Cloud Testing for Mobile Software Systems
Concept and Prototyping
Oleksii Starov
, Sergiy Vilkomir
and Vyacheslav Kharchenko
Department of Computer Science, East Carolina University, Greenville, NC, U.S.A.
Department of Computer Systems & Networks, National Aerospace University KhAI, Kharkiv, Ukraine
Keywords: Cloud Testing, Mobile Systems, Testing Framework, TaaS, Android.
Abstract: This paper describes an approach for increasing the effectiveness of mobile software system testing. A
Cloud Testing of Mobile Systems (CTOMS) framework is presented in the form of a cloud service that
provides the ability to run tests on a variety of remote mobile devices. This framework is based on a
heterogeneous networked system that connects operational computers, mobile devices, and databases with
software applications. Our research focuses on building a concept and a prototype of CTOMS that supports
testing Android mobile applications in the cloud. CTOMS allows multidirectional testing, providing the
opportunities to test an application on different devices and/or operating system (OS) versions and new
device models for their compatibility with the newest OS versions and the most popular applications.
Another new aspect is to embed the test model, specifically the appropriate testing techniques for mobile
development, within the framework. For users, this model will provide suggestions from CTOMS about the
test methods, criteria, coverage, and possible test cases. These suggestions are based on available
configurations, statistics, and resource constraints.
Mobile developments are presented nowadays as a
variety of different applications with different
quality requirements. Interest in critical mobile
applications that require high-level reliability and
security is growing rapidly. For instance, mobile
applications have become commonplace for online
banking (Bank of America, 2013), and some
researchers are discussing the use of smartphones
and tablets at nuclear power plants (Moser, 2012). A
new trend is to use smartphones as components for
mobile cyber-physical systems because the powerful
hardware has a variety of sensors (White et al.,
2010). Examples of such systems include mobile
applications for notifications about hurricanes (Carr,
2012), monitoring cardiac patients (Leijdekkers,
2006), and traffic monitoring (Work and Bayen,
To guarantee the mobile applications’ reliability
and security, sufficient testing is required on a
variety of heterogeneous devices as well as on
different OS. Android development is the most
representative example of how different applications
should function amid a plethora of hardware-
software combinations (uTest, 2013). Adequately
testing all of these platforms is too expensive—
perhaps impossible—especially for small resource-
constrained mobile development companies.
This paper describes a framework that facilitates
the testing of mobile applications. The idea is to
create a cloud service that provides the ability to run
tests on a variety of remote mobile devices (i.e.,
smartphones). The proposed framework is based on
a heterogeneous networked system that connects
operational computers, mobile devices, and
databases with applications. This framework is
presented as a combination of hardware
(smartphones) and software (applications) that
allows for different testing directions. For instance,
it is possible to test a new smartphone model for its
compatibility with mobile applications and to test a
new application on different smartphone models.
This framework—CTOMS—serves as Testing as a
Service (TaaS) for mobile development.
This paper is organized as follows. Section 2
discusses the current state-of-the-art cloud testing
and the testing of mobile systems and applications.
Section 3 presents CTOMS, including its ability for
multidirectional testing, and the use of testing
Starov O., Vilkomir S. and Kharchenko V..
Cloud Testing for Mobile Software Systems - Concept and Prototyping.
DOI: 10.5220/0004416101240131
In Proceedings of the 8th International Joint Conference on Software Technologies (ICSOFT-EA-2013), pages 124-131
ISBN: 978-989-8565-68-6
2013 SCITEPRESS (Science and Technology Publications, Lda.)
models that allow users to choose specific testing
methods for their systems. The issues of security and
performance testing using CTOMS are also analyzed
and described. A high-level CTOMS architecture is
suggested in Section 4. Elements of prototyping and
feasibility studies of the framework are described in
Section 5. Section 6 concludes the paper and
proposes the directions for future work.
A variety of cloud testing services exist to facilitate
software testing (Inçki et al., 2012); (Vilkomir,
2012); (Tilley and Parveen, 2012), including testing
mobile applications (Priyanka et al., 2012). Several
of these services provide remote access to connected
smartphones in order to accomplish their testing
(Perfecto Mobile, 2013); (Keynote Device
Anywhere, 2013). Of these, the Perfecto Mobile
service provides a well-suited example of optimum
core functionality for our proposed CTOMS
framework. Specifically, Perfecto Mobile aids
mobile developers in using remote smartphones for
manual testing, recording of scripts, and automatic
running of tests on a range of models. The CTOMS
framework offers two key enhancements over
existing testing services: multidirectional testing
capabilities and integration of the testing model into
the functionality of the service. Apkudo’s device
analytics (Rowinski, 2012) provide some elements
of multidirectional testing by testing devices on the
top 200 apps from the market. Keynote
DeviceAnywhere Test Planner provides elements of
testing techniques by suggesting a set of devices to
use for testing. While these services do not offer the
total range of desired comprehensive functionalities
compared with CTOMS, they point to the interest in
this testing direction.
A scientific testbed for a cloud solution is
discussed in Konstantinidis et al. (2012). Usage of a
set of plug-ins for different testing methods as a part
of the cloud testing framework is considered in
Jenkins et al. (2011). Cloud testing distributed
systems is analyzed in Tilley and Parveen (2012),
Rhoton and Haukioja (2011), and Coulouris et al.
Many testing approaches were investigated
specifically for Android applications, including
automation of testing and GUI testing (Hu and
Neamtiu, 2011); (Ridene and Barbier, 2011), testing
frameworks that use distributed networked solutions
(Ridene and Barbier, 2011); (Mahmood et al., 2012),
security testing (Mahmood et al., 2012), and others.
Many investigations justify Android selection as
a base mobile platform for the CTOMS prototype.
Thus, according to Gartner, Android devices hold
the best part of the market (Haselton, 2012), and
forecasts by Forbes indicate that the Android
platform will advance to meet enterprise
requirements soon (Fidelman, 2012). Previous
research regarding bugs statistics in Android OS
(Maji, 2010) proves that Android has effectively
organized an open-sourced bug-tracking system that
can be helpful in further investigations.
3.1 Multidirectional Testing
This paper aims to generalize the concept of the
cloud service that provides mobile devices for
testing. The CTOMS framework extends the typical
functionality and provides the ability to test OS
version updates and new hardware devices against
most popular or/and important applications. It is
important to be confident that the legacy mobile
applications will still work properly in a new
Figure 1: Multidirectional testing.
Figure 1 illustrates the concept of multidirectional
testing. It shows the three main types of objects in
the system: applications (apps), devices (hardware),
and versions of OS. CTOMS provides the ability to
test each side on/against others; in other words, it
provides multidirectional testing from all possible
perspectives. All use cases are in high demand.
Simple lines show existing services. Dashed lines
show partially new use cases; see Rowinski (2012)
for an analogue of testing new devices against
applications. Bold arrows show totally new
The current study considers cloud solutions for
the following scenarios:
1. Application developers can test a product on
different devices and/or OS versions.
2. OS developers can test new versions of an OS on
a set of modern devices and the most popular
apps to ensure compatibility.
3. Hardware developers can test new device models
for their compatibility with the newest OS
versions and the most popular applications.
The innovative aspect of our approach is that it
provides testing of new OS version against the top
popular applications (and test cases used for their
development). The relevance of such functionality
becomes obvious if we consider how rapidly new
versions of iOS or Android systems are released.
The same acute situation applies to hardware, thus
the device fragmentation testing matrix for Android
development can have nearly 4,000 separate
Android device models (uTest, 2013).
In general, the need to test OS or hardware
against applications is driven by critical systems that
contain mobile applications. In such systems, it is
very important to guarantee the dependability of
crucial applications with a newer version of the OS
or a new device. Critical mobile applications must
still work properly after OS updates and provide the
same reliability, or else the resulting faults can be
very expensive. The presence of all testing
perspectives in the CTOMS framework provides the
ability to comprehensively test mobile systems
because the hardware components are also key ones
for them.
Android development is the most representative
example of the problem of diverse configurations to
be tested. The variety of heterogeneous devices, OS
versions, screen resolutions, and other parameters is
significant. The popularity of the Android operating
system necessarily makes the question of cloud
testing extremely important. As such, this research
focuses first on building a prototype of the CTOMS
that supports Android testing. The following
sections offer solutions from both the architecture
and implementation points of view. The proposed
CTOMS framework also takes into consideration the
possibility of support for different kinds of devices:
from smartphones to mobile robots and from
microcontroller-based embedded systems to field-
programmable hardware (Kharchenko et al., 2009).
3.2 The CTOMS Structure
The conceptual structure of the CTOMS framework
contains several layers of mechanisms and
functionalities. Each should be analyzed,
architected, and implemented in a final
comprehensive system. Figure 2 illustrates these
layers and their connections with possible use cases:
1. The contributor of the device only invests in
hardware by connecting it to the cloud system.
2. The application developer uploads the
application under test (source codes or binary),
specifies test cases (automated scripts or unit
tests), gets results as pass/fail statistics or
screenshots for checkpoints, manually accesses
selected devices for debugging, etc.
Figure 2: The CTOMS structure and use cases.
3. The application developer uploads the
application under test (source codes or binary),
specifies test cases (automated scripts or unit
tests), gets results as pass/fail statistics or
screenshots for checkpoints, manually accesses
selected devices for debugging, etc.
4. The device producers test new devices against
the base of apps and OS software in CTOMS.
Test scripts available in the system for chosen
apps can be used. Producers can also specify
their own test scenarios.
5. The OS producer connects the devices with new
OS versions to the cloud or uploads update
packages to the database. Then the new OS
versions are tested with apps/scripts/ devices in
the system. OS producers can also specify their
own test scenarios.
Figure 2 shows that each user can provide a “testing
model” while using the framework. For application
developers, this means specifying a test strategy,
namely, what tests need to be performed on what
devices and how. For example, they can specify
their own test scripts; select devices, OS versions,
coverage, methods to check if a test is passed, etc.
For other users, this means not only specification of
test strategy or rules of testing (in case of
contributor), but also how to perform the testing on
connected devices. For example, the settings can be
made for which or how many applications to test.
The technical interface of the system for
contributors, as well as OS and hardware producers,
is almost the same.
The innovative feature of CTOMS is its testing
model that serves as an internal mechanism and
additional service for users (layer inside cloud on the
Figure 2). This aspect is described in more details in
the next section.
The databases in CTOMS store the software
(applications and OS versions), the testing results,
the statistical information about testing, and user
information for granting privileges based on the
billing and for providing multi-tenancy, etc.
This work with devices requires the development
of three additional layers: load balancing of tests
execution, diagnostic facilities, and heterogeneous
device connections. All of these layers (inside the
cloud in Figure 2) should be investigated in
accordance with earlier specified new use cases,
along with the creation of corresponding
methodologies. Load balancing of tests execution
system means algorithms to distribute test cases
between connected smartphones in optimal ways
with respect to time of operation and wait time of
other users. Simultaneously, general scalability
should also be taken into account.
The main high-level architecture solutions for the
challenges above are described in Section 4.
3.3 Testing Model
Another new aspect proposed in this study is to
embed the test model, i.e., appropriate testing
techniques for mobile development within the cloud
framework. Specifically, pair-wise testing (Kuhn et
al., 2009) is considered for this purpose. The pair-
wise testing combinatorial approach aids in dealing
with large amounts of different combinations of
hardware and software parameters that should be
covered by the tests. Coverage evaluation is a crucial
activity within mobile testing. According to the
Android Developers website (Android Developers,
2013), there are nine families of Android OS present
in the market (not counting subversions and builds
without Google APIs), four types of screen
resolution (small, normal, large, extra), and four
levels of screen density. Other parameters, such as
the type of Internet connection (WiFi, 3G), size of
ram, vendor, and processor’s characteristics, should
also be taken into consideration n order to provide
adequate coverage during testing. Some examples of
combinatorial tests based on different configurations
of Android application can be found in Kuhn et al.
(2010). Other techniques, including t-wise testing
(Lei et al., 2007), MC/DC (Chilenski and Miller,
1994), and RC/DC (Vilkomir and Bowen, 2006)
testing criteria, are also considered for integration
with CTOMS.
From a user’s point of view, the embedding of a
testing model operates as suggestions provided by
CTOMS: namely, suggestions relating to what
hardware-software configurations need to be tested,
the testing criterion to choose, the minimal test
coverage required, the risk statistics about particular
device configurations, etc. As a result, a user can
choose an appropriate testing model for a given
situation in terms of desired budget, time,
requirements, etc.
A testing model must also provide the general
organization of testing (i.e., launching tests, storing
results, etc.). This provides an opportunity to collect
statistics in the system, and for instance, to advise a
user that a particular device caused the main part of
defects during other similar applications testing.
CTOMS can also be considered for providing
reliability, performance, and security testing. In the
context of security testing, the following variants of
additional services are proposed:
Implementation of different kinds of static
Whitebox approaches based on decompiling Java
classes (Mahmood et al., 2012).
Model-driven approaches for security and
language-based security analysis.
Automated stress security testing.
For performance testing, it is proposed to use
frame rate counters similar to Windows phone
Emulator (Windows Phone Dev Center, 2013).
For reliability testing, detailed usage of statistics
is proposed in conjunction with the long-term
performing of tests.
A detailed design of the CTOMS framework
aims to provide the ability to extend functionality
with such kinds of testing in a way similar to plug-
ins. The global goal is to create a comprehensive
testing environment.
The architecture of a networked system, such as
CTOMS, can vary significantly in its level of
complexity. For example, the size of the desired
distributed solution is one dependent factor. CTOMS
can be architected as a single PC computer with
connected smartphones, or as a comprehensive cloud
solution that operates hundreds of such PC nodes.
To achieve all of its goals, CTOMS should be
implemented from a large-scale perspective and
must correspond to all cloud computing features,
such as service delivery, scaling, virtualization,
elasticity, multi-tenancy, load balancing, universal
access, etc. The cloud type can also differ from a
public SaaS (Software as a Service) solution to a
private IaaS (Infrastructure as a Service), with the
ability to provide a low-level configuration.
In this paper, trade-off solutions for CTOMS
complexity are suggested. The solutions include
leveraging public cloud providers, such as Google
App Engine (GAE) and Amazon AWS to create a
master application (layer) of the system in PaaS
(Platform as a Service). Computers with connected
mobile devices serve as “leaf nodes” and create a
slave layer of the system. They perform actual
testing on smartphones, collect results, and interact
with a master application through the Internet.
Figure 3 illustrates the high-level architecture of
CTOMS. It shows three layers of the system
interacting with each other. The separation of the
master layer as a public PaaS application gives the
ability to easily achieve both horizontal and vertical
scaling (layer 1 in Figure 3). The master layer does
not depend on the other layers, e.g., leaf nodes.
Besides, it also can leverage the highly scalable
databases and data storage provided by PaaS. The
master application has a comprehensive web
Communication between subsystems through the
Internet (web services) provides a convenient
scaling of layer 3. Leaf computers with smartphones
can be easily connected to a master application in
the cloud. They could also operate emulators (virtual
Android devices marked with a dashed line). The
interface for working with emulators and devices is
the same.
The middle layer 2 is optional. Its goal is to
leverage embedded Hadoop facilities of chosen
PaaS, particularly, MapReduce functionality. The
Figure 3: A variant of the CTOMS architecture.
aim is to organize test distribution and results
gathering by the master application. The algorithms
for effective load balancing of tests execution still
are performed by the master application.
The current study provides a prototyping of the
CTOMS according to a specified distributed
architecture of the desired cloud system. The first
and third layers, according to Figure 3, were
implemented as the first prototype. The resulting
system consists of the two web Java applications:
the master and the leaf node.
The created master application is placed in a
Google App Engine cloud (Google App Engine,
2013), and, from a user’s point of view, looks like a
web interface to upload Android applications’
binaries and test scripts and get testing results.
The MonkeyRunner tool (MonkeyRunner, 2013)
provided with Android SDK (Android SDK, 2013)
is used to perform and automate actual testing on
mobile devices. This was the easiest choice for
acquiring a useful system that reflects all the
required characteristics of similar testing services.
Our aim was not to create a new and better way of
testing automation or manual remote access to
smartphones, but to create a cloud system to
accommodate other improvements. Thus, the current
CTOMS prototype accepts scripts for the
MonkeyRunner tool and provides as testing results a
set of screenshots-checkpoints taken on different
devices to compare by user-tester. Then the user can
specify which screenshot shows a bug.
The following associated issues were elaborated:
scripts parallelization to run on several devices
simultaneously; automation of statistics of fail/pass
checks based on heuristic images comparison; and a
possibility to embed useful supporting frameworks
as AndroidViewClient (AndroidViewClient, 2013)
and ViewServer (ViewServer, 2013).
Users can also run the client application on their
own computers to perform local testing. The
functionalities are the same (except for the absence
of integrated testing techniques). In this case, the
computer plays the role of the leaf node.
Functionality of multidirectional testing is
provided in the CTOMS prototype. It gives an
opportunity to test a connected device against
applications and corresponding tests within the
database of the system. Then a user can compare
screenshots taken on the new device (or device with
installed new OS version) to nominal screenshots
taken on trustworthy old models. This new device
can be connected to the system from a user’s site
and serve in the private (not shared) mode. At the
same time, MonkeyRunner tests do not require
uploading a binary (APK file) and can simulate
common user interactions with Android OS.
The ACTS tool by the National Institute of
Standards and Technology (NIST, 2013) is used to
integrate a combinatorial testing model in CTOMS.
A correspondent interface is provided through the
master application.
All interconnections of the master application
and leaf nodes were implemented through RESTful
web services. This gives the ability to publicly
expose them as Application Programming Interfaces
(Jacobson et al., 2011).
Integration of static code analyzers to provide
quality indicators and reveal security vulnerabilities
was considered, but because such functionality can
serve as a fully separated service, it was moved to
further versions.
The CTOMS cloud framework is presented here as a
way to increase the quality of testing mobile
applications. The main concepts, architecture, and
prototype implementation are described. Integrating
a testing model in CTOMS allows researchers to
apply different testing techniques and provide
performance and security testing of mobile
applications. The current prototype provides ability
of functional testing and detecting user interface
related bugs (e.g., layout or graphic performance
A variety of extensions to the framework are
possible on conceptual and implementation levels.
The future work will focus on case studies and
experiments using CTOMS. In particular, using
possible crowdsourcing benefits provided by
CTOMS, we are going to investigate dependency
between bugs in mobile applications and updates of
OS, and show testing effectiveness of such a cloud
of devices. Development experience and preliminary
analysis of reviews and bug reports in Google Play
indicate high clustering of defects for a specific OS
version or smartphone model. For instance, based on
the latest reviews of Twitter Android app (Twitter,
2013), we can conclude that users have problems
with notifications on different devices under
Android 4.0.1. Some statistics on causes of mobile
app crashes (including crashes by OS version),
provided by Crittercism (Crittercism, 2013), can be
found in Geron (2012).
Android Developers, 2013. Dashboards. Available at:
Android SDK, 2013. Available at: http://developer.
AndroidViewClient, 2013. Extension to MonkeyRunner.
Available at:
Bank of America, 2013. Mobile Banking. Available at:
Carr, D. F. 2012. Hurricane Sandy: Mobile, Social Tools
Help Emergency Management, Brainyardnews (Oct.
2012). Available at: http://www.informationweek.
Chilenski, J. J., Miller, S., 1994. Applicability of Modified
Condition/Decision Coverage to Software Testing,
Software Engineering Journal, Sept. 1994, 193-200.
Coulouris, G., Dollimore, J., Kindberg, T., Blair, G.,
2011. Distributed Systems: Concepts and Design,
Addison-Wesley. 5
Crittercism, 2013. Available at: https://www.
Fidelman, M., 2012. The Latest Infographics: Mobile
Business Statistics For 2012. Available at: http://
Geron, T. 2012. Do iOS Apps Crash More Than Android
Apps? A Data Dive. Available at:
Google App Engine, 2013. Developers portal. Available
Haselton, T., 2012. Android Has 56.1% Of Global OS
Market Share, Gartner Says. Available at:
Hu, C., Neamtiu, I., 2011. Automating GUI testing for
Android applications. In Proceedings of the AST '11,
6th International Workshop on Automation of
Software Test. ACM New York, NY, USA, 77-83.
Inçki, K., Ari, I., Sozer, H., 2012. A Survey of Software
Testing in the Cloud. In Proceedings of 2012 IEEE
Sixth International Conference on Software Security
and Reliability Companion. 18-23.
Jacobson, D., Brail, G., Woods, D., 2011. APIs: A
Strategy Guide. O’Reilly Media, Inc., 60-70.
Jenkins, W., Vilkomir, S., Sharma, P., Pirocanac, G.,
2011. Framework for Testing Cloud Platforms and
Infrastructures. In Proceedings of the CSC 2011,
International Conference on Cloud and Service
Computing, Hong Kong, China, Dec. 12-14, 2011,
Keynote DeviceAnywhere, 2013. The Mobile Testing
Platform. Available at: http://www.keynotedevice
Kharchenko, V., Siora, O., Sklyar, V., 2009. Design and
testing technique of FPGA-based critical systems. In
Proceedings of CADSM 2009, 10th International
Conference - The Experience of Designing and
Application of CAD Systems in Microelectronics
(Polyana-Svalyava, Ukraine, Feb. 24-28, 2009, 305-
Konstantinidis, A., Costa, C., Larkou, G., Zeinalipour-
Yazti, D., 2012. Demo: a programming cloud of
smartphones. In Proceedings of the MobiSys '12, 10th
international conference on Mobile systems,
applications, and services.. ACM New York, NY,
USA, 465-466.
Kuhn, R., Kacker, R., Lei, Y., Hunter, J., 2009.
Combinatorial Software Testing, IEEE Computer.
Volume 42, Number 8, Aug. 2009, 94-96.
Kuhn, R., Kacker, R. N., Lei, Y., 2010. Practical
Combinatorial Testing, NIST Special Publication,
October, 2010.
Lei, Y., Kacker, R., Kuhn, D. R., Okun, V., Lawrence, J.,
2007. IPOG: A General Strategy for T-Way Software
Testing. In Proceeding of ECBS '07, IEEE
Engineering of Computer Based Systems conference,
March 2007, 549-556.
Leijdekkers, Gay, V. 2006. Personal Heart Monitoring and
Rehabilitation System using Smart Phones. In
Proceedings of the International Conference on
Mobile Business. Citeseer, 29.
Mahmood, R., Esfahani, N., Kacem, T., Mirzaei, N.,
Malek, S., Stavrou, A., 2012. A whitebox approach for
automated security testing of Android applications on
the cloud. In Proceedings of AST 2012, Automation of
Software Test, 7th International Workshop.
Maji, K., 2010. Characterizing Failures in Mobile OSes: A
Case Study with Android and Symbian. In Software
Reliability Engineering (ISSRE) 21st International
MonkeyRunner, 2013, Android SDK tool.
Moser K., 2012. Improving Work Processes for Nuclear
Plants (Exelon Nuclear). In American Nuclear Society
Utility Working Conference 2012.
NIST, 2013. ACTS tool. Available at:
Perfecto Mobile, 2013. The MobileCloud Company.
Available at:
Priyanka, Chana, I., Rana, A., 2012. Empirical evaluation
of cloud-based testing techniques: a systematic review,
ACM SIGSOFT Software Engineering Notes archive.
Volume 37, Issue 3, May 2012, ACM New York, NY,
USA, 1-9.
Rhoton, J., Haukioja, R., 2011. Cloud Computing
Architected: Solution Design Handbook, Publisher:
Recursive, Limited, ISBN: 0956355617.
Ridene, Y., Barbier, F., 2011 A model-driven approach for
automating mobile applications testing. In
Proceedings of the 5th European Conference on
Software Architecture: Companion Volume Article
No. 9.
Rowinski D., 2012. Mobile Carriers and OEMs Get
Android App Testing Cloud from Apkudo, ReadWrite
(February 7, 2012). Available at: http://
Tilley, S., Parveen, T., 2012. Software Testing in the
Cloud: Perspectives on an Emerging Discipline,
Information Science Reference, Nov. 2012.
Twitter, 2013. Application for Android. Available at:
uTest, Inc., 2013. The Essential Guide to Mobile App
Testing (free eBook) Available at: http://www.
ViewServer, 2013. Available at: https://github.
Vilkomir, S., 2012. Cloud Testing: A State-of-the-Art
Review, Information & Security: An International
Journal. Volume 28, Issue 2, Number 17, 2012, 213-
Vilkomir, S., Bowen, J. P., 2006. From MC/DC to
RC/DC: formalization and analysis of control-flow
testing criteria, Formal aspects of computing, Volume
18, Number 1, 2006, 42-62.
White, J., Clarke, S., Dougherty, B., Thompson, C., and
Schmidt, D. 2010. R&D Challenges and Solutions for
Mobile Cyber-Physical Applications and Supporting
Internet Services. Springer Journal of Internet Services
and Applications. Volume 1, Number 1 (2010), 45-56.
Windows Phone Dev Center, 2013. Testing Apps for
Windows Phone. Available at:
Work, D. B., Bayen, A. M., 2008. Impacts of the Mobile
Internet on Transportation Cyberphysical Systems:
Traffic Monitoring using Smartphones. In
Proceedings of National Workshop for Research on
High-Confidence Transportation Cyber-Physical
Systems: Automotive, Aviation and Rail, Washington,
DC, Nov. 18-20, 2008.