H. Balakrishnan, “The Cricket
the signal strength lost due to the obstacle. Image processing location-support system,” in Proceedings of the 6th annual
can be done to detect obstacles other than walls to generate a international conference on Mobile computing and networking -
more accurate result. When there are obstacles the signal gets MobiCom ’00, 2000, pp. 32–43.
scattered and this will lead to noise in the signal. Multi- [11] X. Zheng, G. Bao, R. Fu, and K. Pahlavan, “The performance of
simulated annealing algorithms for Wi-Fi localization using Google
Trilateration and triangulation method can be implemented Indoor Map C3 - IEEE Vehicular Technology Conference,” 76th
instead of trilateration technique to compare with the proposed IEEE Veh. Technol. Conf. VTC Fall 2012, 2012.
system. [12] C. Kilinc, S. Al Mahmud Mostafa, R. Ul-Islam, K. Shahzad, and K.
Currently, the system supports only one floor and 3 Andersson, “Indoor taxi-cab: Real-time indoor positioning and
location-based services with ekahau and android OS,” in
reference access points in calculating the position of the
Proceedings - 2014 8th International Conference on Innovative
victim. It can be extended to multi-storey buildings for Mobile and Internet Services in Ubiquitous Computing, IMIS 2014,
predicting the position of the victim. 2014, pp. 223–228.
[13] M. Betke and L. Gurvits, “Mobile robot localization using
landmarks,” IEEE Trans. Robot. Autom., vol. 13, no. 2, pp. 251–
ACKNOWLEDGMENT 263, 1997.
[14] R. Peng and M. L. Sichitiu, “Ad Hoc Positioning System (APS)
First and foremost, we are deeply grateful to her most using AoA,” in INFOCOM’ 03, 2003, no. November 2015.
admired Mata Amritanandamayi Devi our beloved chancellor [15] J. M. Font-Llagunes and J. A. Batlle, “Consistent triangulation for
for her motivation and guidance in incredible and unbelievable mobile robot localization using discontinuous angular
ways. Thank you, heartily, to everyone at Amrita Vishwa measurements,” Rob. Auton. Syst., vol. 57, no. 9, pp. 931–942,
Vidyapeetham Mysuru campus for providing the necessary 2009.
environment, infrastructure, and motivation for conducting our [16] A. T. Kunnath, M. V Ramesh, and V. Selvan, “Signal processing for
Wireless Geophone Network to detect landslides,” in Computer
dissertation work. Applications and Industrial Electronics (ICCAIE), 2010
International Conference on, 2010, pp. 69–73.
REFERENCES [17] D. Moore, J. Leonard, D. Rus, and S. Teller, “Robust distributed
network localization with noisy range measurements,” in
[1] P. B. and V. N. Padmanabhan, “RADAR: An in-building RF based Proceedings of the 2nd international conference on Embedded
user location and tracking system,” Proc. IEEE INFOCOM 2000. networked sensor systems - SenSys ’04, 2004, p. 50.
Conf. Comput. Commun. Ninet. Annu. Jt. Conf. IEEE Comput. [18] Y. Wang, X. Wang, D. Wang, and D. P. Agrawal, “Range-free
Commun. Soc. (Cat. No.00CH37064), vol. 2, no. c, pp. 775–784, localization using expected hop progress in wireless sensor
2000. networks,” IEEE Trans. Parallel Distrib. Syst., vol. 20, no. 10, pp.
[2] K. Pahlavan, X. Li, and J. P. Mäkelä, “Indoor geolocation science 1540–1552, 2009.
and technology,” IEEE Commun. Mag., vol. 40, no. 2, pp. 112–118, [19] A. Boukerche and E. Nakamura, “Localization systems for wireless
2002. sensor networks,” IEEE Wirel. Commun., vol. 14, no. 6, pp. 6–12,
[3] W. K. Zegeye, S. B. Amsalu, Y. Astatke, and F. Moazzami, “WiFi 2007.
RSS fingerprinting indoor localization for mobile devices,” in 2016 [20] E. Martin, G. Friedland, O. Vinyals, and R. Bajcsy, “Precise Indoor
IEEE 7th Annual Ubiquitous Computing, Electronics and Mobile Localization Using Smart Phones,” MM ’10 Proc. Int. Conf.
Communication Conference, UEMCON 2016, 2016. Multimed., vol. 7, pp. 787–790, 2010.
[4] T. Bagosi and Z. Baruch, “Indoor localization by WiFi,” in [21] R. K. Megalingam, A. P. Rajendran, D. Dileepkumar, and A. T.
Proceedings - 2011 IEEE 7th International Conference on Soloman, “LARN: Indoor navigation for elderly and physically
Intelligent Computer Communication and Processing, ICCP 2011, challenged,” in Proceedings of the 3rd IEEE Global Humanitarian
2011, pp. 449–452. Technology Conference, GHTC 2013, 2013, pp. 326–330.
[5] C. Qi, H. Gaoming, and S. Shiqiong, “WLAN user location [22] Z. Yang, C. Wu, and Y. Liu, “Locating in fingerprint space:

12
 Wireless indoor localization with little human intervention,” Proc.
Annu. Int. Conf. Mob. Comput. Netw., pp. 269–280, 2012.
[23] S. Sen, J. Lee, K.-H. Kim, and P. Congdon, “Avoiding multipath to
revive inbuilding WiFi localization,” Proceeding 11th Annu. Int.
Conf. Mob. Syst. Appl. Serv. - MobiSys ’13, p. 249, 2013.
[24] H. Liu et al., “Push the limit of WiFi based localization for
smartphones,” in Proceedings of the 18th annual international
conference on Mobile computing and networking - Mobicom ’12,
2012, p. 305.
[25] A. T. Kunnath, P. Pradeep, and M. V. Ramesh, “Locating and
monitoring emergency responder using a wearable device,” in
Proceedings of the International Conference on Advances in
Computing, Communications and Informatics - ICACCI ’12, 2012,
p. 1163.
[26] H. Liu, J. Yang, S. Sidhom, Y. Wang, Y. Chen, and F. Ye,
“Accurate WiFi based localization for smartphones using peer
assistance,” IEEE Trans. Mob. Comput., vol. 13, no. 10, pp. 2199–
2214, 2014.
[27] N. Kothari, B. Kannan, E. D. Glasgwow, and M. B. Dias, “Robust
indoor localization on a commercial smart phone,” in Procedia
Computer Science, 2012, vol. 10, pp. 1114–1120.
[28] A. Mukhopadhyay, K. S. Subeesh, and S. Mathew, “Context Aware
Information for Localization,” Int. J. Appl. Eng. Res., vol. 10, no.
55, p. 2015.
[29] K. Liu, X. Liu, and X. Li, “Guoguo: Enabling Fine-grained Indoor
Localization via Smartphone,” Proc. of MobiSys, p. 235, 2013.
[30] E.-E.-L. Lau, B.-G. Lee, S.-C. Lee, and W.-Y. Chung,
“ENHANCED RSSI-BASED HIGH ACCURACY REAL-TIME
USER LOCATION TRACKING SYSTEM FOR INDOOR AND
OUTDOOR ENVIRONMENTS,” Int. J. Smart Sens. Intell. Syst.,
vol. 1, no. 2, pp. 534–548, 2008.
[31] M. V Ramesh and K. P. Soman, “Wireless Sensor Network
Localization With Imprecise Measurements Using Only a Quadratic
Solver.,” in ICWN, 2009, pp. 109–112.
[32] M. Cao, B. D. O. Anderson, and A. S. Morse, “Localization with
imprecise distance information in sensor networks,” in Proceedings
of the 44th IEEE Conference on Decision and Control, and the
European Control Conference, CDC-ECC ’05, 2005, vol. 2005, pp.
2829–2834.

View publication stats

13