Content uploaded by Christine Bauer
Author content
All content in this area was uploaded by Christine Bauer
Content may be subject to copyright.
On the (In-)Accuracy of GPS Measures of Smartphones:
A Study of Running Tracking Applications
Christine Bauer
Vienna University of Economics and
Business
Welthandelsplatz 1, D2
1020 Vienna, Austria
+43-1-31336-4420
chris.bauer@wu.ac.at
ABSTRACT
Sports tracking applications are increasingly available on the
market, and research has recently picked up this topic. Tracking a
user’s running track and providing feedback on the performance
are among the key features of such applications. However, little
attention has been paid to the accuracy of the applications’
localization measurements. In evaluating the nine currently most
popular running applications, we found tremendous differences in
the GPS measurements. Besides this finding, our study contributes
to the scientific knowledge base by qualifying the findings of
previous studies concerning accuracy with smartphones’ GPS
components.
Categories and Subject Descriptors
H.3.4 Systems and Software: Performance evaluation (efficiency
and effectiveness)
General Terms
Performance
Keywords
Location-based System, GPS, Global Positioning System,
Accuracy, Smartphone, Sports Tracking, Running Tracking,
Localization, Positioning
1. INTRODUCTION
Globally, the number of smartphone users topped one billion at
the end of 2012 and is estimated to double in less than three years
[18]. Spurred by Apple’s and Google’s platform concept, the
development of mobile applications is booming; about 1.7 million
total applications had been created by the end of 2012 [2]. The use
of smartphones is far beyond its classic domain of application, as
applications for all purposes have turned the smartphone into a
multi-functional device that pervades everyday life [12]. There are
applications for every flavor: games, news, guitar tuners, wine
guides, maps, messengers, travel booking, etc. This trend is also
visible in the sports domain, in particular for running – a topic that
is also increasingly getting attention in research [e.g., 1, 4, 15, 17].
Among other features, most running applications include tracking
a user’s running track and providing feedback on the performance
with respect to the distance run and altitude differences. Typically
these figures are visualized with route mapping and diagrams
(Figure 1). The basis for this feature is rooted in location
technologies, which can be considered standard components in
today’s smartphones: Cell Identifier (Cell ID) based positioning,
localization via Wireless Local Area Network (WLAN), and
localization via Global Positioning System (GPS) [13, 16, 26]. For
sports tracking – in particular, when in competition with other
athletes or for benchmarking, a feature that most running
applications offer – the expectation of users is to receive accurate
tracking information [7]. As a result, WLAN or Cell ID based
tracking are not sufficient for these applications’ purposes because
their positioning results are not fine-granular. Due to its precision,
and worldwide availability [13] without the need for additional
infrastructure [7], GPS is considered the most suitable candidate
for sports tracking applications [7].
29 Comparison of the TOP-10 jogging-applications with GPS-tracking on Android
Accuracy-Test
Sports Tracker took 13 seconds to get a signal, but at least it informs you visually if it has
found a signal already or not. The accuracy yielded mixed results. Total distance of the run
was at 0.99 km, sharing rank 2 with Noom Cardio Trainer & Orux Maps. Unfortunately, total
ascent was 31m, while total descent amounted to 32m. The track on the map would have
been accurate in general, had the start not been off by about 10 meters.
Figure 13 Sports Tracker Web-Interface & map
+ -
Fancy design with a lot of data Fancy design that is very hardware-demanding
Distance accurate Elevation inaccurate
Limited features
Table 7 Sports Tracker Pro/Contra-Summary
Figure 1. Visualized Running Track (left) and Altitude
Differences (right).
However, current smartphones contain very basic GPS receivers
[26]. As a result, the deviations may be larger compared to high-
quality GPS receivers [6]. As smartphones are equipped with
different chip sets, accuracy levels also depend on the respective
devices [25]. Research on measurement accuracy has been
performed to analyze differences between smartphone devices
[14, 25], between platforms [7], between localization technologies
[26], or between GPS based tracking with a smartphone and a
high-quality GPS receiver [16].
In an explorative study analyzing the features of running
applications, we found that identical hardware and platform setup,
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that
copies bear this notice and the full citation on the first page. To copy
otherwise, or republish, to post on servers or to redistribute to lists,
requires prior specific permission and/or a fee.
MoMM2013, 2-4 December, 2013, Vienna, Austria.
Copyright 2013 ACM 978-1-4503-2106-8/13/12 …$15.00.
MoMM2013 Papers Proceedings of MoMM2013
335
as well as positioning method (using GPS that is embedded into
the smartphone), delivered different tracking results for different
GPS based applications. Having delved into detail on this finding,
this paper presents the results on the measurement accuracy of the
nine currently most popular running applications.
The next section provides background information on location
technologies embedded in smartphones and discusses related
work, comparing different positioning modalities with respect to
method and hardware. Section 3 outlines the procedures of our
comparison study. In Section 4, we present the results concerning
deviations in lengths of distance and altitude differences as
reported by the analyzed applications. Section 5 discusses our
findings and concludes with an outlook on future research.
2. BACKGROUND AND RELATED WORK
2.1 Location Technologies in Smartphones
Location based services (LBS) are services that are accessible
through mobile networks and use the geographical position of a
mobile device [24]. The initial attempts in the field ‘context-aware
computing’ produced a plethora of LBS services [21], mainly due
to the fact that devices, at that time, had limited physical-sensing
capabilities and were largely confined to a device’s location. With
the miniaturization of hardware components and proliferation of
smartphones, LBS have become ubiquitously available to
consumers [7] and emerged as an important component of mobile
commerce [19].
Considering the vast availability of LBS applications, we observe
different degrees of granularity and accuracy, and different
requirements on how frequently to refresh location information to
serve the applications’ purposes. While for some applications a
rough estimate may be sufficient, others require more accurate
information [25]. For instance, using the application Foursquare,
people can communicate their current location and their location
history in terms of ‘business premises’, ‘restaurants’, ‘bars’, etc.
The granularity of data required is rather low, as is the needed
frequency for updating localization information. In contrast, other
applications, such as those for car navigation, require high
accuracy and frequent location updates to ensure reliable routing.
In events such as sports competitions, very precise and accurate
positioning information is necessary, which also requires frequent
location updates [7]. To illustrate, Dobson [5] provides a non-
exhaustive list of 18 different ways to gather location information
with different grades of accuracy, depending on the need and
purpose of localization.
Three main location technologies are integrated as standard
components in today’s smartphones: Cell ID, WLAN, and GPS
[13]. Also note that other sensors can be creatively deployed to
sense location, going beyond their primary sensing purposes – for
instance, using a smartphone’s microphone for indoor positioning
[e.g., 20].
Cell ID based localization works as long as there is mobile
network coverage that the smartphone can connect to. The
position of the device is derived from the coordinates of the
serving base station [26]. It is accurate within hundreds of meters
of deviation only. Cell ID information may be combined with a
rough estimation of the round trip time between the device and
base station (so-called ‘timing advance value’), from which the
range between them can be derived. With this combination,
position fixes of higher accuracies can be achieved [13].
The localization with WLAN is based on measuring the proximity
of a mobile device to wireless access points via the intensity of the
received signal (received signal strength; RSS). RSS patterns that
are received at several access points are compared with a table of
predetermined RSS patterns collected at various positions; the
location in the comparison that fits best is adopted as the device’s
position [13, 26]. Based on this mechanism, WLAN based
positioning is accurate within 30 to 50 meters deviation and works
indoors and outdoors [26]. However, as this technology requires
the availability of a wireless hotspot to register in, it is limited to
densely populated areas [13, 26].
The GPS is a satellite based navigation system developed by the
U.S. Department of Defense for military purposes [6]. It allows
locating mobile devices at any place, any time on Earth using
trilateration with range measurements between an observer and a
few visible satellites [6]. At least four independent measurements
(satellites) are required for computing a fixed position [10]. While
the military GPS version is able to provide more accurate
positioning data, its public version is limited to an accuracy of up
to 5 to 10 meters [8, 13]. This accuracy level can also be achieved
with low-cost GPS receivers, such as the ones integrated in
smartphones, when locked onto the minimum number of satellites
[10]. The functionality of GPS is, though, limited to outdoor
positioning as line of sight to the satellites is required [13].
Although the prevalence of GPS receivers has dramatically driven
down cost, size, and power requirements [6], GPS sensing is,
compared to other location technologies, rather resource intensive
[3, 9].
2.2 Studies on Smartphone Tracking
Measurement accuracy using GPS that is embedded in
smartphones has been the subject of many research endeavors.
Menard, Miller, Nowak and Norris [14] analyzed capacity
differences in GPS based positioning in smartphones (using three
different smartphones: Samsung Galaxy S, Motorola Droid X, and
iPhone 4) compared to a vehicle tracker. They concluded that the
three smartphones were an acceptable alternative to other tracking
devices in vehicles. They showed that each smartphone was
accurate within 10 meters about 95 percent of the time.
The differences in positioning accuracy among different Apple
devices (iPhone, iPod Touch, and iPad) have been researched by
von Watzdorf and Michahelles [25]. They found significant
differences in accuracy. However, the different devices under
study used different location technologies, too (WLAN versus a
combination of WLAN, GPS, and Cell ID). In contrast,
Zandbergen [26] evaluated the differences between GPS, WLAN,
and Cell ID based positioning by using only one device
(iPhone 3G) and one application. He found that the WLAN
method has potential for indoor positioning, but outdoors it lags
behind compared to GPS based localization.
Using a self-developed LBS application, Hess, Farahani,
Tschirschnitz and von Reischach [7] compared different
smartphones with different operating systems (Android 2.3.3,
Android 2.3.6, iOS 4.2.1, iOS 4.3.5, and Windows Phone 7) with
different GPS chipsets. They concluded that GPS measurement
accuracy heavily depends on the respective smartphone.
Kost and Brčić [11] analyzed GPS positioning errors based on
weather conditions on two different smartphone devices for one
fixed position, concluding that the research community should
consider that positioning accuracy depends on various factors,
such as space weather related error components.
Mok, Retscher and Wen [16] investigated whether a combination
of different sensors that are embedded into smartphones could
assist GPS positioning to increase the accuracy level. They found
Proceedings of MoMM2013 MoMM2013 Papers
336
that, assisted by an accelerometer and digital compass, GPS
positioning accuracy could be improved.
To the best of our knowledge, though, different tracking
applications run on a single device (and having the same setup for
each application) has not yet been researched.
3. RESEARCH DESIGN
In this work, we analyze nine currently popular running
applications that use GPS based localization in real time while
moving (running). The objective is to compare them with respect
to the accuracy of localization measurements.
3.1 Sample
We chose Android for our study for being the platform with the
highest market share (approximately 80 percent in 2013 [23]). In
order to reflect the highest quality and most popular applications
available at the time of our study, we considered those
applications with the highest download rates and user ratings in
the Google Play Store.
As the Google Play Store does not offer a specific running or
sports category, we had to manually search for applicable
applications in the ‘health and fitness’ category and compare their
download rates and ratings.
We chose the top nine free applications (minimum of one million
downloads by 31-May-2013) as a representative sample for this
evaluation (Table 1). The user rating of all applications in the
sample was high (on average 4.3 stars and above). We had to
exclude the application ‘Nike+ Running’, which was among the
top ten, because it was not compatible with our available testing
device (HTC Desire Bravo).
Finally, we made sure from their descriptions that the applications
relied solely on GPS for tracking the user (i.e., they do not
combine it with further localization technology).
Table 1. Sample description
Application
Downloads
in millions
User rating
Last
actualization
Endomondo
5-10
4.5 (109081)
21-May-2013
Runtastic
5-10
4.6 (76234)
26-Apr-2013
Noom Cardio Trainer
5-10
4.4 (53699)
11-Jan-2012
MyTracks
5-10
4.4 (75482)
17-Apr-2013
Runkeeper
1-5
4.5 (57992)
23-May-2013
Sports Tracker
1-5
4.6 (48275)
16-May-2013
MapMyRun GPS
Running
1-5
4.5 (33468)
10-May-2013
Adidas miCoach
1-5
4.4 (16583)
10-May-2013
Orux Maps
1-5
4.6 (9808)
21-Apr-2013
3.2 Testing for Accuracy
All applications were tested with the smartphone model ‘HTC
Desire Bravo’, following the same procedure for each application.
For applications that allowed using other means for localization
(e.g., Sports Tracker), the respective technology was disabled to
ensure that only GPS data is considered.
First, a distance of exactly 500 meters was measured in a highly
populated (city) location, so that running back and forth along this
track in a straight line would result in a total distance of exactly
one kilometer. As the starting and ending points were the same,
the altitude difference between them was ensured to be zero.
Second, a test person ran the measured track back and forth in a
straight line, with each of the applications in the sample. Before
the start of every run, the GPS signal was ensured to be good
enough for adequate measurement (which is a feature of most
running applications). After a run with an application, the
application itself and the Web interface that extended the
application (if available) were checked for the total distance of the
run (the result of which should have been one kilometer) and any
altitude differences (which should have amounted to zero).
Additionally, we analyzed the visualization of the tracked routes
in the application, as these gave good indications about the
accuracy of the tracking measurements.
4. RESULTS
4.1 Distance
Table 2 shows the measurement data of all applications for
distance in alphabetical order. Figure 2 visualizes the distance
inaccuracies, sorted according to deviation.
Table 2. Accuracy measurements for distance
Application
Distance in
meters
Deviation
in meters
Rank
Adidas miCoach
1000
0
1
Endomondo
940
60
8
MapMyRun GPS Running
1030
30
6
MyTracks
1030
30
6
Noom Cardio Trainer
1010
10
2
Orux Maps
1010
10
2
Runkeeper
980
20
5
Runtastic
940
60
8
Sports Tracker
990
10
2
0
10 10 10
20
30 30
60 60
0
10
20
30
40
50
60
70
Adidas miCoach
Noom Cardio Trainer
Sports Tracker
Orux Maps
Runkeeper
MyTracks
MapMyRun GPS Running
Endomondo
Runtastic
meters
Distance inaccuracies in meters
Figure 2. Distance inaccuracies.
Only one application (Adidas miCoach) measured a total distance
of one kilometer. Noom Cardio Trainer, Orux Maps, and Sports
Tracker showed deviations of 10 meters each. Runkeeper was 20
meters off. MyTracks and MapMyRun GPS Running were off by
30 meters. In comparison, Endomondo and Runtastic showed a
MoMM2013 Papers Proceedings of MoMM2013
337
deviation in terms of distance that was more than double those of
the other applications (60 meters). As is illustrated in Figure 3,
Endomondo’s measurement was far from a straight-line running
track.
19 Comparison of the TOP-10 jogging-applications with GPS-tracking on Android
Accuracy-Test
The testing experience of Endomondo yielded mixed results. Once the application was
started, ittook9secondstogeta“GPSexcellent”sign.Inthose9seconds,theapplication
did not give any feedback on how the signal-search was progressing. Once the signal was
found, the start-button was hit and the running began. The tracked route showed a starting-
point that differentiated 13 meters from the true starting point, although it took about 60
meters until the app actually showed any data on-screen, which would explain why the total
distance measured did come out at 940 meters. The elevation-measurements were very
inaccurate though. According to Endomondo, total ascent amounted to 10 meters, while
total descent amounted to 23 meters, with a minimum altitude of 196m and a maximum
altitude of 219 meters. Needless to say, these measurements are far from even being
realistic, with the starting point being on the same spot as the end point of the run and a
difference in ascent and descent of 13 meters. The highly inaccurate measurement stats are
shown in Table 2, along with other positive and negative standout-factors. Another nasty
byproduct was the fact that the track of the run was not very accurate on the map. The track
showed a zigzag-course and that the street was crossed multiple times, when in fact it was
run right beside the street the whole time.
Figure 7 Endomondo-track shown on Google Maps
+ -
High range of features Elevation & distance inaccurate
Different types of sports Many features are pro-only or require subscription
Broad network in Europe Track on the map inaccurate
Table 2 Endomondo Pro/Contra-Summary
Figure 3. Endomondo’s tracking data shown on Google Maps.
4.2 Altitude Differences
The measurements for altitude differences had to be unified in
order to be comparable. Some applications divided into ascent and
descent measurements; accordingly these figures were added up in
order to get the total deviation. Other applications only provided
figures for total elevation gained (ascent measurements), which
meant negative altitude differences were not shown in the
applications. Assuming that ascent and descent measurements
must be equal when a track is run back and forth, we compensated
the missing data for descent with the equivalent for ascent. (We
are aware that ascent and descent measurements may be prone to
different measurement problems; since some of the applications
did not report negative altitude differences, we had to act on
assumptions.)
Table 3 shows the measurement data of all applications for
altitude differences in alphabetical order; estimates are given in
italics. Figure 4 visualizes the ascent and descent data, sorted by
total altitude deviation.
Runtastic and Noom Cardio Trainer tracked the elevation
correctly. Runkeeper, Adidas miCoach, MapMyRun GPS
Running, and MyTracks were slightly inaccurate, with deviations
ranging from 8 meters to 14.58 meters. Orux (27 meters) and
Endomondo (33 meters) delivered measurement inaccuracies that
were more than double those of other applications measured. And
worst of all, Sports Tracker doubled those inaccuracy
measurements, reporting 63 meters of deviation.
Table 3. Accuracy measurements for altitude differences
Application
Total
ascent in
meters
Total
descent in
meters
Total
deviation
in meters
Rank
Adidas miCoach
5
5
10
4
Endomondo
10
23
33
8
Noom Cardio
Trainer
0
0
0
1
MapMyRun GPS
Running
6
6
12
5
MyTracks
7.29
7.29
14.58
6
Orux Maps
13
14
27
7
Runkeeper
4
4
8
3
Runtastic
0
0
0
1
Sports Tracker
31
32
63
9
Estimates are given in italics.
0 0 4 5 6 7.29 13 10
31
0 0
4 5 6 7.29
14 23
32
0
10
20
30
40
50
60
70
Runtastic
Noom Cardio Trainer
Runkeeper
Adidas miCoach
MapMyRun GPS Running
MyTracks
Orux Maps
Endomondo
Sports Tracker
meters
Elevation inaccuracies in meters
Total ascent in meters Total descent in meters
Figure 4. Elevation inaccuracies.
4.3 Total Deviation
Summing up the distance and altitude inaccuracies for each
application, Figure 5 represents a clear picture of the potential for
improving tracking accuracy.
The deviation levels of Noom Cardio Trainer and Adidas
miCoach show only slight deviations (10 meters each), while
Endomondo presents the highest total deviation measurement with
93 meters.
Proceedings of MoMM2013 MoMM2013 Papers
338
10 10
28
37 42 44.58
60
73
93
0
10
20
30
40
50
60
70
80
90
100
Adidas miCoach
Noom Cardio Trainer
Runkeeper
Orux Maps
MapMyRun GPS Running
MyTracks
Runtastic
Sports Tracker
Endomondo
meters
Total deviation in meters
Figure 5. Total deviation in meters.
5. DISCUSSION AND CONCLUSION
In this paper we presented a study that compares the GPS based
measurements from nine different running applications on a single
smartphone, measured on the same track of one kilometer in an
urban area. The deviations between the distances and altitude
differences measured by the applications were tremendously high.
While the different accuracy levels firstly indicate a quality
ranking of the analyzed applications, this study additionally
contributes to the scientific knowledge base by qualifying the
findings of previous studies in the field.
Our study results show that GPS delivers different results on the
same device. Accordingly, the findings of Menard, Miller, Nowak
and Norris [14] have to be handled with care. The differences that
they found in their study cannot be directly attributed to the
different GPS components used in the analyzed devices.
Moreover, the differences in positioning accuracy found by von
Watzdorf and Michahelles [25] may basically be due to the
different location technologies used for each device setting
(WLAN versus a combination of WLAN, GPS, and Cell ID). This
implies that their setup varied by the involved sensing hardware
and employed technology, causing biased results because their
findings on accuracy can neither be directly attributed to a specific
technology nor to a specific hardware setup.
Zandbergen [26] used a single device research design to analyze
differences between GPS, WLAN, and Cell ID based positioning.
In our study, however, GPS delivered different results for each
analyzed application. Accordingly, the ‘ranking’ of these
technologies’ accuracy levels has to be handled with care.
Hess, Farahani, Tschirschnitz and von Reischach [7] conclude in
their study that GPS measurement accuracy depends on the
respective smartphone. Similar to the study by von Watzdorf and
Michahelles [25], the study setup – mixing hardware (GPS
chipsets) and platforms – does not allow for conclusions
concerning the devices, as the operating system may also have
influenced the results. In addition, we show in our study that GPS
measurements on a given device may vary with the application.
Although we show that previous work has limitations, we
appreciate these studies’ contributions to the knowledge base, as
they provided first indications in the field and triggered further
research with respect to combining location technologies to
deliver more accurate results [e.g., 16, 22].
Our work also faces limitations. Firstly, we did not control for
crowdedness and traffic when tracking the locations. Also, the
phones’ internal activity (lowering read out frequency) as well as
temporary surrounding influences, such as the reflection of signals
disturbing GPS reception, cannot be excluded as influencing
factors. Differences with respect to these issues may have
influenced the GPS measurement results, causing more deviations
for some applications and less for others. Future work should
control for this; it may, for instance, be addressed by running the
track several times with each application. Alternatively, one
runner could wear 9 phones of the same type, each running one of
the apps. Secondly, we did not control for space weather
influence, such as Kos and Brčić [11] did rigidly. Thirdly, we
tracked a rather short distance of one kilometer. It is still unclear
how measurements develop over long distances. If, for instance,
Endomondo would keep its deviations per kilometer, a marathon
(42.195 kilometers) would result in a deviation of 2531.7 meters.
For a runner that maintains a pace of 5 minutes per kilometer, that
would distort his/her performance by 12.6585 minutes. For casual
runners and short distance tracks, that might not be an important
issue. On the other hand, for runners preparing for a marathon,
such a deviation is not acceptable as it can cause severe health
problems. As a result, longer distances should be evaluated in
future investigations.
In conclusion, this paper provided an overview of the currently
most popular running applications for smartphones and evaluated
them for measurement accuracies with GPS. We identified
deviating results for each application, which implies influencing
factors on GPS accuracy apart from GPS component and location
of the track, which were held constant for the study. As previous
studies did not emphasize this issue, we suggest that future work
should pay more attention to these influencing factors.
6. ACKNOWLEDGEMENTS
We thank Dominik Günsberg for his contribution of running the
track with each application of the sample.
7. REFERENCES
[1] Ahtinen, A., Isomursu, M., Huhtala, Y., Kaasinen, J.,
Salminen, J. and Häkkilä, J. 2008. Tracking Outdoor Sports:
User Experience Perspective. In E. Aarts, J. Crowley, B.
Ruyter, H. Gerhäuser, A. Pflaum, J. Schmidt and R. Wichert,
Ambient Intelligence. Springer Berlin Heidelberg.
http://dx.doi.org/10.1007/978-3-540-89617-3_13.
[2] Austin, S. 2013. The surprising numbers behind Apps. Digits:
Tech News & Analysis from the WSJ. The Wall Street
Journal, http://blogs.wsj.com/digits/2013/03/11/the-
surprising-numbers-behind-apps/.
[3] Bareth, U. and Kupper, A. 2011. Energy-efficient position
tracking in proactive location-based services for smartphone
environments. In Proceedings of the 2011 IEEE 35th Annual
Computer Software and Applications Conference
(COMPSAC 2011) (Munich, Germany, 18-22 July, 2011).
IEEE, 516-521.
http://dx.doi.org/10.1109/COMPSAC.2011.72.
[4] Bauer, C. and Waldner, F. 2013. Reactive Music: When User
Behavior affects Sounds in Real-Time. In Proceedings of the
MoMM2013 Papers Proceedings of MoMM2013
339
CHI 2013 Extended Abstracts Extended Abstracts on Human
Factors in Computing Systems (Paris, France, 27 April - 2
May, 2013). ACM, 739-744.
http://dx.doi.org/10.1145/2468356.2468488.
[5] Dobson, S. 2005. Leveraging the subtleties of location. In
Proceedings of the sOc-EUSAI 2005 (Grenoble, 12-14
October, 2005). ACM, 189-193.
http://dx.doi.org/10.1145/1107548.1107597.
[6] Hedgecock, W., Maroti, M., Sallai, J., Volgyesi, P. and
Ledeczi, A. 2013. High-Accuracy Differential Tracking of
Low-Cost GPS Receivers. In Proceedings of the MoniSys
2013 (Taipei, Taiwan, 25-28 June, 2013). ACM.
[7] Hess, B., Farahani, A. Z., Tschirschnitz, F. and von Reischach,
F. 2012. Evaluation of fine-granular GPS tracking on
smartphones. In Proceedings of the 1st ACM SIGSPATIAL
International Workshop on Mobile Geographic Information
Systems (MobiGIS 2012) (Redono Bench, CA, 6 November,
2012). ACM, 33-40.
http://dx.doi.org/10.1145/2442810.2442817.
[8] Hofmann-Wellenhof, B., Lichtenegger, H. and Collins, J.
1993. Global positioning system: theory and practice.
Springer, Vienna, Austria; New York, NY.
[9] Jurdak, R., Corke, P., Dharman, D. and Salagnac, G. 2010.
Adaptive GPS duty cycling and radio ranging for energy-
efficient localization. In Proceedings of the 8th ACM
Conference on Embedded Networked Sensor Systems
(SenSys’10) (Zürich, Switzerland, 3-5 November, 2010).
ACM, 57-70. http://dx.doi.org/10.1145/1869983.1869990.
[10] Kaplan, E. D. and Hegarty, C. 2005. Understanding GPS:
Principles and Applications. Artech House, City.
[11] Kos, S. and Brčić, D. 2013. Smartphone application GPS
performance during various space weather conditions: a
preliminary study. In Proceedings of the 21st International
Symposium on Electronics in Transport (ISEP 2013)
(Ljubljana, Slovenia, 25-26 March, 2013).
[12] Kranz, M., Möller, A., Hammerla, N., Diewald, S., Plötz, T.,
Olivier, P. and Roalter, L. 2013. The mobile fitness coach:
Towards individualized skill assessment using personalized
mobile devices. Pervasive and Mobile Computing 9, 2 (April
2013), 203-215.
http://dx.doi.org/10.1016/J.Pmcj.2012.06.002.
[13] Küpper, A. 2005. Location-based services: fundamentals and
operation. John Wiley, West Suxxes, UK.
[14] Menard, T., Miller, J., Nowak, M. and Norris, D. 2011.
Comparing the GPS capabilities of the Samsung Galaxy S,
Motorola Droid X, and the Apple iPhone for vehicle tracking
using FreeSim_Mobile. In Proceedings of the 14th
International IEEE Conference on Intelligent Transportation
Systems (ITSC), 2011 (Washington, DC, 5-7 October, 2011).
IEEE, 985-990.
http://dx.doi.org/10.1109/ITSC.2011.6083141.
[15] Mladenov, M. and Mock, M. 2009. A step counter service for
Java-enabled devices using a built-in accelerometer. In
Proceedings of the 1st International Workshop on Context-
Aware Middleware and Services: affiliated with the 4th
International Conference on Communication System
Software and Middleware (COMSWARE 2009) (Dublin,
Ireland, 2009). ACM, 1-5.
http://dx.doi.org/10.1145/1554233.1554235.
[16] Mok, E., Retscher, G. and Wen, C. 2012. Initial test on the
use of GPS and sensor data of modern smartphones for
vehicle tracking in dense high rise environments. In
Proceedings of the Ubiquitous Positioning, Indoor
Navigation, and Location Based Service (UPINLBS 2012)
(Helsinki, Finland, 3-4 October, 2012). IEEE, 1-7.
http://dx.doi.org/10.1109/UPINLBS.2012.6409789.
[17] Neville, J., Wixted, A., Rowlands, D. and James, D. 2010.
Accelerometers: An underutilized resource in sports
monitoring. In Proceedings of the 6th International
Conference on Intelligent Sensors, Sensor Networks and
Information Processing (ISSNIP 2010) (Brisbane, Australia,
7-10 December, 2010). IEEE, 287-290.
http://dx.doi.org/10.1109/ISSNIP.2010.5706766.
[18] Ramanathan, V. M. 2012. Worldwide Smartphone Users
Cross 1 Billion Mark: Report. ibitimes.com - Tech/Sci.
ibitimes.com, http://www.ibtimes.com/worldwide-
smartphone-users-cross-1-billion-mark-report-847769.
[19] Rao, B. and Minakakis, L. 2003. Evolution of mobile
location-based services. Communications of the ACM 46, 12
(December 2003), 61-65.
http://dx.doi.org/10.1145/953460.953490.
[20] Rossi, M., Seiter, J., Amft, O., Buchmeier, S. and Tröster, G.
2013. RoomSense: an indoor positioning system for
smartphones using active sound probing. In Proceedings of
the 4th Augmented Human International Conference
(Stuttgart, Germany, 7-8 March, 2013). ACM, 89-95.
http://dx.doi.org/10.1145/2459236.2459252.
[21] Schmidt, A., Beigl, M. and Gellersen, H.-W. 1999. There is
more to context than location. Computers & Graphics 23, 6
(1999), 893-902. http://dx.doi.org/10.1016/S0097-
8493(99)00120-X.
[22] Soria Morillo, L. M., Ortega Ramírez, J. A., Alvarez García,
J. A. and Gonzalez-Abril, L. 2012. Outdoor exit detection
using combined techniques to increase GPS efficiency.
Expert Systems with Applications 39, 15 (2012), 12260-
12267. http://dx.doi.org/j.eswa.2012.04.047.
[23] Tsukayama, H. 2013. Apple’s iOS loses marketshare to
Android, Windows Phone. The Washington Post.
Washington, DC,
http://www.washingtonpost.com/business/technology/apples-
ios-loses-marketshare-to-android-windows-
phone/2013/08/07/37b5a5ae-ff6c-11e2-96a8-
d3b921c0924a_story.html.
[24] Virrantaus, K., Markkula, J., Garmash, A., Terziyan, V.,
Veijalainen, J., Katanosov, A. and Tirri, H. 2001. Developing
GIS-supported location-based services. In Proceedings of the
2nd International Conference on Web Information Systems
Engineering (Kyoto, Japan, 3-6 December, 2001). IEEE, 66-
75. http://dx.doi.org/10.1109/WISE.2001.996708.
[25] von Watzdorf, S. and Michahelles, F. 2010. Accuracy of
positioning data on smartphones. In Proceedings of the 3rd
International Workshop on Location and the Web (SDoW
2010) (Shanghai, China, 8 November, 2010). ACM, 2.
http://dx.doi.org/10.1145/1899662.1899664.
[26] Zandbergen, P. A. 2009. Accuracy of iPhone locations: A
comparison of assisted GPS, WiFi and cellular positioning.
Transactions in GIS 13, s1 (June 2009), 5-25.
http://dx.doi.org/10.1111/J.1467-9671.2009.01152.x.
Proceedings of MoMM2013 MoMM2013 Papers
340































