Android security

Original article | doi: 10.25007/ajnu.v6n3a97
Academic Journal of Nawroz University (AJNU) 135
Android Security: A Review
Omar M. Ahmed 1 and Amira B. Sallow2
1Department of Computer Science, Faculty of Science, Zakho University, Duhok, Kurdistan Region – Iraq
2Department of Computer Science and Information Technology, College of Computer Science & Information
Technology, Nawroz University, Duhok, Iraq
ABSTRACT
Smartphones are used by billions of people that means the applications of the smartphone is increasing, it is out of
control for applications marketplaces to completely validate if an application is malicious or legitimate. Therefore, it
is up to users to choose for themselves whether an application is safe to use or not. It is important to say that there
are differences between mobile devices and PC machines in resource management mechanism, the security solutions
for computer malware are not compatible with mobile devices. Consequently, the anti-malware organizations and
academic researchers have produced and proposed many security methods and mechanisms in order to recognize
and classify the security threat of the Android operating system. By means of the proposed methods are different
from one to another, they can be arranged into various classifications. In this review paper, the present Android
security threats is discussed and present security proposed solutions and attempt to classify the proposed solutions
and evaluate them.
KEYWORDS: Android, Security, Smartphone.
I. INTRODUCTION
Android is a platform for mobile that is designed and
developed to be totally open source. An advanced level
of software and hardware can be used by Android
applications, in addition to local and server data and
information, uncovered by the operating system in order
to gets value and innovation to consumers. On the way
to guarantee the security of user’s application,
information, and data Android platform should be
having a powerful security mechanism. In order to
provide security for Open source platform, it requires a
powerful and severe security architecture. With
multilayered security, the architecture of Android
platform is designed that gives the flexibility required for
an open source platform (Android, 2017).
Android applications are increasing very fast, the
informal application developers, and the open source
platform encourages the malicious software developers
to Exploit of such application and vulnerable platform
and steal the user’s private data to unintentionally
open source platform, it permits or authorise the using of
third-party market applications. Malicious software can
damage the developer reputation and the apps markets.
Furthermore, because the Android platform is
considered one of the access control of the device,
consume excessive battery, steal private data, and even
transform the smartphone from normal into a botnet
zombie (Appbrain, 2017).
There are huge diversity of Android vulnerabilities are
existing which they could appear in many layers of
Android operating system’s security, like framework
layer or application layer or even Linux Kernel layer.
Vulnerabilities as well occurs in benign or any
applications by the unintended include of design flaws or
coding faults. As mentioned previously, the flawed
Android operating system gives the opportunity for
hackers or attackers to use it. There is many of security
threats exists on Android phones, like Denial of Service
attacks, colluding, repackaging apps to inject malicious
code, permission escalation, and unauthorized access
between the application.
The objective of this paper is for expanding the coverage
of malicious application growth, and Android security
threats. In this review paper, the main proposed works in
Android operating system privacy and security will be
covered, and furthermost of the existing deployed tools
and techniques.
1. Android Platform Security Architecture
Android is developed by Google at first and then via Open
Handset Alliance (OHA) is promoted. The Android
platform is located on highest of the Linux kernel which is
Academic Journal of Nawroz University (AJNU)
Volume 6, No 3(2017), 6 pages
Received 1 May 2017; Accepted 29 August 2017
Regular research paper: Published 30 August 2017
Corresponding author’s e-mail: omar.m.a@gmail.com
Copyright ©2017 Omar M. Ahmed and Amira B. Sallow
This is an open access article distributed under the Creative
Commons Attribution License.

136 Academic Journal of Nawroz University (AJNU)
consists of the APIs, libraries, and middleware by writing
in c programming language, and above the application
framework the application software is running which
contains other libraries which they are Java-compatible.
Like open source licenses by Google, the Android’s source
code is published. By means of a stack of software
components, the Android operating system is based,
which is organized by five major layers like appeared in
Figure 1. Android operating system and platform
components and layers are described and analyzed as
Threats and Defenses of Android Security (Brähler, 2010).
Fig 1: Android operating system architecture
The Linux kernel is the most important among all layers
of the whole system. This layer for the embedded
environment is customized especially containing limited
resources. The entire Android operating system is
constructed on top of the Linux kernel and by Google
extra architectural changes has been made. This layer as
well performances as a deliberation layer between the
software and hardware layers. The basic system
functionality is provided Linux kernel such as the
management of device, memory, and process. Beside, an
array of device drivers is provided by Linux kernel
which makes the job simpler while communicating the
peripheral devices with Android.
The native libraries of the Android on top of the Linux
kernel were developed. For handling different types of
data on the device this layer is enabled. Different useful
libraries are provided for Android OS as the well-
function. These libraries are developed for a particular
hardware and were written in C or C++ language. As an
example of native libraries such as for Internet security
the SSL libraries are used, for providing various media
codecs the Media framework is used, to show 2D or 3D
graphics content the OpenGL is used, the common
library libc, for storing data the SQLite database engine
is used, and to display HTML content the Web browser
engine WebKit is used.
The Android runtime has an element named Dalvik
Virtual Machine (DVM), which is an interpreter for byte
code that has been converted from Java Virtual Machine
(JVM) particularly designed and enhanced for Android.
From Linux core features the Dalvik takes advantage
such as memory management, multitasking execution
environment, and multi-threading, which is for Java
language it is very essential. Dalvik provides control to
applications to work as a process immediately with its
own VM and the Linux kernel. While Dalvik is utilizing
JVM, which gives clients a group of APIs and libraries to
design and create Android application mainly via Java

programming language.
The application framework is written in Java and it is a
collection of services that together make an environment
for managing and running Android applications. As Java
classes, the services are provided to applications. These
services are allowed for application developers to use in
their applications. The main services of application
framework are View System, Notifications Manager,
Resource Manager, Content Providers, and Activity
Manager.
Application layer which is the last layer is placed on the
upper of the Android operating system stack. This layer
involve both of the preinstalled application and third-
party application developed and created by informal
application developers, such as Contacts Manager,
Browser, and Email apps.
2. Android Security Threats
Android operating system security is designed as a
permission-based mechanism which manages and
control the admission and approval of third-party
Android apps to reach critical resources. This
permission-based mechanism is extensively criticized for
the inefficient permission management and controlling
the application permissions, by end-users, marketers,
and developers. Let's say, all permission requests from
an app can either be accepted by users to install it or not.
Here, the major security threats of the Android will be
discussed, which makes the user’s information
vulnerable to leak and places the privacy at risk (Faruki
et al., 2015).
2.1. Information Leakage
The design of the present Android architecture restricts
applications from using or accessing other applications
or resources except it is approved by the users. Before
installing and using an application the user has to grant
all access requests of the resources. Information leakage
happens when without any restriction from OS the users
grant resources. However, permission control
mechanism of the Android Operating system
ineffectively protect user’s resource and privacy from
malware.
2.2. Privilege Escalation
Privilege escalation threats are utilized by taking
advantage of freely accessible Android kernel
vulnerabilities to obtain higher or all access to resources
which is typically protected from a user or app. Such
kind of threat can be a consequence of unauthorized
activities from apps with extra privileges than intended,
as a result, can lead to many sensitive information
leakages. To obtain access to the critical permissions the
Android exported components can be exploited.
2.3. Repackaging Apps
In the Android operating system, one of the most
common and important security threat is Repackaging.
Repackaging by using reverse-engineering techniques
can apply decompiling/disassembling of .apk files and
inserting malicious code inside the main source code. By
using of repackaging techniques as a normal app with
distract the malicious code can be disguised. The
differentiate between a normal app and a repackaged
malicious code can be difficult because as the same way
as the legitimate one the repackaged app typically
appears to function.
2.4. Denial of Service Attack
The growing quantity of mobile devices which they are
connected to the Internet as a big network which could
be a stage for evolution of DoS attacks. Because the
smartphones are not fortified or having less protections
compared with PCs, the developers of malicious
applications find it as an appropriate platform for DoS
attacks. The primarily objectives of DoS attacks are
targeting Overusing limited CPU, battery power,
memory, and network bandwidth.
2.5. Colluding
The colluding threat is happening from the side of the
users. Here, users deploy a group of applications that
have the same certificate and grant various sorts of
permissions could be non-sensitive or sensitive. These
applications after installing it can get access to all their
resources and permissions by taking advantage of a
shared UID.
3. Proposed Solutions
Android security solutions separated into two kinds: 1)
Static; 2) Dynamic which both can use for vulnerability
assessment, analysis, and detection. Static methods are
fast, yet it needs to manage false-positives sensibly.
Dynamic methods, however time-consuming, are
exceptionally useful when applications are extremely
obscured. There are also hybrid methods that merge both
dynamic and static methods together with the limitations
of both.
3.1. Crowdroid
In (Burguera, Zurutuza, & Nadjm-Tehrani, 2011) they
proposed a Crowrdoid system which is a behavior based
malware detection. Which is having two mechanisms, a
crowd sourcing application that should be deployed on
user-devices and other is for malware detection which is
a remote-server. The crowd sourcing application
communicates with the remote server by sending the
behavioral data and this data could be as an application
log file. The application log file contains fundamental
device info, behavioral data, and a list of installed apps.
This data is processed at the remote-server to produce
feature vectors which later might be analyzed by 2-
means partition clustering for recognizing if the app
malicious or benign. An application report is created and
deposited in the remote-server’s database.
3.2. AndroSimilar
In (Faruki, Ganmoor, Laxmi, Gaur, & Bharmal, 2013)
they propose AndroSimilar method which is used

statistically implausible feature selection by means of
similarity digest hashing mechanism. The proposed
solution efficiently detects code obfuscated malware,
control flow obfuscation, method renaming, and junk
method addition. Signature produced through the
proposed method is sufficiently solid to detect unknown
samples obfuscated with several code obfuscation
techniques which they are not detected by the common
Antivirus application. Manually analyzing the suspected
samples detected by the method in which through
present Android malware signatures is verifying their
similarity. This proposed signature method is strong in
counter to repackaged apps.
3.3. Kirin
In (Bahman Rashidi, Fung, & Vu, 2014) they proposed
Kirin method the main goal of this method is to
qualifying malicious applications at install time utilizing
certification process on applications. On app’s requested
permissions this method uses a set of previously defined
security rules for finding the matched permission
requests that are considered as a malicious. The defined
rules are based on the sensitive permissions which lead
to abusing of dangerous activities and permissions.
3.4. RecDroid
In (B Rashidi, Fung, & Vu, 2014) they proposed the
RecDroid method it is a framework through
crowdsourcing for resource accessing permission
control. RecDroid tries to help users to choose the correct
choice to decide if a permission request ought to be
denied or accepted. RecDroid is an approval framework
that brings together the user’s permission responses to
the application’s permission requests, also for evaluating
the level of experience of users this method is utilized
and to figure a proper response to the permission request
even to be rejected or accepted. RecDroid depends on a
slight group of seed expert users that might generate
dependable recommendations for a slight group of apps.
To ignore malicious responses and detecting the
malicious clients this method in addition utilizes a game-
theoretic Bayesian.
3.5. Aurasium
In (Xu, Saïdi, & Anderson, 2012) they proposed
Aurasium technique which is a very useful technique
that takes control of execution of applications, by forcing
arbitrary security rules at runtime. In order to have the
ability to do that, Aurasium includes code for rules
enforcement to Android applications by repackages it.
Aurasium can apply security rules across multiple
applications not only at individual application. Any
privacy and security violations are informed to the user.
So, it rejects the necessity of manipulating an Android
operating system to monitor application behavior. It
interferes when the application accessing sensitive data
like phone identifiers, messages, contacts and
implementing shell-commands by requesting the user
for approval regarding the same.
3.6. FireDroid
In (Russello, Jimenez, Naderi, & van der Mark, 2013) they
propose FireDroid which is a framework that is policy-
based by interleaving process system calls can
implementing security policies. In this method an
application monitor is generated for tracking all
processes spawned in Android operating system and
based on humanly managed policies can accept or reject
them. At runtime, the FireDroid can identify if an app is
executing illegitimate or potentially harmful actions by
interrupting the system calls the app executes.
Regardless of if the malware is repackaged or a new type
of an existing one: FireDroid can detect and implement
the proper security policies once the malware executes
harmful system calls.The objective of FireDroid is that it
is entirely transparent to the apps along with Android
operating system. By this the users are not complicated
with a heavy interaction.
3.7. DroidScope
In (Yan & Yin, 2012) they proposed DroidScope which is
a Virtual Machine Introspection (VMI) that is for
Android applications is dynamic analysis framework.
Contrasting other dynamic analysis frameworks, it
doesn't be located inside the emulator but creating
Dalvik-level and OS-level semantics by locating outside
the emulator. By this, even in the kernel, the privilege
escalation attacks can be detected. It likewise makes the
attackers mission of disorderly analysis hard. upon
QEMU emulator the DroidScope is built, and likewise,
make a set of APIs available to modify analysis needs to
human analysts.
3.8. RiskMon
In (Jing, Ahn, Zhao, & Hu, 2014) they propose the
RiskMon method which is integrated runtime behaviors
and user’s expectations of authorized apps to produce a
risk assessment baseline in which can seizures suitable
behaviors of apps. User’s expectations on the apps are
the key part of the operating system. Initially, the user’s
expectations of the deployed applications are gathered
and the position of permission sets based on their
relevancy to the corresponding app. After that,
depending on the gathered data from the user, it
generates or produce the risk assessment baseline for the
apps. Lastly, utilizing the produced baseline, the method
ranks deployed apps depending on the risk of the
application’s interactions.
3.9. RiskRanker
In (Grace, Zhou, Zhang, Zou, & Jiang, 2012) they propose
RiskRanker is a practical active procedure to recognize
zero-day Android malicious apps. It attempts to assess
potential security risks caused by unauthorized
applications. They design an automated system with the
purpose of analyzing the harmful behavior of
applications dynamically. The proposed method

achieves a two-stage risk analysis. Firstly, they
recognizes applications with medium and high risk. The
purpose of recognizing these applications it traces non-
obfuscated implementations of applications that invoke:
privacy violation attacks, illegal cost creation, and
launching root exploits. Secondly, so as to determine
those applications that encrypt exploit code to avoid the
previous stage analysis it achieves an additional
investigation by analyzing suspicious application
behavior. To solve this encounter, they make a set of
heuristics to map applications to associated risk types
(Low, Medium, and High risk).
3.10. DroidRanger
In (Zhou, Zhou, Jiang, & Ning, 2012) they present a
survey to estimate the security of applications on some
existing unofficial Android application markets and
Google Play. The DroidRanger has two-stage analysis to
detect zero-day malware and existing known malware.
They utilize a permission-based behavioral footprinting
scheme so as to detect known malware. The other stage,
to identify certain inherent behaviors of a zero-day
malware they apply a heuristics-based filtering scheme.
The summary of all previous proposed solutions is
presented in Table.1.
Table.1. Proposed solutions
Proposed
Solutions
Objective Mechanisms
Assessment Analysis Detection Static Dynamic
Crowrdoid
AndroSimilar
Kirin
RecDroid
Aurasium
FireDroid
DroidScope
RiskMon
RiskRanker
DroidRanger
4. Conclusion
Alongside the expanding of Android smartphones, the
quantity of Android applications, as well as malware is
expanding daily. Regardless of existing Android security
system, malware exploits the existing security system’s
weakness to reach and access the granted resources. By
this means, various solutions have been proposed in
order to control and prevent the vulnerabilities in
Android platform. In this review paper, the proposed
solution is divided into two groups static and dynamic
and into three goals assessment, analysis, and detection.
REFERENCES
Android. (2017). Android Security Overview. Retrieved
from https://source.android.com/security/
Appbrain. (2017). Number of android applications.
Retrieved July 28, 2017, from
http://www.appbrain.com/stats/number-of-android-
apps
Brähler, S. (2010). Analysis of the Android Architecture.
Karlsruhe Institute for Technology, 52. Retrieved from
http://os.ibds.kit.edu/downloads/sa_2010_braehler
-stefan_android-architecture.pdf
Burguera, I., Zurutuza, U., & Nadjm-Tehrani, S. (2011).
Crowdroid: Behavior-Based Malware Detection
System for Android. Proceedings of the 1st ACM
Workshop on Security and Privacy in Smartphones and
Mobile Devices - SPSM ’11, 15.
https://doi.org/10.1145/2046614.2046619
Faruki, P., Bharmal, A., Laxmi, V., Ganmoor, V., Gaur, M.
S., Conti, M., & Rajarajan, M. (2015). Android security:
A survey of issues, malware penetration, and
defenses. IEEE Communications Surveys and Tutorials,
17(2), 998–1022.
https://doi.org/10.1109/COMST.2014.2386139
Faruki, P., Ganmoor, V., Laxmi, V., Gaur, M. S., &
Bharmal, A. (2013). AndroSimilar : Robust Statistical
Feature Signature for Android Malware Detection.
Proceedings of the 6th International Conference on Security
of Information and Networks, (September 2015), 152–159.
Grace, M., Zhou, Y., Zhang, Q., Zou, S., & Jiang, X. (2012).
RiskRanker: Scalable and Accurate Zero-day Android
Malware Detection. 10th International Conference on
Mobile Systems, Applications, and Services, 281–294.
Jing, Y., Ahn, G.-J., Zhao, Z., & Hu, H. (2014). RiskMon :
Continuous and Automated Risk Assessment of
Mobile Applications. Proceedings of the 4th ACM
Conference on Data and Application Security and Privacy
- CODASPY ’14, 99–110.
Rashidi, B., Fung, C., & Vu, T. (2014). On lightweight
mobile phone application certification. Proceedings of
the ACM MobiCom Workshop on Security and Privacy in
Mobile Environments, 235–245.

Rashidi, B., Fung, C., & Vu, T. (2014). RecDroid: A resource
access permission control portal and recommendation
service for smartphone users. 2014 ACM MobiCom
Workshop on Security and Privacy in Mobile
Environments, SPME 2014, 13–17.
Russello, G., Jimenez, A. B., Naderi, H., & van der Mark,
W. (2013). FireDroid: hardening security in almost-
stock Android. Proceedings of the 29th Annual Computer
Security Applications Conference, 319–328.
Xu, R., Saïdi, H., & Anderson, R. (2012). Aurasium:
Practical Policy Enforcement for Android
Applications. Proceedings of the 21st USENIX
Conference, 27. Retrieved from
https://www.usenix.org/system/files/conference/
usenixsecurity12/sec12-
final60.pdf%5Cnhttp://dl.acm.org/citation.cfm?id=2
362793.2362820
Yan, L. K., & Yin, H. (2012). DroidScope: Seamlessly
Reconstructing the OS and Dalvik Semantic Views for
Dynamic Android Malware Analysis. In USENIX
Security Symposium.
Zhou, W., Zhou, Y., Jiang, X., & Ning, P. (2012). Detecting
repackaged smartphone applications in third-party
android marketplaces. Proceedings of the Second ACM
Conference on Data and Application Security and Privacy -
CODASKY ’12, 317–326.

Android security

More Related Content

What's hot (20)

Similar to Android security (20)

More from Dr Amira Bibo (8)

Recently uploaded (20)

Android security