Journal of Network and Computer Applications 61 (2016) 59–80

Contents lists available at ScienceDirect

Journal of Network and Computer Applications

journal homepage: www.elsevier.com/locate/jnca

Review

Authentication in mobile cloud computing: A survey

Mojtaba Alizadeh a,b, Saeid Abolfazli c,n, Mazdak Zamani d, Sabariah Baharun b,
Kouichi Sakurai a
a
Department of Informatics, Graduate School and Faculty of Information Science, Kyushu University, Fukuoka, Japan
b
Malaysia-Japan International Institute of Technology (MJIIT), Universiti Teknologi, Malaysia
c
YTL Communications and Xchanging, Malaysia
d
Department of Computer Science, Kean University, NJ, USA

art ic l e i nf o a b s t r a c t

Article history: Mobile cloud computing (MCC) is the state-of-the-art mobile distributed computing model that incor-
Received 29 March 2015 porates multitude of heterogeneous cloud-based resources to augment computational capabilities of the
Received in revised form plethora of resource-constraint mobile devices. In MCC, execution time and energy consumption are
21 September 2015
significantly improved by transferring execution of resource-intensive tasks such as image processing, 3D
Accepted 18 October 2015
Available online 6 November 2015
rendering, and voice recognition from the hosting mobile to the cloud-based resources. However,
accessing and exploiting remote cloud-based resources is associated with numerous security and privacy
Keywords: implications, including user authentication and authorization. User authentication in MCC is a critical
Cloud computing requirement in securing cloud-based computations and communications. Despite its critical role, there is
Mobile cloud computing
a gap for a comprehensive study of the authentication approaches in MCC which can provide a deep
Security
insight into the state-of-the-art research. This paper presents a comprehensive study of authentication
Authentication
methods in MCC to describe MCC authentication and compare it with that of cloud computing. The
taxonomy of the state-of-the-art authentication methods is devised and the most credible efforts are
critically reviewed. Moreover, we present a comparison of the state-of-the-art MCC authentication
methods considering five evaluation metrics. The results suggest the need for futuristic authentication
methods that are designed based on capabilities and limitations of MCC environment. Finally, the design
factors deemed could lead to effective authentication mechanisms are presented, and open challenges
are highlighted based on the weaknesses and strengths of existing authentication methods.
& 2015 Elsevier Ltd. All rights reserved.

Contents

1. Introduction and motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

2. Authentication in mobile cloud computing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
2.1. Mobile cloud computing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
2.2. User authentication in mobile cloud computing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
2.3. MCC vs. cloud computing authentication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
3. The state-of-the-art of authentication approaches in MCC: taxonomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
3.1. Cloud-side authentication methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
3.1.1. Identity-based authentication methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
3.1.2. Context-based authentication methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
3.2. User-side authentication methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
3.2.1. Identity-based authentication methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
3.2.2. Context-based authentication methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3.3. Evaluation criteria for authentication in MCC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3.3.1. Usability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
3.3.2. Efficiency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

n
Corresponding author.
E-mail addresses: amojtaba2@live.utm.my, 9IE14015W@s.kyushu-u.ac.jp (M. Alizadeh), abolfazli@ieee.org (S. Abolfazli), mzamani@kean.edu (M. Zamani),
sabariahb@utm.my (S. Baharun), sakurai@csce.kyushu-u.ac.jp (K. Sakurai).

http://dx.doi.org/10.1016/j.jnca.2015.10.005
1084-8045/& 2015 Elsevier Ltd. All rights reserved.
60 M. Alizadeh et al. / Journal of Network and Computer Applications 61 (2016) 59–80

3.3.3. Security and robustness. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72

3.3.4. Privacy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
3.3.5. Adaptable to MCC environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
4. Prospective authentication algorithms in MCC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
4.1. Mobile device characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
4.2. Usability preferences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
4.3. Security and privacy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
4.4. Mobility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
4.5. Support heterogeneity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
4.6. Adaptiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
5. Open challenges. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
5.1. Heterogeneous infrastructure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
5.2. Seamless handover . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
5.3. Identity privacy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
5.4. Resource scheduling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
6. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Acknowledgement. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77

1. Introduction and motivation
The mobile cloud computing (MCC) is “a rich mobile computing
technology that leverages unified elastic resources of varied clouds
and network technologies toward unrestricted functionality, storage,
and mobility to serve a multitude of mobile devices anywhere,
anytime through the channel of Ethernet or the Internet regardless of
heterogeneous environments and platforms based on the pay-as-
you-use principle.” (Sanaei et al., 2013). MCC incorporates cloud
computing, mobile computing, and wireless networking and aims to
provide cloud-based services to the mobile consumers (Abolfazli
et al., 2014a; Fernando et al., 2013). In MCC, execution time and
energy consumption are significantly improved by transferring
execution of resource-intensive application from the hosting mobile
to the cloud-based resources.Therefore, once MCC is fully deployed,
the mobile devices do not require high resources, such as central
processing unit (CPU), random access memory (RAM), storage, and
particularly battery, because the entire data or complex computing
are manipulated in the remote cloud-based resources (Ko et al., 2012;
Abolfazli et al., 2012; Liu et al., 2015). MCC has emerged as a subset of
cloud computing to enable intensive on-demand elastic computing
and storage on the go to the potential mobile users. Mobile devices,
particularly tablet personal computers, smart phones, and PDAs are
becoming an integral part of today's lifestyle as they are convenient
and effective communication endpoint. The swift development of
mobile computing has become a forceful pattern in IT technology's
development in commerce and related fields. According to Cisco
visual networking index statistics (Cisco, 2014), the usage of smart-
phone and global mobile data traffic grew 50 and 81 percent in 2013,
respectively. Nevertheless, performance and functionality of mobile
devices are hindered by several limitations, particularly computing
and storage resources (i.e., CPU, RAM), wireless communication
throughput, battery life, local data safety, communication security,
and mobility impeding development of the quality of service (Abol-
fazli et al., 2014a). The idea of remote computing and the process of
augmenting mobile devices using remote cloud-based computing
and storage resources is envisioned to overcome the inherent chal-
lenges and shortcomings in mobile computing (Aminzadeh et al.,
2015). This is carried out by utilizing other resource providers besides
the mobile device to host the delivery of resource-intensive mobile
applications (Dinh et al., 2013; Alizadeh et al., 2013a,b).
Although MCC is proven to be advantageous in augmenting
computational capabilities of mobile devices and conserving their
native resources, leveraging remote resources introduces several
challenges, including reliability, security, trust, and privacy (Khalil
et al., 2014; Xuanxia et al., 2014; Khan et al., 2013a; Sood, 2012).
Successful diffusion of cloud computing technology with mobile
devices incites users desires for efficient and also secure service
delivery. Furthermore, in MCC environment, typical mobile devi-
ces communicate through the combination of heterogeneous
wireless networks, which is more energy-intensive compared to
wired communication. Hence, reducing mobile devices' resource
consumption is an important and critical problem in delivering
sustainable and long-lasting on-demand services to the end-users
(Shon et al., 2014). Although mobile devices' resource poverty can
be alleviated by cloud computing and cloud-based augmentation
techniques (Abolfazli et al., 2014a), inadequate security manage-
ment inhibits development and successful deployment of cloud-
connected security-sensitive applications in broad areas, including
health-care, financial services, and e-government services.
Researchers in several efforts (Yang et al., 2014; Li and Li, 2014;
Si et al., 2014; Xia et al., 2014; Sookhak et al., 2014; Kaewpuang
et al., 2013; Rahimi et al., 2013; Yang et al., 2013; Ma and Wang,
2012; Satyanarayanan et al., 2009; Ra et al., 2011) have studied
varied aspects of MCC, including task outsourcing, heterogeneity,
virtualization, energy saving, and remote auditing, aiming to
enhance the MCCs performance and efficiency. However, security
(as another crucial aspect of MCC), particularly authentication is
overlooked. The security challenges in MCC are twofold, namely
cloud security and mobile network security because of the co-
existence of cloud computing and mobile computing in MCC (Peng
et al., 2014; Morrow, 2011; Zissis and Lekkas, 2012; Dijiang et al.,
2011). One of the most important security issues for MCC users is
authentication and authorization (Esposito and Ciampi, 2015; Yu
and Wen, 2012; Riley et al., 2011). As an example, a lost or stolen
mobile device could be abused to access a host and download
sensitive data from the cloud, if a mobile user is registered with a
particular cloud service provider, both mobile device and cloud
server should authenticate each other in order to secure the
communication when the mobile user accesses the cloud from
different locations using heterogeneous networks and various
mobile devices (Clarke et al., 2002).
Several studies (Xu et al., 2013; Wang et al., 2013; Noureddine and
Bashroush, 2013; Ghazizadeh et al., 2014; Singh and Singh, 2012; Guo
et al., 2012; Dinesha and Agrawal, 2012; Li et al., 2013; Zhi-Hua et al.,
2012; Zhang et al., 2012; Yongqing and Xiang, 2012; Yassin et al.,
2012; Wang and Jia, 2012; Sang-Ho et al., 2012; Ahn et al., 2011) have
been conducted to propose suitable authentication schemes in cloud
computing. However, authentication in MCC, as one of the most
crucial security countermeasures, has not been studied yet. Moreover,
 M. Alizadeh et al. / Journal of Network and Computer Applications 61 (2016) 59–80 61

several efforts have been undertaken to study varied aspects of the Table 1
MCC. Khan et al. (2013a) evaluated and identified the security issues List of acronyms and corresponding full forms.
of the existing security schemes in MCC infrastructure. Furthermore,
Acronym Full form
the security issues and challenges of MCC are discussed in Khan et al.
(2013b,c), Popa et al. (2013), Jin et al. (2013), Kumar and Rajalakshmi 2D 2 dimensional
(2013), Alizadeh and Wan (2013), Hui et al. (2013) and Shahzad and 2G 2nd generation
Hussain (2013) by surveying the current state of mobile cloud devices 3G 3th generation
4G 4th generation
security vulnerabilities, and exploring the various possible solutions. AES Advanced encryption standard
However, authentication as one of the most crucial security coun- AS Authentication server
termeasures, has not been studied yet. Therefore, we aim to fill the CER Crossover error rate
gap by conducting a comprehensive survey to assess and analyze CPU Central processing unit
FAR False acceptance rate
various authentication schemes in MCC aiming to furnish an
FRR False rejection rate
insightful view of the state-of-the-art of authentication methods GPS Global positioning system
in MCC. IA Integrated authentication
The significance of applying appropriate authentication meth- ID Identifier
IMEI International mobile station equipment identity
ods and also lack of suitable authentication mechanisms in MCC
IMSI International mobile subscriber identity
based on security and usability criteria motivates us to evaluate LTE Long term evolution
and analyze the state-of-the-art authentication approaches. MCC Mobile cloud computing
The main contributions of this paper are: MDA Message digest algorithm
NFC Near field communication

We provide a description of MCC security challenges. OTP One-time pad
PDA Personal digital assistant

Comprehensive survey of the state-of-the-art authentication PIN Personal identification number
methods in MCC is provided. PRNG Pseudorandom number generator

Security and performance of authentication mechanisms are QR Quick response
ROC Relative operating characteristic
analyzed for MCC based on five critical metrics.

We identify and discuss several important factors deemed could RSA
SI
Rivest Shamir Adleman
State identifier
contribute to the successful development of future authentica- SLA Service-level agreement
tion methods for mobile devices in MCC environment. SMS Short message service

Several open challenges that ground future researches are SNR Signal to noise ratio
SSL Secure socket layer
discussed.
TCG Trusted computing group
TLS Transport layer security
Authentication in MCC benefits communications and networking TNC Trusted network connect
communities by providing a comprehensive insight into the domain TPA Third party agent
TTP Trusted
so future wireless communication technologies and architectures can
URI Uniform resource identifier
efficiently and effectively furnish cloud-based resources to the mobile VM Virtual machine
users with high security and low footprint. Discussed evaluation WAN Wide area network
criteria highlight effective factors as a guideline to design suitable Wi-Fi Wireless fidelity
authentication schemes, which can benefit the research community. WiMAX Worldwide interoperability for microwave access
WLAN Wireless local area network
The open challenges grant future research directions toward pro-
ZKP Zero knowledge proof
posing a suitable authentication scheme that mitigates security issues
in MCC. In this paper, the terms mobile devices and smartphones are
used interchangeably with similar notion. Table 1 shows the list of computing phenomenon. MCC as the state-of-the-art mobile dis-
acronyms used in the paper. tributed computing technology incorporates three principal tech-
Section 2 provides a brief introduction to MCC and discusses nologies, namely mobile computing (Imielinski and Korth, 1996),
the evaluation criteria to analyze authentication methods in this cloud computing (Mell and Grance, 2011), and wireless networking
environment. Section 3 surveys the state-of-the-art authentication (Lei et al., 2013). Therefore, MCC can be defined as “a rich mobile
methods in MCC. In Section 4, the important factors that are
computing technology that leverages unified elastic resources of
deemed could benefit design and development of future MCC
varied clouds and network technologies towards unrestricted func-
authentication methods are presented. The open challenges are
tionality, storage, and mobility to serve a multitude of mobile devices
discussed in Section 5. Finally, we conclude this study in Section 6.
anywhere, anytime through the channel of Ethernet or the Internet
regardless of heterogeneous environments and platforms based on
2. Authentication in mobile cloud computing the pay-as-you-use principle” (Sanaei et al., 2013).
In MCC, a shared pool of various configurable cloud-based
In this section, we present a brief introduction over MCC from computing resources is utilized to enhance and optimize mobile
authentication point of view. In the first part, definition of MCC is devices computing capabilities such as executing resource-
presented based on the existing studies. We then describe possible intensive applications. MCC has penetrated into a very large
MCC architectures and its different components. Furthermore, number of domains, and researchers are increasingly adopting
authentication in MCC is presented and the significant role of cloud computing to augment mobile devices in critical domains,
authentication in successful adoption of cloud-based mobile appli- particularly health-care (An et al., 2014; Al-Zoube and Alqudah,
cations is highlighted.
2014; Hoang and Chen, 2010; Doukas et al., 2010), education
2.1. Mobile cloud computing (Mahalingam and Rajan, 2013; Chen et al., 2013; Dong et al., 2012;
Huang, 2011), remote monitoring (Xu et al., 2012; Zhang et al.,
In this part, some of the credible MCC definitions are provided to 2014), tourist industry (Song et al., 2012; Pal and Henderson, 2013;
furnish the fundamental knowledge of this rapidly emerging Li et al., 2011), and transportation (Chandra et al., 2013).
62 M. Alizadeh et al. / Journal of Network and Computer Applications 61 (2016) 59–80

On the other hand, MCC can also be referred to a mixture of cloud
computing and mobile web, which are dominant preferred tools of
mobile users when consuming Internet services and applications (Liu
et al., 2010; Christensen, 2009). MCC services focus on furnishing
varied cloud-based services, particularly computing and data storage
to mobile users. Therefore, cloud-connected mobile users can per-
form infinite computing and data storage on demand.
Cloud resources are different for mobile users compared to
immobile users. Immobile users consume computing and storage
resources from the private or public cluster of visualized servers,
known as cloud data centers mostly via wired connectivity.
However, intrinsic and non-intrinsic limitations of mobile devices,
including resource poverty, interruptible battery, and wireless
communication (Abolfazli et al., 2014b) are obligating highly het-
erogeneous types of cloud-based resources (Sanaei et al., 2013) to
fulfill varied computing requirements of mobile users in different
occasions. Throughout the MCC efforts, four types of computing
resources (known as cloud-based resources) are introduced,
namely distant immobile clouds, proximate immobile computing
entities, proximate mobile computing entities, and hybrid
resources (Abolfazli et al., 2014a; Satyanarayanan et al., 2009). In
the distant immobile cloud, the mobile user connects to the sta-
tionary cloud servers in distance through the risky channel of the
Internet. Though the distant stationary cloud server such as the
public cloud provides more secure enforcements, they are vul-
nerable to security breaches and crashes due to bulky volume of
sensitive data such as Amazon EC2 crash (Cachin and Schunter,
2011). The second cloud-based resource type is the stationary
computers that are located near the mobile nodes. These machines
are available for the mobile device in public places such as air-
ports, coffee shops, and malls, and can process resource-intensive
parts of mobile applications. The security and privacy issues such
as lack of strong security approaches, security infrastructure
limitations, and insufficient maintenance, hinder utilization of
such nearby resources (He et al., 2015a). Third resource type refers
to the proximate mobile computing entities such as PDAs, Tablets,
Laptops, wearable computers, and smartphones that provide
resources to the other nearby mobile nodes. The security and
privacy threats such as eavesdropping, denial-of-service, Trojans,
malware, viruses, worms, and mobile loss, are the critical concerns
using these kinds of resources (He et al., 2015b; Louk
et al., 2014; Allam et al., 2014; Mylonas et al., 2013; Wang et al.,
2012). The last type is the hybrid infrastructures which are com-
prised of various kinds of distant and proximate computing
machines.
The mobile devices are connected to the cloud-based resources
dominantly through the risky channel of the Internet via the wireless
medium, though Internet-free connection to nearby or private
resources is also conceivable. Therefore, the remote computing and
data transmission are completed in collaboration of mobile clients,
cloud-based resources, and heterogeneous wireless technologies.
According to the classification of cloud-based resources, four possible
architectures depicted in Fig. 1, can be plausible for MCC.
Each of the plausible MCC architectures has different security and
privacy requirements depend on the type of cloud-based resources
and wireless communication technology/medium. The security and
privacy threats within different parts of MCC, including cloud
resources (Xiao and Xiao, 2013), mobile devices (La Polla et al., 2013),
and wireless networking makes designing a secure framework more
challenging. Though mobile devices have some resource limitations to
perform complex cryptographic algorithms, the cloud resources can
process resource-intensive algorithms instead of mobile devices.
In this research, the security issues as one of the important con-
cerns in MCC are considered, and some proposed solutions are
reviewed. The user authentication is highly important to protect
networks from different security threats (Furnell et al., 2000, 2008;

Distant Immobile Proximate Mobile Computing

Proximate Immobile Computing
Cloud

BTS Internet
Access Point
Access Point

Hybrid Computing

Access Point

BTS

BTS
Internet

Fig. 1. Four mobile cloud computing architecture models – (a) distant immobile clouds perform elastic computing, (b) proximate immobile computing entities near the user
perform elastic computing, (c) proximate mobile computing entities in user vicinity perform elastic computing on behalf of user and (d) hybrid model converges varied types
of cloud-based resources to perform elastic computing.
 M. Alizadeh et al. / Journal of Network and Computer Applications 61 (2016) 59–80
Clarke and Furnell, 2007; Simmons, 1988; Weiwei et al., 2011). Suc-
cessful adoption of MCC highly necessitates robust and effective
authentication solutions by which users can utilize the cloud-based
services for their mobile devices anytime, anywhere, from any mobile
device with low computing cost on the native resources. The MCC
authentication is different from a typical mobile device because in
MCC environment, the mobile device connects to the Internet to
perform authentication. Furthermore, the resource-intensive parts of
authentication mechanism can be transferred and processed in cloud
servers using a suitable algorithm.
During the last few years, authentication for cloud computing
has been investigated in several researches (Ghazizadeh et al.,
2014; Xu et al., 2013; Wang et al., 2013; Noureddine and Bash-
roush, 2013; Singh and Singh, 2012; Guo et al., 2012; Dinesha and
Agrawal, 2012; Chow et al., 2010; Li et al., 2013; Zhi-Hua et al.,
2012; Zhang et al., 2012; Yongqing and Xiang, 2012; Yassin et al.,
2012; Wang and Jia, 2012; Sang-Ho et al., 2012; Ruj et al., 2012);
however, comprehensive study of MCC authentication, which is
crucial in design and development of future authentication
methods is lacking and demands further efforts. The lack of secure
and efficient authentication methods necessitates a vital need to
conduct a comprehensive research to gain deep insight into the
filed. Different authentication methods aiming to improve the
MCC security are analyzed in the following parts.
2.2. User authentication in mobile cloud computing
User authentication in MCC is the process of validating the
identity of the mobile user to ensure that the user is legitimate to
access mobile cloud resources (Schwab and Yang, 2013). Authen-
tication as a critical aspect of security enforcement approaches in
MCC is essential to protect users against existing security and
privacy issues by preventing unauthorized access to the mobile
cloud user information (Park et al., 2011; Zhu et al., 2009). The
security and privacy issues of mobile cloud users are the main
hurdles to the successful and rapid MCC deployment, which exist
in three MCC components, namely cloud, wireless communication,
and mobile device. Therefore, considering characteristics and
computing limitations of mobile devices, effective and efficient
MCC authentication solutions are expected to be lightweight with
the least possible computing, memory, and storage overheads.
The aim of effective authentication solutions is to minimize the
security threats to mobile devices. Discussion over the security
and privacy threats in MCC is out of scope of this paper, and thus
we only point the most important threats and provide relevant
references for interested readers. Some of the most important
security threats to mobile users are information leakage, denial of
service, malfunction of devices and theft or loss of the device (Park
et al., 2011). Moreover, security threats found in mobile devices
can manifest as attacks via the services offered through the
wireless networks, including network profiling, information lea-
kages by sniffing, session hijacking, and jamming (La Polla et al.,
2013; Xiao and Xiao, 2013; Morshed et al., 2011; Zissis and Lekkas,
2012; Khan et al., 2013a). Furthermore, the potential threats in the
cloud computing system include denial of services, information
leakages due to mismanagement, authentication threats and
control of access with default applications (Jang et al., 2011), can
be considered in MCC.
In the following parts, The differences between authentication
in cloud computing environment and MCC are discussed in the
following part.
2.3. MCC vs. cloud computing authentication
Authentication mechanisms in MCC are different from cloud
computing in several ways. The capabilities and limitations of
63
mobile devices introduce some challenges for designing effective
and efficient authentication mechanisms. Some of the most
important differences between authentication requirements and
principles in MCC and cloud computing are described below.

Resource limitations: Resource limitation among mobile devices
refers to incapacitation in computational power, battery life-
span, and storage capacity in comparison to typical computers
in cloud networks. Computational performance and function-
alities of mobile devices are significantly hindered by such as
incapacitation. Consequently, most of the mobile devices are
incapable of efficiently executing sophisticated resource-
intensive encryption algorithms, for example, RSA algorithm
with 2048 bits (Sheng and Gong, 2010). However, non-mobile
cloud users are benefiting from plenty of local computational
resources, high speed wired Internet connection, and contin-
uous power source, which allow resource-intensive authentica-
tion algorithms without serious effect on user experience
(Qureshi et al., 2011). Therefore, mobile devices require robust
but lightweight authentication mechanisms that can ensure
authenticity of users without draining local resources (Yang
et al., 2010).

Mobile device sensors: Mobile device sensors such as touch
screen, gyroscope, accelerometer, camera, digital compass, and
microphone give the researcher this opportunity to add other
authentication factors, particularly biometrics to improve the
level of security in MCC (Giuffrida et al., 2014; Lane et al., 2010;
Jeong et al., 2013; Le et al., 2013). The authentication mechan-
isms can benefit from the various types of mobile device
sensors, which can measure user's biometric attributes, such
as fingerprint, and facial, retina, iris, voice, gait and keystroke
patterns that are used as authentication factors (Omri et al.,
2012; Al Rassan and AlShaher, 2014). Although the authentica-
tion methods in cloud computing can benefit from peripheral
accessories and equipments on end-user computers, additional
cost can create a hurdle. Besides the significant benefits of latest
sensors, they introduce security breach points too that compli-
cate designing authentication methods in MCC. For instance,
researchers in Owusu et al. (2012) could unveil credentials of a
user by decoding accelerometer sensor readings on smartphone
when the user enters his/her credentials.

High mobility: Mobility can originate latency due to WAN
latency that is intensified by signal handoff in the presence of
heterogeneous networks. The miniature nature and mobility of
mobile devices can intensify chance of robbery and loss leading
to high probability of user privacy and security violation in the
absence of robust authentication solutions (Khalil et al., 2014).
In addition, in contrast to static computers, the quality of
connection to the Internet is not stable in MCC because of
mobility of peers (Ardagna et al., 2014). Furthermore, fast
authentication procedure is desired to protect seamless con-
nectivity for mobile devices in roaming (Lingfeng and Hoang,
2013; Mansoor et al., 2015; Jana and Bandyopadhyay, 2013).

Network heterogeneity: The mobile devices connect to Hetero-
geneous Network (HetNet) (Sanaei et al., 2013; Lei et al., 2013),
which has various kinds of radio access such as Wi-Fi, WiMAX,
2G, 3G, 4G and LTE to accomplish the data traffic demand in
MCC. In addition, a proper handoff scheme is critical for
heterogeneous networks to have seamless connectivity and
authentication plays key role in handoff procedure (Avelar
et al., 2015; Bin et al., 2012; Chu et al., 2012). An authentication
method must be designed based on security and performance
requirements of each network technology that is a challenging
part in MCC. In cloud computing environment, the user is
typically immobile and authentication procedure can be done
without using HetNet.
64 M. Alizadeh et al. / Journal of Network and Computer Applications 61 (2016) 59–80

Wired or wireless communication: The type of communication
technology can be considered in designing security frameworks
and risk evaluation schemes to (near) optimally fit authentica-
tion methods based on limitation and capabilities of commu-
nication technology. Typically, the wireless communication that
is used to access cloud resources introduces some security and
privacy issues in MCC compared to mostly used wired com-
munications in cloud environment (Ardagna et al., 2014; Mor-
shed et al., 2011). Moreover, the probability of failure of security
protection such as authentication in MCC is potentially high due
to inconsistency of wireless communications (Honggang et al.,
2014). Besides, the user authentication procedure in cloud
environment can be done through the wired network without
connecting to various wireless networks, which makes design-
ing the authentication mechanism less challenging compare to
MCC environment.
The differences between MCC authentication and cloud com-
puting that are recommended to be considered to propose a sui-
table authentication mechanism are summarized in Table 2.
3. The state-of-the-art of authentication approaches in MCC:
taxonomy
In this section, we comprehensively analyze various methods
based on the type of authentication components used in authen-
tication procedure. We classify authentication methods into two
main categories, namely, cloud-side and user-side authentication
approaches. Each category is again divided into two sub-categories
based on types of authentication credentials. The credential is
defined as a unique identifier that can be used for node authen-
tication. There are two types of credentials based on this classifi-
cation, namely, identity-based and context-based credentials
(Aboudagga et al., 2005). Figure 2 illustrates our devised taxonomy
of authentication in MCC which is described as follows.
3.1. Cloud-side authentication methods
In cloud-side authentication approaches, most of the authen-
tication steps are processed in the cloud server. The cloud
resources bring some advantages to improve the performance of
authentication method by providing elastic processing and storage
capabilities. Furthermore, different kinds of authentication factors
can be incorporated to authentication procedure based on security
requirements of the user. In addition, the cloud-based authenti-
cation methods are more flexible, efficient, and adjustable com-
pared to other authentication methods because of using unlimited
resources of cloud servers.
Although, the cloud-based authentication introduces some
benefits in case of performance and usability, it introduces some
security and privacy issues. In the era that mobile users are con-
suming cloud services from the plethora of different cloud vendors
who store user's data on multiple instances around the globe
(redundant data for data safety), the user's private authentication
Table 2
MCC and cloud authentication comparison.
Metrics MCC Cloud Computing
Resource limitations  N/A
Mobile device features √ N/A
High mobility  N/A
Network heterogeneity  
Wired or wireless communication  N/A
(  ) – challenge, (√) – opportunity and – (N/A) not effective.
information such as password and biometrics are highly exposed
to risk (Rahul and Sharda, 2013). Therefore, robust authentication
method is a critical requirement for mobile-cloud environment.
We further classify cloud-side authentication methods into two
groups of identity-based and context-based that are explained as
follows.
3.1.1. Identity-based authentication methods
In identity-based authentication methods, users are authenti-
cated through user identification attributes such as unique ID,
password, and (seldom) biometrics. However, the user attributes
are usually fixed (users try to simplify the authentication process
by selecting similar user names and passwords for varied vendors)
in this kind of authentication, which introduces some issues such
as exposing private biometrics information and user ID to different
service providers leading to weakening their authentication power
(Zhi-Hua et al., 2012). Consider a particular user who is main-
taining identical user names and passwords for multiple services
such as Gmail, YouTube, and Facebook. If security of one of these
servers is compromised, all other accounts are in risk due to
similarity of the credentials. The authenticator entity checks the
user attributes directly without extra analyzing indirect proce-
dures such as the user behavioral analysis. The authentication
provider is the responsible entity for identity management and
performing primary user authentication.
Below we critically review some of the most credible identity-
based authentication methods.

Multifactor-based authentication: Jeong et al. (2013) proposed a
multi-factor authentication method for MCC. Different authen-
tication features such as basic ID/password, mobile identifica-
tion number, and various bio-information of user are combined
to preserve security. This method enhanced authentication
efficiency and performance by processing these authentication
factors in bulk not one by one in mobile cloud environment.
Proposed architecture comprises of four main entities, includ-
ing mobile devices, storage, a management server, and a cluster
host. The authors assumed that Transport Layer Security (TLS)
protocol/Secure Sockets Layer (SSL) is utilized for communications
between the authentication system and wireless access points. The
task of each entity is applied according to the multi-factor
authentication architecture, and the five employed authentica-
tion parameters are used to authenticate the mobile users,
including: (1) ID/password, (2) international mobile equipment
identity (IMEI), (3) international mobile subscriber identity (IMSI),
(4) voice recognition, (5) face recognition. The cluster host dis-
tributes these five parameters to individual VMs to improve per-
formance of the authentication process. The management server
manages the load balancing on the VMs in the clustered host. To
protect the user in case of mobile device loss and theft, IMEIs and
IMSIs are used. All the user authentication information are stored
in the storage. It is worth noticing that the provided hash function
encrypts IMSI and IMEI to prevent exposure of these two private
IDs. The details of this architecture are described in Fig. 3.
The procedure of this authentication method is similar to
normal multi-factor authentication methods that use biometrics
with other factors together; however the authors designed a
method to be more fit to MCC as they utilized processing capacity
of VMs in MCC environment. The individual VMs process each
factor of authentication simultaneously to reduce the time of an
authentication procedure.
Although, this method improved the performance of authen-
tication compared to other methods in this area, the privacy issues
of using biometrics factors are neglected in security and privacy
evaluation. The bio-information are very sensitive data that are
 M. Alizadeh et al. / Journal of Network and Computer Applications 61 (2016) 59–80 65

State-of-the-art
MCC Authentication

Cloud-side User-side

Identity-based Context-based Identity-based Context-based

FDZ [94] TrustCube [92]

Multi-factors [103] Cloud-based Biometric [105]
QR code-based [119] SMCBA [106]

MDA [117] SeDiCi 2.0 [120] NemoAuth [104]

Fig. 2. Taxonomy of the state-of-the-art authentication in MCC.

ID/PW
Face Recognition
Voice Recognition
IMEI IMEI
Management Server
Hash

VM1

VM2

Wireless AP VMn
Smartphone L2 Switch
Cluster Host

IMEI
ID/PW Face Voice Log
IMSI

Storage
Fig. 3. The multi-factor authentication architecture.

recommended to be confidential and appropriate encryption
algorithms can be applied to preserve the confidentiality of data.
Another security issue that is being ignored in this method, is
mutual authentication between mobile users and MCC server. The
user can check authenticity of the MCC servers to prevent different
attacks such as man-in-the-middle, where an attacker can monitor
or alter communication messages, and masquerading attack,
which means an attacker impersonates either user or authenti-
cator. Although authors could fit devised authentication method
into MCC domain by utilizing the processing power of the cloud
server to improve authentication performance.

Message digest authentication: Dey et al. (2013) proposed an
authentication scheme using message digest (MD) called MDA.
This method is designed based on existing mobile device
hardware and platforms to protect mobile user against different
potential security attacks. The vulnerability of the system is
computed by vulnerability score, Sv, which is a measure of the
number of attacks that the method can prevent. The Sv is
calculated according to the following equation:
Sv ¼ ðN success Þ=ðNÞ
where N is the number of attacks that are launched on the
authentication method, and Nsuccess is the number of recorded
successful attacks. The amount of Sv is between 0.0 and 1.0.
Mutual authentication is an important security countermeasure
that is considered in this method. The procedure of this method
includes two phases; in the first phase, cloud server checks
mobile user authentication and subsequently the mobile device
verifies authenticity of the cloud in the second phase. First
mobile device sends the authentication request to the cloud
server as shown in Fig. 4.
As Figure. 4 shows, the user has two message digests, MDcloud
and MDuser that are used to create MD. The password is hashed and
XORed by user ID, to protect user from an attacker during
authentication. The Pseudo Random Number Generator (PRNG) is
an algorithm to generate random number using seed and State
Identifier (SI). The mobile node sends ♯CF J ETk fEauth_keyi fMDg J SIg to
the cloud server. In this message, ♯CF is a column reference of
stored mobile user and Eauth_keyi is the nth sequence of bits, which
are generated by PRNG. The ETk is calculated by XOR-ing hashed
user password and userID.
The cloud server checks user authentication after receiving
mobile device authentication request message. The procedure of
this phase is shown in Fig. 5.
Firstly, the cloud server checks ♯CF, to find userID and hashed
password. Then, it generates T k to decrypt the received message
ð1Þ from the mobile device. After decrypting the message, the cloud
server obtains MD, and compares both message digests to check
user authenticity. The authenticity of the mobile device is verified
only if MD and MDn messages are match. Once the mobile user is
authenticated to the cloud server, the procedure of cloud server
authentication will be initiated. The cloud server sends its digital
signature that is encrypted by its private key, Pkpriv_cloud , to the
66 M. Alizadeh et al. / Journal of Network and Computer Applications 61 (2016) 59–80

MDcloud MDuser MDcloud MDuser

Hash
Hash

If MD* matches MD, then the cloud MD

MD
is legitimate
Auth_keyi
MD*
User ID From Cloud End

ENC

DScloud DEC
User ID PRNG Hash {Password}
Seed Tk
Pkpub_cloud
XOR SI
DEC
Hash ||
{Password} ETk{PKpub_cloud || MDuser ||MDcloud ||#CF}

Fig. 6. The cloud server authentication procedure in MDA.

Tk
together. Furthermore, user privacy is preserved by hashing userID
and Password. Nevertheless, the security and privacy of this
ENC
method are protected compared to other methods, but the pro-
cedure of this authentication scheme is somehow complicated.
Additionally, it is recommended to transfer some processing steps
that are processed in the mobile device, to the cloud server to
improve the performance of the scheme, however, doing such kind
#CF
of improvement will be more complicated.

Fig. 4. The procedure of the authentication request message in MDA method.

3.1.2. Context-based authentication methods
In Context-based methods, the users are authenticated by
sent to mobile
#CF || ETk{Eauth_keyi {MD*} || SI} device analyzing multiple passive user information such as IP address,
If MD* matches MD,
then user is legitimate DScloud device location, user biometrics, signal to noise ratio, and beha-
vioral features of users (Aboudagga et al., 2005). The context-
E {MD }
DEC DEC MD* MD ENC based methods are more autonomous compared to the identity-
based method because the authentication procedure is completed
#CF SI with the minimal user interaction. However, the usability of
authentication mechanism can be improved by reducing user
TK PRNG
involvement during the authentication procedure; the accuracy of
Hash Pkpriv_cloud context-based method is lower than identity-based methods
Seed
(Cloud Private Key) because authentication procedure depends on the accuracy of the
User ID XOR Hash {Password} result of analyzing user pattern information. Furthermore, ana-
lyzing the MCC user private information such as location, bio-
metrics, calling pattern, and web searching pattern, which can be
MDcloud used to improve the accuracy of context-based method, increases
the privacy issue. In contrast to the accuracy and privacy issues,
MDuser
context-based methods can get some benefits from the various
#CF ||user ID||hash {Password} ||MDcloud ||MDuser} smartphone capabilities in MCC environment for both capturing
Fig. 5. The mobile device authentication procedure in MDA.
required user information and processing the retrieved data to
improve accuracy.
mobile device. The mobile device decrypts the cloud server mes-

Cloud-ready biometric: Omri et al. (2012) proposed to use user
sage to check the authenticity of the cloud server. If the MD
handwriting as an authentication factor to access the cloud
matches with mobile device MD, then the cloud server is
securely. The mobile user writes his password manually using
authenticated. The details of the procedure are shown in Fig. 6.
his smartphone touch screen and sends the image to cloud server
In MDA method, a secure authentication scheme based on
to be check the validity of password. There are two criteria to
message digest is proposed. Furthermore, most of security and check authentication of users, first the unique handwriting of the
privacy criteria that must be achieved to propose a suitable user and the second is the password. In the proposed method, the
authentication algorithm are applied. One of the important connection between the cloud and the mobile phone is established
security criteria is mutual authentication that is achieved in this by a Hadoop server. The architecture of the system is described in
method by authenticating both mobile device and the cloud server Fig. 7.
 M. Alizadeh et al. / Journal of Network and Computer Applications 61 (2016) 59–80
Cloud Server
Smartphone
Decryption
3456
Encryption Image Processing
Fig. 7. Handwriting-based recognition system for MCC.
The uniqueness of biometrics features is useful beneficial in
improving the security of different authentication methods;
however, some usability and privacy issues are risen by using of
these features. Unlike usual methods, some biometrics such as
fingerprint and iris, which are more confidential, advised not to be
used as authentication credentials to preserve user privacy.
Moreover, the privacy risk for handwriting is lower than other
biometrics; however, the accuracy of using handwriting is low as
well. It is recommended that low accuracy authentication metrics
such as handwriting can be applied to other methods such as
using ID and Password together, if handwriting authentication fails,
the system can ask for other methods. Furthermore, MCC cap-
abilities are not considered to improve performance of the
method, and the mobile phone should process resource-intensive
handwriting image encrypting. Additionally, some other security
enforcements such as mutual authentication, and anonymity can
be considered to make authentication method more secure against
different attacks.
3.2. User-side authentication methods
In user-side authentication methods, most of the authentica-
tion steps are processed in mobile devices. In the recent years, the
mobile devices' functionalities have been improved rapidly, hence
these mobile devices are capable of running resource-intensive
applications. Furthermore, the mobile devices are equipped with
high-technology sensors that can be used to capture and analyze
different inputs. In the user-side methods, these capabilities of
mobile devices are utilized to check authenticity of mobile users.
Compared to the conventional PCs, smartphones are able to use
a network through several paths and remain connected all the
time. However, this also presents huge threats to users of smart-
phones. Additionally, there is the serious issue in case of losing or
misplacing the mobile device, which means the loss of valuable
user confidential information (Park et al., 2011).
Transferring resource-intensive processing tasks to the cloud as
one of the main MCC goals contrasts with the processing of the
authentication mechanism inside the mobile devices in user-side
authentication methods, which makes the user-side approaches less
efficient and secure for cloud-connected mobile devices compared to
cloud-side methods. Furthermore, most of the required information
for authentication are stored in the mobile device in user-side
authentication methods, which increases the user privacy risk in
case of device robbery or loss.
67
3.2.1. Identity-based authentication methods
Similar to identity-based methods in cloud-side, authentication
methods that use user identities; however, mobile device pro-
cesses and analyzes user attributes to check user authentication
instead of cloud servers. The private user identities such as bio-
metrics are stored locally in the mobile device during authenti-
cation procedure in user-side identity-based authentication
mechanisms, which increase the privacy issues, especially in case
of loss or stolen mobile device.

Fuzzy vault, digital signature, and zero-knowledge combination:
Schwab and Yang (2013) proposed an authentication scheme
called FDZ to authenticate mobile device in cloud computing
environment. In this method, entity authentication for mobile
device's user who wants to connect to the cloud server is
provided. This proposed method is based on zero-knowledge
authentication, digital signature, and fuzzy vault. Firstly, the
secure encrypted channel between the mobile device and the
server is created, then entity authentication will be processed.
The authors used Diffie–Hellman key exchange protocol to
create a shared Advanced Encryption Standard (AES) session
key. In addition, RSA key pairs are used to protect the Diffie–
Hellman key exchange against attacks especially man-in-the-
middle attack, and fuzzy password are used to avoid some
drawbacks of traditional password. This approach is resistant to
some of the popular security treats and attacks such as
impersonation, loss of device, man-in-the-middle, and reply
attacks. Asymmetric RSA key pairs and a fuzzy picture password
system are used respectively to authenticate the client and
mobile device's user by the server. Finally, a secure encrypted
channel between mobile devices and the cloud servers is
created using Diffie–Hellman (DH) key exchange, after complet-
ing authentication steps. The different steps of proposed
method are shown in Fig. 8.
In this method, Diffie–Hellman key exchange protocol is used
to provide key management between server and the client. The
difficulty of the discrete logarithm problem makes this key
exchange protocol more secure (Forouzan, 2008). It is assumed
that an attacker can easily sniff R1 and R2, even in this case,
finding x and y is computationally difficult. There is no existing
solution to solve the discrete logarithm problem since the Diffie–
Hellman protocol was introduced. The authors proposed to choose
more than 300 decimal for the prime p, and to do not reuse x and y
values. Furthermore, this prime p should be selected in the way
that (p  1) has at least one prime factor, which must be greater
than 60 digits. One of the attacks against Diffie–Hellman protocol
is a man-in-the-middle attack; however, in this scheme, RSA keys
is used for digital signatures to protect the protocol against this
attack. The proposed protocol provides encryption services using
AES, which uses large key sizes such as 128, 192, or 256 bits to
prevent a brute force attack.
As analyzing authentication approaches is the main goal of this
paper, we focus on the authentication part of this method. For
authentication, fuzzy password system is applied, where the
mobile user can select seven images among provided images. If
the user selects five images correctly, he will be granted to access
to server resources; in other cases, the server will reject the
authentication request. The encryption system is provided using
the AES encryption algorithm that is a secure encryption algorithm
against brute force attack.
However, the proposed method uses some security algorithms
that make the method more secure against various threats; it has
some drawbacks. First of all, the details of the graphical password
system are not specified, which make security analysis more dif-
ficult. Moreover, the proposed graphical password is not secure
68 M. Alizadeh et al. / Journal of Network and Computer Applications 61 (2016) 59–80

Amazon Cloud
Android (Client)
(Server)

R1
a. Calculate R2 = g y mod p a. Calculate R1 = gx mod p
2 1
b. Calculate Session Key K b. Send R1

a. Send R2
b. Sign R2 With RSAcli- R2, Enc (Sign(R2)) a. Calculate Session Key K
3 4
c. Encrypt with K b. Verify Android Signature
d. Send Signed/Encrypted R2

a. Sign R1 With RSA serv-,

Enc (Sign(R1))
6 Verify Cloud Signature 5 b. Encrypt with K and
c. send Signed/Encrypted R1
K=AES Session Key, RSA=Asymmetric Key, -=Private Value

Fig. 8. Entity authentication protocol.

enough, hence more secure graphical password can be proposed Grid DATA
to improve the security of authentication part. Additionally, the
method needs more computing power to process different steps of
this authentication method. Various cryptography algorithms such
as RSA key exchange, AES encryption, Diffie–Hellman key
exchange are used in this approach, which need more computa-
tional resources.

QR code-based: According to Oh et al. (2011), the QR code, which

is typically a 2-dimensional (2D) code can be used for the
authentication scheme in the MCC. “Quick Response” or QR is a
form of the matrix that allows quick decoding, utilizing a form
of mass storage of high density and uses the Reed–Solomon
error correction. Typically, the IDs and passwords, certificates of
authentication, and One-Time Password (OTP), which are the
most commonly authentication methods, increase the network
traffic even if it is only a one-time process of user authentica-
tion. In QR code-based scheme, users are able to authorize a
whole new set of information in contrast to information using a
new form of data that has three types of QR code by changing
the user's information to three different versions of QR code and
Circulation Loop Method
keeping all the QR codes in a distributed format in the cloud Sequential Save
server in circulated method. This method would be appropriate
to show the necessary information for authentication of the
user via its strong points such as compressibility, error correc-
tion capability, and high data integration.

The image, ID, and password of mobile user are converted to QR

code to implement this kind of code. After creating this code, QR
will be stored in one divided cell, which shapes as a grid, and it
will be used as an authentication certification for using different
services of cloud servers as shown in Fig. 9.
This protocol uses fuzzy password system using the similar
method used by Schwab and Yang (2013), for authenticate the mobile
user. Fuzzy password is a usable method for authentication because it
reduces the difficulty of remembering a password with enough
length and randomness of password. The user can select seven
images for getting access to the server, and if he chooses five images
out of seven images correctly, he will be granted to access to the
resources. Each image is defined as a number between 1 and 255 by
the system, which makes 2557 possible combination of password.
The protocol is resistant to different attacks such as man-in-
middle, replay, and sniffing attacks. For protecting system against
man-in-the-middle, both client and server use digital signature,
then send information for both party authentication. Replay attack
is impossible because the Diffie–Hellman key exchange generates
a secure channel randomly. Sniffing attack is avoided using a
temporary session key, which encrypts the exchanged data
Fig. 9. The grid shaped QR Code for storing distributed server storage.
between the server and the client. Finally, using a visual password
for user authentication makes the protocol resistant to imperso-
nation attack.
In Oh et al's method (Oh et al., 2011), the network traffic is
reduced by utilizing QR codes as an authentication system in
mobile cloud environment. The evaluation of security of this
method is difficult because the details of the proposed authenti-
cation method are not provided.

SeDiCi 2.0: Grzonkowski et al. (2011) proposed the SeDiCi 2.0 pro-
tocol, which is another form of Zero Knowledge Proof (ZKP)
technique. This technique provides mutual authentications, which
are supposed to be more secure when it comes to phishing attack
as compared to the present system of using third party protocols.
SeDiCi 2.0 (Grzonkowski, 2010) is part of the protocol known as
the TTP (Third Trusted Party) protocol, which uses the ZKP
technique. The main goal is to provide an improved solution for
 M. Alizadeh et al. / Journal of Network and Computer Applications 61 (2016) 59–80
phishing attempts by offering mutual authentication, where users
do not have to disclose their passwords at each of the websites
that they visit. The user runs his authentication on the browser
that domain is controlled Third Trusted Party (TTP), and can login
to the system if the name of a service is on the trusted list. Fig. 10
shows the different steps of SeDiCi 2.0 authentication protocol.
There are three parties that are participating in the protocol,
including Service (S), Authentication Service (AS), and Client (C).
The client communicates with both authentication and con-
sumer services to start authentication procedure. The same policy
of typical web-browsers is applied in the case of using web-based
applications. In SeDiCi 2.0 protocol, a plug-in-based implementa-
tion is utilized to allow the application to bypass the browsers'
policy. The URI or other identifiers is required to find user location.
The URI has two useful characteristics; it contains authentication
service and user name, and it is also globally unique. Furthermore,
the users are required to have control over the authentication
domain, which can be considered as a second factor for authen-
tication (Adida, 2007). The user never type password at his visited
websites, which is the only revealed information in login step. On
the other hand, if malicious servers obtain the login information,
the adversary will attack the user. One of the SeDiCi 2.0 protocol
advantages is that the physical token is not required. Nonetheless,
a plug-in is required in the case of utilizing the user browser,
which overwrites the standard websites' security mechanisms
because a web-browser communicates with external services in
this way.
3.2.2. Context-based authentication methods
The context-based methods in the user-side analyze user
behavior features, similar to the cloud-side methods. The only
difference between cloud-side and user-side context-based
authentication method is that the mobile device processes and
evaluates user information instead of cloud server.
Typically, a context-based authentication mechanism needs
more computation power compared to the identity-based meth-
ods, and processing these kinds of resource-intensive mechanisms
by mobile devices, introduces performance issues due to resource
limitations of mobile devices. Therefore, context-based user-side
authentication methods are less appropriate in MCC compared to
cloud-based methods. Furthermore, in context-based methods,
various kinds of users' sensitive information are stored inside
Fig. 10. SeDiCi 2.0 authentication protocol.
69
smartphones that increases the user privacy risk due to device loss
compared to more reliable cloud environment.

TrustCube: Chow et al. (2010) proposed an extended version of
their previous authentication method known as TrustCube
(Song et al., 2009) by integrating the implicit authentication
(Shi et al., 2011) to perform mobile client authentication (both
the initial method and extended method feature common name
of TrustCube). TrustCube is a cloud-based authentication solu-
tion that is policy-based and utilizes an open standard. It also
supports the combination of different authentication methods
for the sake of robustness and adaptability. The policy-based
authentication has several unique advantages such as the
utilization of policies that are user-specific and finely grained,
which can be immediately updated according to users' prefer-
ences. In addition, TrustCube uses a framework with federated
authentication, more similar to the OpenID; the algorithms of
the implicit authentication are not specified, and the top-level
system description is provided. This system is developed with
an implicit authentication, which utilizes mobile data such as
SMS messages, calling logs, location, and website accesses, in
the current mobile environment. The mobile device constraints
in input requirements make using complicated passwords more
difficult, and this leads to select short passwords and PINs,
which has the higher rates of security risks.
TrustCube supports a broad range of policies, including the
platform, device's runtime environment, and user. The author
called this method as TrustCube because of using the wide range
of policies. Fig. 11 shows the architecture of this framework.
In TrustCube scheme, the Trusted Network Connect (TNC)
protocol is used for authentication between the authentication
server and the smartphone; the OpenID protocol is used to
redirect service requests to the integrated authentication service.
The Android is used for developing client agent because it can run
a background monitoring service, and this ability is critical for data
collection of implicit authentication.
After receiving the mobile client's request by the web server,
the request details are redirected to the integrated authenticated
(IA) service. Subsequently, the IA retrieves the policy for each
access request, and then sends it to the client device after
extracting the needed information through the trusted network.
The client device generates the report, and sends it back to IA
server. Finally, the authentication roles are applied, and authenti-
cation results are determined by the IA, and the authentication
result will be sent to the web server.
The usage patterns of users include the calling pattern, SMS
activity, website access, and location need to be stored during the
implicit authentication method; however, it leads to privacy issues
for the users; the problem can be solved using a suitable encryp-
tion method. The proposed system is compatible with different
authentication methods from passwords to TCG-style integrity
measurements. Furthermore, the proposed system can support
cloud-oriented authentication methods because of its flexibility.

Securing mobile cloud computing using biometric authentication
(SMCBA): In 2014, Al Rassan and AlShaher (2014) proposed an
authentication algorithm based on fingerprint. In this method,
the fingerprint image is captured by existing mobile device
camera, which does not need to implement sensors in the
mobile device. The whole process of capturing and matching
fingerprint is hosted on the cloud server to take all benefits
from cloud. The main idea of this method is alike to other
normal finger recognition methods that use mobile device
camera to capture fingerprint. The procedure of this method is
described in Fig. 12.
70 M. Alizadeh et al. / Journal of Network and Computer Applications 61 (2016) 59–80
The procedure is initiated with the capturing fingerprint image
to be processed on the cloud server. After capturing, the pre-
processing of the image is applied to convert RGB to gray-scale
image and other steps such as reducing the blur effect, ridge
enhancement, and segmentation are completed. This pre-
processed image is sent to feature extraction phase, and in the
final phase, the server checks the similarity of the extracted fea-
tures to store information of user fingerprint.
The privacy issues of using biometrics introduce the require-
ment of applying privacy preserving approaches. In a similar
situation, some cryptographic algorithms should be applied to the
captured image by the mobile device before sending it to the cloud
server, however, the fingerprint image is sent in plain text in this
method. Furthermore, the details of utilizing MCC processing and
storage resources are not clearly explained in this approach, and
the fit utilization framework for MCC is advised to be designed. In
the other word, the adaptability to MCC is not clearly defined in
this method. In addition, the accuracy of fingerprint that is
Integrated Authentication Service Implicit Authentication Service
Policy Engine
TNC Server Database
Web Server Mobile Client
Fig. 11. The TrustCube architecture.
Fingerprint Image
Pre-processing Image
Core-point Detection
Enrolment
Feature Extraction
No
Matching Database
Yes
User is Accepted
Fig. 12. Fingerprint recognition procedure in MCC.
captured by mobile device camera, is lower than using sensors
to capture the fingerprint images; therefore, it is recommended to
add other authentication factors such as using ID and Password to
this method.

NemoAuth: Le et al. (2013) proposed an authentication method
called NemoAuth based on the mnemonic multimodal approach.
NemoAuth utilizes different mobile device sensors such as gyro-
scopic, gravity, orientation, proximity, pressure, ambient light,
temperature, touch screen, and moisture sensors as well as other
facilities such as microphone and camera to measure and extract
the biometric features of mobile device user. In general, the
dynamic knowledge and biometric based approaches are combined
to improve accuracy of authentication method in NemoAuth.
The procedure of NemoAuth is similar to biometric based
methods that predefines and trains user's signature profile during
system setup step. The user's signature includes a set of multi-
modal signatures, and each signature is composed of a set of
mnemonic and atomic motions. The atomic actions that associate
with the mnemonics help users to memorize the secret keys more
IA Engine conveniently. There are varied types of atomic actions that can be
utilized according to types of mobile device sensors. As an
example, the set of atomic actions for touch screen can be taped,
Database
line, hold, circle, and cross, and a mobile user can use a fingertip to
tap at specific position or hold the fingertip for certain duration on
the mobile screen that shows the mnemonic image. Fig. 13 shows
an example of using mnemonic image to ease atomic actions that
is tap or hold a determined position of the mobile screen. In
Fig. 13, the mnemonic image is composed of 16 elements, and each
element is located at a determined position of the mobile screen.
There is no need to remember the position of image that user
wants to tap or hold for certain duration of time, because the user
can just remember the memorized elements of mnemonic image.
Furthermore, the user can select desirable signature profile
according to preferable level of security and usability. In addition,
each signature profile consists of a set of duple that shows the kind
of authentication method and the trigger time. The user can set
signature profile to use different authentication methods in the
different period of the day; for example, the mobile device can
automatically enable voice signature during non-bed time and GPS
authentication at home.
The main objective of the NemoAuth is to utilize different
capabilities of the mobile device to improve the usability of
authentication by using mnemonic images. However, this method
simplifies remembering a password for users and provides dif-
ferent options according to mobile device capacities, but the per-
formance and accuracy of authentication are in question because
the performance metrics such as False-Acceptance Rate (FAR),
False-Rejection Rate (FRR), Relative Operating Characteristic (ROC),
and Crossover Error Rate (CER) are not evaluated in this study.
Furthermore, applying a multi-modal method needs enough pro-
cessing and storage power that can be provided by the cloud
server; however, the framework to transfer these intensive pro-
cessing steps is not provided. The suitable algorithm to transfer
intensive processing phases to cloud can be designed to improve
performance.
3.3. Evaluation criteria for authentication in MCC
In this section, the criteria that are used to evaluate authenti-
cation schemes in MCC are discussed. We focus on five critical
criteria, including, usability, efficiency, security and robustness,
privacy, and adaptability to MCC environment. Furthermore, each
criterion is defined based on different sub-criteria for precise
 M. Alizadeh et al. / Journal of Network and Computer Applications 61 (2016) 59–80
evaluation. The authentication methods presented in this survey
are evaluated based on the criteria explained as follows.
Mnemonic Image
Fingertip
Fig. 13. An example of atomic action using mnemonic image.
Table 3
Usability of MCC authentication schemes: pros and cons.
Scheme
Multi-factors (Jeong et al., 2013)
MDA (Dey et al., 2013)
Cloud-based biometrics (Omri et al., 2012)
FDZ (Schwab and Yang, 2013)
QR code-based (Oh et al., 2011)
SeDiCi 2.0 (Grzonkowski et al., 2011)
TrustCube (Chow et al., 2010)
SMCBA (Al Rassan and AlShaher, 2014)
NemoAuth (Le et al., 2013)
71
3.3.1. Usability
Usability in authentication refers to “the extent to which a
product can be used by specified users to achieve specified goals with
effectiveness, efficiency and satisfaction in a specified context of use”
ISO (1998). Accordingly, several criteria are introduced to evaluate
the usability of existing authentication schemes such as effortless
memorization, fine-grained protection, and easy-to-learn (Craw-
ford et al., 2013; Braz and Robert, 2006; Bonneau et al., 2012). The
effortless memorization means that no need to remember any
secrets, and to make the procedure of authentication clear and
easy for users, which means the authentication method is easy-to-
learn. In addition, the security level of authentication procedure is
tunable based on the users preferences in fine-grained protection.
The usability as one of the most important criterion of authenti-
cation can be preserved to increase the acceptance rate of method
by end-users. Interaction in MCC authentication is performed
through mobile devices, which have some limitations such as the
small size of screen and small keyboard that introduces some
difficulties for users to input required information to complete
authentication procedure. The usability limitations and benefits of
existing authentication methods are summarized in Table 3.
3.3.2. Efficiency
Efficiency in MCC authentication is achieved when the provi-
sion of authentication service does not require too much resources
(Lopez et al., 2004; Abolfazli et al., 2013). In mobile cloud envir-
onment, the mobile device accesses cloud resources, however,
some parts of authentication protocol are processed in the mobile
device. Furthermore, transferring the most resource consuming
parts of authentication protocol to cloud server by designing
proper method can improve efficiency of authentication proce-
dure. The efficiency of each method is discussed in Table 4.
Pros Cons

Using voice and face recognition, which is easy to
User should memorize complicated password
provide and hard to replicate instead of complex
Using biometrics and password at the same time,
password demands more computational resources

Using just one factor for authentication simplifies the
User should memorize complicated password to
authentication procedure achieve high level of security

Inserting password using handwriting is more intuitive
Any other authentication factor is not available if the
than using tiny keyboard of the mobile devices to enter system fails to identify the handwriting image
password

4-digit handwritten password which is easy to
remember

Using graphical password, which is easier to remember
Memorizing some secrets makes the procedure more
difficult for a mobile user

No need to memorize all images
The authentication procedure is complicated to the user

The authentication is simple to the user as the
Using complicated password, which introduces mem-
authentication responsibility is carried out by TTP orizing difficulty for user

Using the fine-grained method that can be customized
Using specific mobile data patterns is required to
based on user preference improve the accuracy of TrustCube method, which is

Using implicit authentication method that does not difficult for the users
need the user interferes

Easy procedure for user is provided by no longer
Using error-prone biometric reduces the usability of the
needing to memorize passwords authentication mechanism

Applying atomic actions using mnemonics to make
User should memorize some secrets
memorizing secrets easier for user

Using biometrics is easy for the users
Using both secret and biometrics makes procedure
more difficult for the users

Using fine-grained method that can be customized
based on user situations
72 M. Alizadeh et al. / Journal of Network and Computer Applications 61 (2016) 59–80

Table 4
Efficiency evaluation of MCC authentication schemes: pros and cons.

Scheme Pros Cons

Multi-factors (Jeong
Processing authentication factors in bulk, which makes authen-
Processing several parameters such as ID/password, IMEI, IMSI, voice
et al., 2013) tication procedure faster and face recognition, make the authentication procedure more

Using powerful cloud resources to process authentication factors complicated

Each factor is processed by one VM in cloud

MDA (Dey et al.,
The authentication procedure requires just ID/password to make
The mobile device must process several steps to send an authenti-
2013) procedure faster cation request to the cloud

The number of communication messages is high because of applying
mutual authentication, which increases the authentication procedure
overhead

Cloud-based bio-
The powerful cloud server processes the handwriting image
Handwriting pattern is the error-prone method as the mobile user
metrics (Omri et instead of the mobile device to reduce the authentication time may write the same digits in different styles
al., 2012)
The authentication process is simple to the mobile user because of
using handwriting to enter the password

FDZ (Schwab and
Using graphical password can reduce authentication time
Several encryption algorithms such as RSA, AES, and Diffie–Hellman
Yang, 2013) key exchange are utilized to process authentication, which reduce the
efficiency of the authentication mechanism

QR code-based (Oh et
The network traffic is reduced using compressed codes like QR
Some time-consuming operations such as converting the user data to
al., 2011)
Only user name and password is utilized for authentication QR code, are processed by the mobile device

SeDiCi 2.0 (Grzon-
Using just password as the only authentication factor makes the
Number of communication messages are high because of using TTP
kowski et al., procedure simple
2011)

TrustCube (Chow et
No need to process data by mobile device because the IA engine
Processing several patterns such as SMS activities, calling pattern,
al., 2010) processes all the information and location, needs many computation power, make the authentication
procedure more complicated

SMCBA (Al Rassan
The authentication procedure does not require any password to
The mobile device processes resource-intensive task such as pro-
and AlShaher, reduce authentication time cessing the user fingerprint image
2014)

NemoAuth (Le et al.,
Using mnemonics, which are easier to remember than typical
Several authentication factors are utilized for user authentication,
2013) password, makes authentication mechanism faster which increases the authentication procedure time

3.3.3. Security and robustness
Security and robustness metrics are highly important criteria to
evaluate the authentication methods. These criteria show the
strength and weakness of the algorithm under different attacks in
various situations. According to our review, security criteria
include resistant to attacks such as impersonation, replay, and
man-in-the-middle attack. Impersonation attack means that an
attacker can masquerade user and access to the confidential
information (Truong et al., 2012); however, in reply attack, an
attacker can obtain the user information by replying the pre-
viously sent messages (Yoon et al., 2012). In addition, in the man-
in-the-middle attack, an attacker can access to all the messages
that are transferred between mobile device and servers, to get
intended information by analyzing these messages.
Furthermore, preferably an authentication method can provide
some security services such as untraceability, mutual authentica-
tion, and user unlinkability to have more secure protection. If an
attacker cannot trace a user using obtained messages, the untra-
ceability of the user is preserved; and if an attacker cannot link
sniffed messages to a specific user, the unlinkability of the
authentication algorithm is well-preserved (Park, 2004). In addi-
tion, mutual authentication means that both user and server
should be authenticated to each other (He et al., 2011). The
security services and limitations of current authentication
schemes are discussed in Table 5.
3.3.4. Privacy
Privacy is a significantly critical requirement in authentication
methods to ensure that the user is known only to legitimate
entities. The authentication method should protect private infor-
mation of users from eavesdropping during the authentication
procedure. Moreover, anonymity is one of the best approaches to
preserve both user and server privacy (Zeng et al., 2009; Kim et al.,
2008; Chang et al., 2009). Other privacy criteria include resistant
to phishing and eavesdropping attacks (Grzonkowski et al., 2011).
In phishing attack; an attacker collects credentials of user or server
to launch impersonation attack, and in eavesdropping attack, an
attacker can read the communication messages. The privacy issues
and provided protection mechanisms of the MCC authentication
schemes are summarized in Table 6.
3.3.5. Adaptable to MCC environment
This criterion shows how much the proposed method is sui-
table for MCC environment based on the capabilities and limita-
tions of this computing paradigm. The limitations and capabilities
of MCC are advised to be considered. For example, some authen-
tication methods are proposed based on resource limitation of
mobile devices such as PDA, laptop, cellphone. However, in MCC
environment, the mobile device can transfer resource-intensive
parts of the authentication algorithm to the cloud server to solve
its processing power limitation. In addition, the user needs enough
bandwidth to connect to the cloud server for uploading the
required data, which introduce some issues such as network
bandwidth and availability of the network.
Among different categories that mentioned above, the most
important category is compatibility of proposed methods with
MCC environment. After defining different evaluation criteria, the
 M. Alizadeh et al. / Journal of Network and Computer Applications 61 (2016) 59–80 73

Table 5
Security and robustness of MCC authentication schemes: pros and cons.

Scheme Pros Cons

Multi-factors (Jeong et
Using TLS and SSL for communication between network entities
Mutual authentication is not applied
al., 2013)
Utilizing several authentication factors

MDA (Dey et al., 2013)
Providing mutual authentication procedure
User untraceability is not considered

Hashing ID and password before transmission

Cloud-based biometrics
The uniqueness of handwriting style prevents an attacker to do any actions, even if he knows
Mutual authentication is not applied
(Omri et al., 2012) the user password
Lack of security attack-resistant
mechanisms

Mutual authentication is not applied

FDZ (Schwab and Yang,
Providing secure channel between mobile device and server using Diffie–Hellman protocol
The password space of proposed gra-
2013) phical password is not large enough

Resistant against device loss, impersonation, and MITM attack
Mutual authentication is not applied

Using AES encryption algorithm

QR code-based (Oh et
Using Diffie–Hellman protocol to prevent replay attack
Lack of security attack-resistant
al., 2011) mechanisms

Preventing impersonation attack using graphical password
The password space of proposed gra-
phical password is not large enough

Mutual authentication is not applied

SeDiCi 2.0 (Grzon-
Mutual authentication is provided
User untraceability and unlinkability
kowski et al., 2011) mechanism are not considered

TrustCube (Chow et al.,
It is difficult to impersonate the user, because an attacker must access to different user infor-
Mutual authentication is not applied
2010) mation such as calling pattern, SMS activity, website access, and location

SMCBA (Al Rassan and
Uniqueness of user fingerprint authentication
Mutual authentication is not applied
AlShaher, 2014)
User untraceability and unlinkability
mechanism are not considered

NemoAuth (Le et al.,
Using the user activity patterns, which are similar to the biometrics and cannot be replicated
Mutual authentication is not applied
2013)
User untraceability and unlinkability
mechanism are not considered

existing authentication approaches can be assessed to find the
most important issues of existing authentication algorithms.
All reviewed/evaluated authentication methods have been
presented in Table 7. These methods have been evaluated based on
security and performance evaluation criteria, which are discussed
in Section 3.3.
4. Prospective authentication algorithms in MCC
In this section, we present guidelines deemed beneficial for
designing relatively more secure, efficient, and adaptive authentica-
tion systems for MCC. The important factors that are essential to be
applied in coming MCC authentication approaches are discussed and
illustrated in Fig. 14. We classified these factors into five classes of
mobile device characteristics: (i) usability, (ii) efficiency, (iii) security
and robustness, (iv) privacy, and (v) adaptability to MCC environ-
ment. The comparison of proposed methods based on mentioned
criteria is discussed in Table 7. For ranking purpose, we deployed two
different scales based on Likert scale; firstly we use very low, low,
moderate, and high for usability, efficiency, and adaptability criteria.
Secondly, we use poor, fair, good, and very good for security and
robustness as well as privacy (shown in Table 7).
The results in this table advocate lack of adaptivity of current
MCC authentication schemes for MCC ecosystem. Furthermore,
most of these schemes are based on traditional methods that
previous researchers recommended to be used in conventional
mobile computing environment. In the other word, the proposed
schemes hardly considered capabilities and limitations of mobile
devices. The results of this comparison indicate that MCC-ready
authentication methods are required to focus more on mobile-
friendliness when exploiting cloud resources for mobile users.
4.1. Mobile device characteristics
The important features in the mobile device perspective are sto-
rage capacity, memory, battery limitation, and processing power.
Among these factors, energy considered as the critical resource since
the mobile devices consumes the considerable amount of energy to
remain connected to the Internet and also process authentication
procedure (Abolfazli et al., 2014a). In addition, the suitable authen-
tication scheme is compatible to the normal mobile devices with
typical resource capability. Furthermore, the consideration of the
structure and limitations such as processing power and battery of
current mobile devices makes designing an authentication method
more complicated, and effects selecting the security mechanisms
such as encryption type and key management scheme (Aiash and
Loo, 2013).
According to our review, the mobile device should be equipped
with the high quality camera in some of the authentication approa-
ches such as Jeong et al. (2013) and Al Rassan and AlShaher (2014) to
get better results. In the other word, the quality of the mobile device
camera affects the accuracy of authentication decision. Hence, no
need to install extra hardwares such as different kinds of sensors is
preferable in designing authentication method.
4.2. Usability preferences
Security is often recognized as one of the important barriers to
usability of authentication methods (Braz et al., 2007). The trade-off
between usability and performance is an important criterion to
74 M. Alizadeh et al. / Journal of Network and Computer Applications 61 (2016) 59–80

Table 6
Privacy evaluation of MCC authentication schemes: pros and cons.

Scheme Pros Cons

Multi-factors (Jeong et
Secure against phishing and eavesdropping attacks using TLS and SSL for communication
Privacy issues of using biometrics
al., 2013) between network entities
Anonymity of both user and server is
not preserved

Lack of providing user biometrics pro-
tection mechanism

MDA (Dey et al., 2013)
Providing user and server anonymity using XoR and hash function
The untraceability is not applied, which

Secure against phishing and eavesdropping attacks introduce privacy issue for the user

Cloud-based bio-
Protect user privacy by using just handwriting pattern instead of using other private biometrics
Anonymity of both user and server is
metrics (Omri et not preserved
al., 2012)
Lack of providing secure mechanism
against phishing and eavesdropping
attack

FDZ (Schwab and Yang,
Secure against phishing and eavesdropping attacks
Anonymity of both user and server is
2013) not preserved
QR code-based (Oh
Secure against phishing and eavesdropping attacks using Diffie–Hellman protocol
Lack of user and server anonymity
et al. (2011)) preservation
SeDiCi 2.0 (Grzon-
Not using private information preserves the user privacy
User and server anonymity mechanism
kowski et al., 2011) are not considered
TrustCube (Chow et al.,
Using OpenID to protect user anonymity
Using many private information of user
2010)
Protect user against phishing and eavesdropping attacks to process authentication

SMCBA (Al Rassan and
The mobile device processes fingerprint data, then sends the result of the authentication to the
Fingerprint as one of the most private
AlShaher, 2014) cloud server not the biometrics information to prevent eavesdropping risk information of user is not protected

Lack of providing secure mechanism
against phishing and eavesdropping
attacks

NemoAuth (Le et al.,
The NemoAuth method uses less private biometrics than typical biometrics methods
Using many private information of user
2013) to process authentication

No privacy mechanism is provided to
protect the user private information

evaluate authentication schemes. However, the authentication
method is expected to protect varied network entities against
security and privacy threats while it is sufficiently user-friendly. The
usability dramatically drops when the authentication methods are
designed with major focus on security that leads to more complexity
and less usability. Furthermore, the authentication mechanism can be
in accordance with the user preferable model to reach user satis-
faction. As an example, the user might prefer graphical password
authentication over other methods because of easiness of entering
the password by tapping the touch screen. The usability analysis of
various kinds of authentication methods is investigated in Braz and
Robert (2006), which shows the biometric authentication such as
voice recognition is more usable for the users because of inherent
characteristics of these kinds of authentication methods such as easy-
to-use and effortless-memorizing. Based on our review, the graphical
password and biometrics authentication methods are two types of
the trending methods in the field of MCC authentication.
4.3. Security and privacy
In MCC, mitigating the security and privacy issues during the
offloading procedure is a critical task because neglecting these issues
inhibits adoption of MCC services by the end-users (Khan et al., 2014).
The security and privacy issues of MCC such as revealed mobile user
location, stolen device, insecure access to rogue access points, near
field communication (NFC) hacking (Xuanxia et al., 2014; Yu and
Wen, 2013; Yu-Jia and Li-Chun, 2011; Yu and Wen, 2012; Zhou and
Huang, 2012; Itani et al., 2010; Ren et al., 2011; Weiwei et al., 2011;
Sue-Chen et al., 2011) can be mitigated by suitable authentication
methods. The user privacy is highly important, especially in MCC,
where the authentication information are stored in the cloud servers.
Among different approaches, biometric information such as
fingerprint, iris, retina, face, and voice recognition are highly vul-
nerable to privacy attacks, henceforth the proper privacy protection
techniques can be applied to the authentication methods. The mutual
authentication is an important feature that can be provided by
authentication mechanism. In the other words, mobile device can
check authentication of cloud server while the cloud server can check
mobile end-user authentication. Another important feature of
authentication method is untraceability that means if an attacker can
sniff the transferring packets, the relation between different packets
should not be revealed.
4.4. Mobility
In contrast to wired network that all the nodes are static, the
mobility is the inherent characteristic of mobile devices in MCC,
which introduces new challenges to design authentication proto-
cols (Chen and Hoang, 2013). The mobile devices rapidly roam in
different wireless networks to preserve connectivity, hence the
authentication delay should be acceptable according to the wire-
less network's standards. Furthermore, the packet loss and sig-
naling cost as two important criteria in mobile device roaming
procedure can be considered.
4.5. Support heterogeneity
Heterogeneity in MCC can be discussed in three main cate-
gories including mobile device, cloud, and wireless networks. The
diversity of infrastructure, hardware, software, architectures, and
technologies in MCC makes this environment more complicated.
Sanaei et al. (2013) explained open challenges and issues regard-
ing heterogeneity in MCC comprehensively. Mobile devices in
heterogeneous networks switch between different types of
 M. Alizadeh et al. / Journal of Network and Computer Applications 61 (2016) 59–80
Table 7
Comparison of evaluated authentication methods.
networks to preserve network connectivity, henceforth there is a
need to consider the characteristics of these heterogeneous net-
works in authentication mechanisms. As an example, when a
mobile device connects to a network, different parameters such as
security and QoS will be defined by the network as service level
agreement, and the preferences of each network differ from other
networks, hence these various requirements are recommended to
be considered.
4.6. Adaptiveness
In the cloud-based authentication methods, some parts of
mechanism are processed on the cloud server, hence a suitable
authentication methods are required to manage the load balancing
of VMs in cloud servers, among the reviewed authentication
methods in MCC, Jeong et al. (2013) designed a server to manage
load balancing. In addition, some managing services such as
resource scheduler are required, where one of the mobile device
or cloud server can be selected to process the specific steps of
authentication mechanism based on the intrinsic properties of the
entities (Sanaei et al., 2012).
5. Open challenges
This section highlights several open challenges related to
authentication in MCC that demand future research and devel-
opment efforts. Addressing these challenges can significantly
enhance authentication in MCC leading to successful adoption
of MCC.
75
5.1. Heterogeneous infrastructure
The heterogeneity is an intrinsic characteristic of MCC because
of using various wireless networks to preserve mobile device
connectivity. The variety of wireless technologies includes WLAN
and cellular, as well as the diversity of technologies of mobile
devices, including software, hardware, and architecture, make
designing authentication method more complicated and difficult,
because the authentication approach should be applicable and
comply with the network and mobile device requirements. Several
security concerns are arisen in heterogeneous networks' authen-
tication procedure such as initially authentication and handover
authentication when the mobile node is roaming between differ-
ent types of networks. Furthermore, as different communication
technologies offer different security levels, consequently, some
networks are less able to provide the required security level for
mobile nodes. The diversities of wireless networks, mobile devi-
ces, and cloud technologies are marginally considered in designing
authentication methods for MCC. It is recommended to design
multilevel authentication methods, which are adaptive to the
capabilities and limitations of various network entities. In multi-
level authentication methods, such as Thorwat and Shetty (2014),
Naik and Koul (2013) and Dinesha and Agrawal (2012), the
security level can be selected automatically and extra security
plugins can be added based on a tradeoff between sensitivity and
available resources.
5.2. Seamless handover
The authentication as a critical phase in handover procedure,
is a challenging task to achieve seamless handover in MCC
76 M. Alizadeh et al. / Journal of Network and Computer Applications 61 (2016) 59–80
Mobile Device Characteristics
Mobile architecture
Memory
CPU processing power
Battery limitations
Security and Privacy
Mutual Authentication
Untraceability
Resistant to attacks
Privacy protection
Support Heterogeneity
Mobile Devices
Wireless Networks
Cloud Servers
Fig. 14. Effective factors in designing authentication scheme in MCC.
(Gani et al., 2014). The authentication standards are defined in
different ways for various wireless technologies such as Wi–Fi,
WiMAX, ZigBee, and Cellular, which supports specific authentica-
tion protocols according to the defined architecture, software and
hardware. Furthermore, some network technologies support sev-
eral authentication protocols, which makes the appropriate
authentication selection method more challenging for the mobile
device. In addition, the protection of seamless connectivity is a
challenging task, especially in high mobility networks such as
MCC, and authentication has a key role to reduce handover latency
for seamless handover achievement. The suitable authentication
method must fulfill the acceptable latency rate requirement dis-
cussed in literature (Tolia et al., 2006), which are provided for the
mobile user in MCC.
5.3. Identity privacy
In MCC, the mobile user privacy preservation is more challenging
compared to the immobile users. In terms of user privacy, both
communication content and the user personal data need to be pro-
tected (Zhu et al., 2009). The mobility feature of the MCC user may
connect to different access points owned by several kinds of networks,
and the targeted network has to check the user authenticity during
handover procedure using his identity information. Furthermore, the
user information may be stored in parallel cloud storage to improve
performance and reliability, which increases the risk of disclosure of
the user private information (Barsoum et al., 2013). Henceforth, a
proper privacy preservation mechanism in MCC is a challenging task.
5.4. Resource scheduling
Energy conservation of the mobile device is highly important in
MCC, where the battery limitation of the mobile device is
Usability Preferences
Minimize User Effort
Infrequent-Errors
Memorizing-effortless
Mobility
Latency
Packet Loss
Signaling Cost
Adaptiveness
Cloud server manger
Resource scheduler
mitigated by providing shared pool computational processing
resources in cloud (Abolfazli et al., 2014b; Zhu et al., 2014; Ver-
belen et al., 2014; Shojafar et al., 2015). The mobile processing
power constraint demands transferring the resource-intensive
parts of the authentication mechanisms to the cloud for proces-
sing. Computing off-loading technique reduces the time of appli-
cation execution on mobile devices, which results in reduction of
power consumption. However, the energy consumption of trans-
ferring the data is not considered in most of the current studies
(Yao et al., 2013). Therefore, designing the appropriate resource
scheduling algorithm, which decides on transferring resource-
intensive parts of an authentication mechanism to the cloud, is a
challenging task that affects authentication latency, performance,
and energy consumption.
6. Conclusions
Security and privacy are among the most important issues in
MCC that decelerate development of this technology. In the core of
MCC, authentication is the most critical process to preserve
security and privacy of end-users. Although authentication is not
new in computing, it is immature in MCC due to unique features,
requirements, opportunities, and challenges existing in mobile-
cloud environments. Mobility, resource poverty, small form factor,
and pervasive usability of mobile devices on one hand, and wire-
less communication, cloud resource provisioning, computation
offloading, and heterogeneity on the other hand necessitate pro-
posing authentication mechanisms that are developed for mobile-
cloud users. Therefore, it is essential to critically analyse existing
authentication methods taking into account unique characteristics
and challenges of MCC to highlight the security and efficiency
 M. Alizadeh et al. / Journal of Network and Computer Applications 61 (2016) 59–80
issues of the current methods and suggest the future research
directions toward fully unleashing the power of mobile cloud.
In this study, we present a comprehensive survey of authenti-
cation in MCC that critically examines various existing authenti-
cation approaches to portrait the state-of-the-art and also high-
light the most important challenges as future research directions.
We present an in-depth discussion over authentication in MCC
and highlight its challenging characteristics compared to cloud
computing technology. Furthermore, the existing authentication
methods in MCC are evaluated and analyzed based on criteria such
as usability, efficiency, security, privacy and adaptability to MCC.
The current authentication methods are proposed based on tra-
ditional authentication schemes without considering adaptability
to MCC specifications. The results of the evaluation show that
some important factors such as user preferences, mobility, het-
erogeneity, mobile device characteristics, and MCC-friendliness
are highly critical to be considered when designing the future
authentication mechanisms for MCC. The results also suggest that
the most appropriate authentication method in MCC would be
hybrid adaptive methods with varied degrees of fidelity that can
be adopted depending on user location, Internet connectivity,
native resource level, and remote resource proximity. While the
most effective MCC solutions perform resource-intensive complex
authentication processes using remote resources in the cloud that
offers high fidelity, lightweight, yet robust approaches running
natively on the end-user device can be utilized in the absence of
remote cloud resources leading to lower fidelity. Such approaches
in near future, that demand future R&D efforts can autonomously
determine fidelity level considering user's context.
Acknowledgement
Authors acknowledge support from Malaysia-Japan Interna-
tional Institute of Technology (MJIIT) center at Universiti Teknologi
Malaysia, Japan Student Services Organization (JASSO), and
Kyushu University, Fukuoka, Japan.
References
Abolfazli S, Sanaei Z, Shiraz M, Gani A. MOMCC: market-oriented architecture for
mobile cloud computing based on service oriented architecture. In: 1st IEEE
international conference on communications in china workshops. Beijing,
China: IEEE Computer Society; 2012. p. 8–13. doi: http://dx.doi.org/10.1109/
ICCCW.2012.6316481.
Abolfazli S, Sanaei Z, Alizadeh MM, Gani A, Xia F, Member S. An experimental
analysis on cloud-based mobile augmentation in mobile cloud computing. IEEE
Trans Consum Electron 2014;99(1):1–9.
Abolfazli S, Sanaei Z, Ahmed E, Gani A, Buyya R. Cloud-based augmentation for mobile
devices: motivation, taxonomies, and open challenges. IEEE Commun Surv Tutor
2014a;16(1):337–68. http://dx.doi.org/10.1109/SURV.2013.070813.00285.
Abolfazli S, Sanaei Z, Gani A, Xia F, Yang LT. Rich mobile applications: genesis,
taxonomy, and open issues. J Netw Comput Appl 2014b;40(April):345–62. http:
//dx.doi.org/10.1016/j.jnca.2013.09.009.
Aboudagga N, Refaei MT, Eltoweissy M, DaSilva LA, Quisquater J-J. Authentication
protocols for ad hoc networks: taxonomy and research issues. In: Proceedings
of the 1st ACM international workshop on quality of service and security in
wireless and mobile networks. Montreal, Canada: ACM; 2005. p. 96–104.
http://dx.doi.org/10.1145/1089761.1089777.
Adida B. Beamauth: two-factor web authentication with a bookmark. In: Pro-
ceedings of the 14th ACM conference on computer and communications
security. Alexandria, USA: ACM; 2007. p. 48–57. http://dx.doi.org/10.1145/
1315245.1315253.
Ahn H, Chang H, Jang C, Choi E. User authentication platform using provisioning in
cloud computing environment. In: 3rd international conference on advanced
communication and networking. Brno, Czech republic: Springer-Verlag; 2011. p.
132–8. doi: http://dx.doi.org/10.1007/978-3-642-23312-8_16.
Aiash M, Loo J. Introducing a novel authentication protocol for secure services in
heterogeneous environments using Casper/FDR. Int J Commun Syst 2013;
27(12):3600–18. http://dx.doi.org/10.1002/dac.2561.
77
Alizadeh M, Wan HH. Challenges and opportunities of mobile cloud computing. In:
2013 9th international on wireless communications and mobile computing
conference. Sardinia, Italy: IEEE; 2013. p. 660–6.
Alizadeh M, Hassan WH, Behboodian N, Karamizadeh S. A brief review of mobile
cloud computing opportunities. Res Notes Inf Sci 2013a;12(April):155–60. http:
//dx.doi.org/10.4156/rnis.vol12.27.
Alizadeh M, Hassan WH, Zamani M, Khodadadi T, Chaeikar SS. A prospective study
of mobile cloud computing. Int J Adv Comput Technol 2013b;5(11).
Allam S, Flowerday SV, Flowerday E. Smartphone information security awareness: a
victim of operational pressures. Comput Secur 2014;42(May):56–65. http://dx.
doi.org/10.1016/j.cose.2014.01.005.
Al Rassan I, AlShaher H. Securing mobile cloud computing using biometric
authentication (SMCBA). In: 2014 International conference on computational
science and computational intelligence, vol. 1. Las Vegas, USA: IEEE Computer
Society; 2014. p. 157–61. http://dx.doi.org/10.1109/CSCI.2014.33.
Al-Zoube MA, Alqudah YA. Mobile cloud computing framework for patients' health
data analysis. Biomed Eng: Appl, Basis Commun 2014;26(02) 1450020–9.
http://dx.doi.org/10.4015/S1016237214500203.
Aminzadeh N, Sanaei Z, Ab Hamid SH. Mobile storage augmentation in mobile cloud
computing: taxonomy, approaches, and open issues. Simul Model Pract Theory
2015;50(January):96–108. http://dx.doi.org/10.1016/j.simpat.2014.05.009.
An N, Huynh C-T, Lee B, Hong C, Huh E-N. An efficient block classification for media
healthcare service in mobile cloud computing. Multimed Tools Appl 2014:1–15.
http://dx.doi.org/10.1007/s11042-014-2039-6.
Ardagna CA, Conti M, Leone M, Stefa J. An anonymous end-to-end communication
protocol for mobile cloud environments. IEEE Trans Serv Comput 2014;
7(3):373–86. http://dx.doi.org/10.1109/TSC.2014.2304717.
Avelar E, Marques L, dos Passos D, Macedo R, Dias K, Nogueira M. Interoperability issues
on heterogeneous wireless communication for smart cities. Comput Commun
2015;58(March):4–15. http://dx.doi.org/10.1016/j.comcom.2014.07.005.
Barsoum A, Hasan A, Member S. Enabling dynamic data and indirect mutual trust for cloud
computing storage systems. IEEE Trans Parallel Distrib Syst 2013;24(12):2375–85.
Bin M, Xiaofeng L, Xianzhong X. Risk-aware vertical handoff algorithm for security
access support in heterogeneous wireless networks. In: 5th international
conference on biomedical engineering and informatics. Chongqing, China:
IEEE; 2012. p. 1515–9. http://dx.doi.org/10.1109/BMEI.2012.6512937.
Bonneau J, Herley C, van Oorschot PC, Stajano F. The quest to replace passwords: a
framework for comparative evaluation of web authentication schemes. In: 2012
IEEE symposium on security and privacy. San Francisco, USA: IEEE; 2012. p.
553–67. http://dx.doi.org/10.1109/SP.2012.44.
Braz C, Robert J-M. Security and usability: the case of the user authentication
methods. In: Proceedings of the 18th international conference of the associa-
tion francophone d'interaction homme–machine. Montreal, Canada: ACM;
2006. p. 199–203.
Braz C, Seffah A, MRaihi D. Designing a trade-off between usability and security: a metrics
based-model. In: Baranauskas C, Palanque P, Abascal J, Barbosa S, editors. Human-
computer interaction. Lecture notes in computer Science, vol. 4663. Berlin, Heidel-
berg: Springer; 2007. p. 114–26. http://dx.doi.org/10.1007/978-3-540-74800-7_9.
Cachin C, Schunter M. A cloud you can trust. IEEE Spectrum 2011;48(12):28–51.
Chandra DG, Prakash R, Lamdharia S. Mobile ticketing system for automatic fare
collection model for public transport. In: 2013 5th international conference on
computational intelligence and communication networks. Mathura, India:
IEEE; 2013, p. 600–3. http://dx.doi.org/10.1109/CICN.2013.131.
Chang C-C, Lee C-Y, Chiu Y-C. Enhanced authentication scheme with anonymity for
roaming service in global mobility networks. Comput Commun 2009;32
(4):611–8. http://dx.doi.org/10.1016/j.comcom.2008.11.032.
Chen L, Hoang DB. Addressing data and user mobility challenges in the cloud. In:
2013 IEEE sixth international conference on cloud computing. Santa Clara, USA:
IEEE; 2013. p. 549–56. http://dx.doi.org/10.1109/CLOUD.2013.26.
Chen M, Ma Y, Liu Y, Jia F, Ran Y, Wang J. Mobile learning system based on cloud com-
puting. J Netw 2013;8(11):2572–7. http://dx.doi.org/10.4304/jnw.8.11.2572-2577.
Chow R, Jakobsson M, Masuoka R, Molina J, Niu Y, Shi E, et al. Authentication in the
clouds: a framework and its application to mobile users. In: Proceedings of the
2010 ACM workshop on cloud computing security workshop. Chicago, USA:
ACM; 2010. p. 1–6. http://dx.doi.org/10.1145/1866835.1866837.
Christensen JH. Using RESTful web-services and cloud computing to create next
generation mobile applications. In: 24th ACM SIGPLAN conference companion
on Object oriented programming systems languages and applications. New
York, USA: ACM; 2009. p. 627–34.
Chu C-H, Ouyang Y-C, Jang C-B. Secure data transmission with cloud computing in
heterogeneous wireless networks. Secur Commun Netw 2012;5(12):1325–36.
http://dx.doi.org/10.1002/sec.409.
Cisco T. Cisco visual networking index: global mobile data traffic forecast update,
2013–2018. Technical Report. Cisco; 2014.
Clarke NL, Furnell SM. Advanced user authentication for mobile devices. Comput
Secur 2007;26(2):109–19. http://dx.doi.org/10.1016/j.cose.2006.08.008.
Clarke NL, Furnell SM, Rodwell PM, Reynolds PL. Acceptance of subscriber
authentication methods for mobile telephony devices. Comput Secur 2002;
21(3):220–8. http://dx.doi.org/10.1016/S0167-4048(02)00304-8.
Crawford H, Renaud K, Storer T. A framework for continuous, transparent mobile
device authentication. Comput Secur 2013;39(Part B, November):127–36. http:
//dx.doi.org/10.1016/j.cose.2013.05.005.
Dey S, Sampalli S, Ye Q. Message digest as authentication entity for mobile cloud
computing. In: 32nd international performance computing and communica-
tions conference. San Diego, USA: IEEE; 2013. p. 1–6. http://dx.doi.org/10.1109/
PCCC.2013.6742784.
78 M. Alizadeh et al. / Journal of Network and Computer Applications 61 (2016) 59–80
Dijiang H, Zhibin Z, Le X, Tianyi X, Yunji Z. Secure data processing framework for
mobile cloud computing. In: 2011 IEEE conference on computer communica-
tions workshops. Shanghai, China: IEEE; 2011. p. 614–8. http://dx.doi.org/10.
1109/INFCOMW.2011.5928886.
Dinesha HA, Agrawal VK. Multi-level authentication technique for accessing cloud
services. In: 2012 international conference on computing, communication and
applications. Dindigul, India: IEEE; 2012. p. 1–4. doi: http://dx.doi.org/10.1109/
ICCCA.2012.6179130.
Dinh HT, Lee C, Niyato D, Wang P. A survey of mobile cloud computing: architecture,
applications and approaches. Wirel Commun Mob Comput 2013;13(18):1587–611.
Dong J, Han J, Liu J, Yin H. Shallow analysis of mobile learning environment under Cloud
computing. In: 2011 7th international conference on MEMS, NANO and smart sys-
tems. Switzerland, Kuala Lumpur, Malaysia: Trans Tech Publications; 2012. p.
1447–50. http://dx.doi.org/10.4028/www.scientific.net/AMR.403-408.1447.
Doukas C, Pliakas T, Maglogiannis I. Mobile healthcare information management
utilizing cloud computing and android OS. In: 2010 annual international con-
ference of the IEEE engineering in medicine and biology society. Buenos Aires,
Argentina: IEEE; 2010. p. 1037–40.
Esposito C, Ciampi M. On security in publish/subscribe services: a survey. IEEE Commun
Surv Tutor 2015; PP (99). http://dx.doi.org/10.1109/COMST.2014.2364616.
Fernando N, Loke SW, Rahayu W. Mobile cloud computing: a survey. Future Gener
Comput Syst 2013;29(1):84–106. http://dx.doi.org/10.1016/j.future.2012.05.023.
Forouzan BA. Introduction to cryptography and network security. McGraw-Hill
Higher Education; 2008.
Furnell SM, Dowland PS, Illingworth HM, Reynolds PL. Authentication and super-
vision: a survey of user attitudes. Comput Secur 2000;19(6):529–39. http://dx.
doi.org/10.1016/S0167-4048(00)06027-2.
Furnell S, Clarke N, Karatzouni S. Beyond the PIN: enhancing user authentication for
mobile devices. Comput Fraud Secur 2008;2008(8):12–7. http://dx.doi.org/
10.1016/S1361-3723(08)70127-1.
Gani A, Nayeem GM, Shiraz M, Sookhak M, Whaiduzzaman M, Khan S. A review on
interworking and mobility techniques for seamless connectivity in mobile
cloud computing. J Netw Comput Appl 2014;43(August):84–102. http://dx.doi.
org/10.1016/j.jnca.2014.04.009.
Ghazizadeh E, Zamani M, Ab Manan JL, Alizadeh M. Trusted computing strengthens cloud
authentication. Sci World J 2014;2014:17. http://dx.doi.org/10.1155/2014/260187.
Giuffrida C, Majdanik K, Conti M, Bos H. I sensed it was you: authenticating mobile
users with sensor-enhanced keystroke dynamics. In: Dietrich S, editor. Detection
of intrusions and malware and vulnerability assessment SE - 6. Lecture notes in
computer science, vol. 8550. Egham, London, UK: Springer International Pub-
lishing; 2014. p. 92–111. http://dx.doi.org/10.1007/978-3-319-08509-8_6.
Grzonkowski S. SeDiCi: an authentication service taking advantage of zero-
knowledge proofs. In: Sion R, editor. Financial cryptography and data secur-
ity. Lecture notes in computer science, vol. 6052. Springer; 2010. p. 426.
Grzonkowski S, Corcoran PM, Coughlin T. Security analysis of authentication pro-
tocols for next-generation mobile and CE cloud services. In: 2011 IEEE inter-
national conference on consumer electronics. Berlin, Germany: IEEE; 2011. p.
83–7. http://dx.doi.org/10.1109/ICCE-Berlin.2011.6031855.
Guo M-H, Liaw H-T, Hsiao L-L, Huang C-Y, Yen C-T. Authentication using graphical
password in cloud. In: 15th international symposium on wireless personal multi-
media communications. Taipei, Taiwan: IEEE Computer Society; 2012. p. 177–81.
He D, Ma M, Zhang Y, Chen C, Bu J. A strong user authentication scheme with smart
cards for wireless communications. Comput Commun 2011;34(3):367–74. http:
//dx.doi.org/10.1016/j.comcom.2010.02.031.
He D, Chan S, Guizani M. User privacy and data trustworthiness in mobile crowd
sensing. IEEE Wirel Commun 2015a;22(1):28–34. http://dx.doi.org/10.1109/
MWC.2015.7054716.
He D, Chan S, Guizani M. Mobile application security: malware threats and
defenses. IEEE Wirele Commun 2015b;22(1):138–44. http://dx.doi.org/10.1109/
MWC.2015.7054729.
Hoang DB, Chen L. Mobile cloud for assistive healthcare (MoCAsH). In: 2010 IEEE
Asia-Pacific services computing conference, APSCC. Hangzhou, China: IEEE;
2010. p. 325–32. http://dx.doi.org/10.1109/APSCC.2010.102.
Honggang W, Shaoen W, Min C, Wei W. Security protection between users and the
mobile media cloud. IEEE Commun Mag 2014;52(3):73–9. http://dx.doi.org/
10.1109/MCOM.2014.6766088.
Huang B. The study of mobile education development based on 3G technique and
cloud computing. In: 2011 international conference on uncertainty reasoning
and knowledge engineering. Kuala Lumpur, Malaysia: IEEE; 2011. p. 86–9.
http://dx.doi.org/10.1109/URKE.2011.6007847.
Hui S, Zhuohua L, Jiafu W, Keliang Z, Security and privacy in mobile cloud com-
puting. In: 2013 9th international on wireless communications and mobile
computing conference. Sardinia, Italy: IEEE; 2013. p. 655–9. http://dx.doi.org/
10.1109/IWCMC.2013.6583635.
Imielinski T, Korth H. Introduction to mobile computing. In: Imielinski T, Henry KF,
editors. Mobile computing, the Kluwer international series in engineering and
computer science, vol. 353. US: Springer; 1996. p. 1–43 [Chapter 1]. URL
http://faculty.winthrop.edu/domanm/csci411/Handouts/NIST.pdf.
ISO W. 9241-11. Ergonomic requirements for office work with visual display
terminals (VDTs). Technical Report. The international organization for stan-
dardization; 1998.
Itani W, Kayssi A, Chehab A. Energy-efficient incremental integrity for securing
storage in mobile cloud computing. In: 2010 international conference on
energy aware computing. Cairo, Egypt: IEEE; 2010. p. 1–2. http://dx.doi.org/10.
1109/ICEAC.2010.5702296.
Jana D, Bandyopadhyay D. Efficient management of security and privacy issues in
mobile cloud environment. In: 2013 annual IEEE india conference. Mumbai,
India: IEEE; 2013. p. 1–6. http://dx.doi.org/10.1109/INDCON.2013.6726077.
Jang EY, Kim HJ, Park CS, Kim JY, Lee J. The study on a threat countermeasure of
mobile cloud services. Korea Inst Inf Secur Cryptol 2011;21(1).
Jeong Y-S, Park JS, Park JH. An efficient authentication system of smart device using
multi factors in mobile cloud service architecture. Int J Commun Syst 2015;
28(4):659–74. http://dx.doi.org/10.1002/dac.2694.
Jin AH, Sangmin S, Namgi K, Byoung-Dai L. A study of secure data transmissions in
mobile cloud computing from the energy consumption side. In: 2013 interna-
tional conference on information networking. Bangkok, Thailand: IEEE; 2013. p.
250–5. http://dx.doi.org/10.1109/ICOIN.2013.6496385.
Kaewpuang R, Niyato D, Wang P, Hossain E. A framework for cooperative resource
management in mobile cloud computing. IEEE J Sel Areas Commun 2013;
31(12):2685–700. http://dx.doi.org/10.1109/JSAC.2013.131209.
Khalil I, Khreishah A, Azeem M. Consolidated identity management system for
secure mobile cloud computing. Comput Netw 2014;65(June):99–110. http:
//dx.doi.org/10.1016/j.comnet.2014.03.015.
Khan AN, Mat Kiah M, Khan SU, Madani SA. Towards secure mobile cloud com-
puting: a survey. Future Gener Comput Syst 2013a;29(5):1278–99. http://dx.
doi.org/10.1016/j.future.2012.08.003.
Khan AN, Mat Kiah ML, Madani Sa, Khan AUR, Ali M. Enhanced dynamic credential
generation scheme for protection of user identity in mobile-cloud computing. J
Supercomput 2013b;66(3):1687–706. http://dx.doi.org/10.1007/s11227-013-0967-y.
Khan AN, Kiah MLM, Khan SU, Madani SA, Khan AR. A study of incremental cryp-
tography for security schemes in mobile cloud computing environments. In:
2013 IEEE symposium on wireless technology and applications. Kouching,
Malaysia: IEEE; 2013. p. 62–7. http://dx.doi.org/10.1109/ISWTA.2013.6688818.
Khan A, Kiah MLM, Madani S, Ali M, Khan A, Shamshirband S. Incremental proxy re-
encryption scheme for mobile cloud computing environment. J Supercomput
2014;68(2):624–51. http://dx.doi.org/10.1007/s11227-013-1055-z.
Kim S, Rhee HS, Chun JY, Lee DH. Anonymous and traceable authentication scheme
using smart cards. In: International conference on information security and
assurance. Busan, South Korea: IEEE; 2008. p. 162–5. http://dx.doi.org/10.1109/
ISA.2008.52.
Ko SKV, Lee JH, Kim SW. Mobile cloud computing security considerations. J Secur
Eng 2012;9(2):143–50.
Kumar R, Rajalakshmi S. Mobile cloud computing: standard approach to protecting
and securing of mobile cloud ecosystems. In: 2013 international conference on
computer sciences and applications. Wuhan, China: IEEE; 2013. p. 663–9.
http://dx.doi.org/10.1109/CSA.2013.161.
Lane ND, Miluzzo E, Lu H, Peebles D, Choudhury T, Campbell AT. A survey of mobile
phone sensing. IEEE Commun Mag 2010;48(9):140–50.
La Polla M, Martinelli F, Sgandurra D. A survey on security for mobile devices. IEEE
Commun Surv Tutor 2013;15(1):446–71.
Le Z, Zhang X, Gao Z. NemoAuth: a mnemonic multimodal approach to mobile user
authentication. In: 2013 IEEE region 10 conference (31194). Xi'an, China: IEEE;
2013. p. 1–6.
Lei L, Zhong Z, Zheng K, Chen J, Meng H. Challenges on wireless heterogeneous
networks for mobile cloud computing. IEEE Wirel Commun 2013;20(3) http://
dx.doi.org/10.1109/MWC.2013.6549281.
Li C, Li L. Phased scheduling for resource-constrained mobile devices in mobile
cloud computing. Wirel Person Commun 2014;77(4):2817–37. http://dx.doi.
org/10.1007/s11277-014-1669-3.
Li R, Zhang Y, Wang Z, Sun X. The implementation of the travel cloud interpretation
system. In: 2011 international conference on computer science and network
technology. Harbin, China: IEEE Computer Society; 2011. p. 1449–51. http://dx.
doi.org/10.1109/ICCSNT.2011.6182238.
Li X, He J, Zhang T. A service-oriented identity authentication privacy protection
method in cloud computing. Int J Grid Distrib Comput 2013;6(1):77–86.
Lingfeng C, Hoang DB. Addressing data and user mobility challenges in the cloud.
In: 2013 IEEE sixth international conference on cloud computing. Santa Clara,
USA: IEEE; 2013. p. 549–56. http://dx.doi.org/10.1109/CLOUD.2013.26.
Liu L, Moulic R, Shea D. Cloud service portal for mobile device management. In:
2010 IEEE 7th international conference on e-business engineering. Shanghai,
China: IEEE; 2010. p. 474–8. http://dx.doi.org/10.1109/ICEBE.2010.102.
Liu J, Ahmed E, Shiraz M, Gani A, Buyya R, Qureshi A. Application partitioning
algorithms in mobile cloud computing: taxonomy, review and future directions.
J Netw Comput Appl 2015;48(February):99–117. http://dx.doi.org/10.1016/j.
jnca.2014.09.009.
Lopez J, Oppliger R, Pernul G. Authentication and authorization infrastructures
(AAIs): a comparative survey. Comput Secur 2004;23(7):578–90. http://dx.doi.
org/10.1016/j.cose.2004.06.013.
Louk M, Lim H, Lee H. An analysis of security system for intrusion in smartphone
environment. Sci World J 2014;2014:1–12. http://dx.doi.org/10.1155/2014/
983901.
Ma RKK, Wang C-L. Lightweight application-level task migration for mobile cloud
computing. In: IEEE 26th international conference on advanced information
networking and applications (AINA), Fukuoka, Japan: IEEE; 2012. p. 550–7.
http://dx.doi.org/10.1109/AINA.2012.124.
Mahalingam T, Rajan AV. Cloud and mobile computing: affordances of the 21st
century teaching and learning. In: 2013 international conference on current
trends in information technology. Dubai, UAE: IEEE; 2013. p. 125–8. http://dx.
doi.org/10.1109/CTIT.2013.6749490.
 M. Alizadeh et al. / Journal of Network and Computer Applications 61 (2016) 59–80
Mansoor N, Muzahidul Islam AKM, Zareei M, Baharun S, Wakabayashi T, Komaki S.
Cognitive radio ad-hoc network architectures: a survey. Wirel Person Commun
2015;81(3):1117–42. http://dx.doi.org/10.1007/s11277-014-2175-3.
Mell P, Grance T. The NIST definition of cloud computing draft. 2011.
Morrow S. Data security in the cloud. In: Rajkumar B, James B, Andrzej G, editors.
Cloud computing: principles and paradigms. John Wiley & Sons, Inc.; 2011.
p. 573–92.
Morshed MSJ, Islam MM, Huq MK, Hossain MS, Basher MA. Integration of wireless
hand-held devices with the cloud architecture: security and privacy issues. In:
2011 international conference on P2P, parallel, grid, cloud and internet com-
puting. Barcelona, Spain: IEEE; 2011. p. 83–8. http://dx.doi.org/10.1109/3PGCIC.
2011.22.
Mylonas A, Kastania A, Gritzalis D. Delegate the smartphone user? Security
awareness in smartphone platforms Comput Secur 2013;34(May):47–66. http:
//dx.doi.org/10.1016/j.cose.2012.11.004.
Naik T, Koul S. Multi-dimensional and multi-level authentication techniques. Int J
Comput Appl 2013;75(12):17–22.
Noureddine M, Bashroush R. An authentication model towards cloud federation in
the enterprise. J Syst Softw 2013;86(9):2269–75. http://dx.doi.org/10.1016/j.
jss.2012.12.031.
Oh D-S, Kim B-H, Lee J-K. A study on authentication system using QR code for
mobile cloud computing environment. In: 6th international conference on
future information technology. Loutraki, Greece: Springer-Verlag; 2011. p.
500–7. http://dx.doi.org/10.1007/978-3-642-22333-4_65.
Omri F, Hamila R, Foufou S, Jarraya M. Cloud-ready biometric system for mobile security
access. In: Benlamri R, editor. Networked digital technologies, communications in
computer and information science, vol. 294. Berlin, Heidelberg: Springer; 2012.
p. 192–200. http://dx.doi.org/10.1007/978-3-642-30567-2_16 [Chapter 16].
Owusu E, Han J, Das S, Perrig A, Zhang J. ACCessory: password inference using
accelerometers on smartphones. In: HotMobile '12, proceedings of the twelfth
workshop on mobile computing systems and applications. New York, USA:
ACM; 2012. p. 9:1–6. http://dx.doi.org/10.1145/2162081.2162095.
Pal S, Henderson T. MobOCloud: extending cloud computing with mobile oppor-
tunistic networks. In: Proceedings of the 8th ACM MobiCom workshop on
Challenged networks. Miami, Florida, USA: ACM; 2013. p. 57–62. http://dx.doi.
org/10.1145/2505494.2505503.
Park C-S. Authentication protocol providing user anonymity and untraceability in
wireless mobile communication systems. Comput Netw 2004;44(2):267–73.
http://dx.doi.org/10.1016/j.comnet.2003.09.001.
Park J, Yi K, Park J. SSP-MCloud: a study on security service protocol for smartphone
centric mobile cloud computing. In: Park JJ, Arabnia H, Chang H-B, Shon T,
editors. IT convergence and services lecture notes in electrical engineering.
Lecture notes in electrical Engineering, vol. 107. Netherlands: Springer; 2011.
p. 165–72. http://dx.doi.org/10.1007/978-94-007-2598-0_18.
Peng S, Yu S, Yang A. Smartphone malware and its propagation modeling: a survey.
IEEE Commun Surv Tutor 2014;16(2):925–41. http://dx.doi.org/10.1109/
SURV.2013.070813.00214.
Popa D, Cremene M, Borda M, Boudaoud K. A security framework for mobile cloud
applications. In: 11th IEEE conference on roedunet international conference.
Sinaia, Romania: IEEE; 2013. p. 1–4. doi: http://dx.doi.org/10.1109/RoEduNet.
2013.6511724.
Qureshi SS, Ahmad T, Rafique K, Shuja Ul I. Mobile cloud computing as future for
mobile applications—implementation methods and challenging issues. In: 2011
IEEE international conference on cloud computing and intelligence systems.
Beijing, China: IEEE Computer Society; 2011. p. 467–71. http://dx.doi.org/10.
1109/CCIS.2011.6045111.
Ra M-R, Sheth A, Mummert L, Pillai P, Wetherall D, Govindan R. Odessa: enabling
interactive perception applications on mobile devices. In: Proceedings of the
9th international conference on mobile systems, applications, and services.
New York, USA: IEEE Computer Society; 2011. p. 43–56.
Rahimi MR, Venkatasubramanian N, Vasilakos AV. MuSIC: mobility-aware optimal
service allocation in mobile cloud computing. In: IEEE sixth international
conference on cloud computing. Santa Clara Marriott, USA: IEEE; 2013. p.
75–82. http://dx.doi.org/10.1109/CLOUD.2013.100.
Rahul S, Sharda DJK. Security and privacy issues in cloud computing. Int J Eng Res
Technol 2013;2(3).
Ren W, Yu L, Gao R, Xiong F. Lightweight and compromise resilient storage out-
sourcing with distributed secure accessibility in mobile cloud computing.
Tsinghua Sci Technol 2011;16(5):520–8.
Riley M, Akkaya K, Fong K. A survey of authentication schemes for vehicular ad hoc
networks. Secur Commun Netw 2011;4(10):1137–52. http://dx.doi.org/10.1002/
sec.239.
Ruj S, Stojmenovic M, Nayak A. Privacy preserving access control with authenti-
cation for securing data in clouds. In: 12th IEEE/ACM international symposium
on cluster, cloud and grid computing. Ottawa, Canada: IEEE; 2012. p. 556–63.
Sanaei Z, Abolfazli S, Shiraz M, Gani A. SAMI: service-based arbitrated multi-tier
infrastructure model for mobile cloud computing. In: Proceedings of the IEEE
MobiCC'12. Beijing, China: IEEE; 2012. p. 14–9.
Sanaei Z, Abolfazli S, Gani A, Buyya R. Heterogeneity in mobile cloud computing:
taxonomy and open challenges. IEEE Commun Surv Tutor 2013;16(1):369–92.
http://dx.doi.org/10.1109/SURV.2013.050113.00090.
Sang-Ho S, Dong-Hyun K, Yoo K-Y. A lightweight multi-user authentication scheme
based on cellular automata in cloud environment. In: 2012 IEEE 1st interna-
tional conference on cloud networking. Paris, France: IEEE; 2012. p. 176–8. doi:
http://dx.doi.org/10.1109/CloudNet.2012.6483680.
79
Satyanarayanan M, Bahl P, Caceres R, Davies N. The case for VM-based cloudlets in
mobile computing. IEEE Pervas Comput 2009;8(4):14–23. http://dx.doi.org/
10.1109/MPRV.2009.82.
Schwab D, Yang L. Entity authentication in a mobile-cloud environment. In: 8th
annual cyber security and information intelligence research workshop: federal
cyber security R and D program thrusts. Oak Ridge, United States: ACM; 2013.
http://dx.doi.org/10.1145/2459976.2460024.
Shahzad A, Hussain M. Security issues and challenges of mobile cloud computing.
Int J Grid Distrib Comput 2013;6(6):37–50. http://dx.doi.org/10.14257/
ijgdc.2013.6.6.04.
Sheng X, Gong W. Mobility can help: protect user identity with dynamic credential.
In: 2010 eleventh international conference on mobile data management. Kan-
sas City, USA: IEEE; 2010. p. 378–80. http://dx.doi.org/10.1109/MDM.2010.73.
Shi E, Niu Y, Jakobsson M, Chow R. Implicit authentication through learning user
behavior. In: Burmester M, Tsudik G, Magliveras S, Ilić I, editors. Information
security. Lecture notes in computer science, Vol. 6531. Berlin, Heidelberg:
Springer; 2011. p. 99–113. http://dx.doi.org/10.1007/978-3-642-18178-8_9.
Shon T, Cho J, Han K, Choi H. Toward advanced mobile cloud computing for the
internet of things: current issues and future direction. Mob Netw Appl 2014;
19(3):404–13. http://dx.doi.org/10.1007/s11036-014-0509-8.
Shojafar M, Javanmardi S, Abolfazli S, Cordeschi N. FUGE: A joint meta-heuristic
approach to cloud job scheduling algorithm using fuzzy theory and genetic
method. Cluster Compute 2015;18(2):829–44.
Si P, Zhang Q, Richard YF, Zhang Y. QoS-aware dynamic resource management in
heterogeneous mobile cloud computing networks. China Commun 2014;
11(4):144–59.
Simmons GJ. A survey of information authentication. Proc IEEE 1988;76(5):603–20.
http://dx.doi.org/10.1109/5.4445.
Singh M, Singh S. Design and implementation of multi-tier authentication scheme
in cloud. Int J Comput Sci Issues 2012;9(5 5-2):181–7.
Song Z, Molina J, Lee S, Lee H, Kotani S, Masuoka R. TrustCube: an infrastructure
that builds trust in client. In: Gawrock D, Reimer H, Sadeghi A-R, Vishik C,
editors. Future of trust in computing. Vieweg þ Teubner; 2009.
p. 68–79. http://dx.doi.org/10.1007/978-3-8348-9324-6_8.
Song R, Li S, Yao Z, Li X. Design and implementation of the web content adaptation
for intelligent tourism cloud platform. In: 2012 international conference on
control engineering and communication technology. Liaoning, China: IEEE;
2012. p. 759–62. http://dx.doi.org/10.1109/ICCECT.2012.100.
Sood SK. A combined approach to ensure data security in cloud computing. J Netw
Comput Appl 2012;35(6):1831–8. http://dx.doi.org/10.1016/j.jnca.2012.07.007.
Sookhak M, Talebian H, Ahmed E, Gani A, Khan MK. A review on remote data
auditing in single cloud server: taxonomy and open issues. J Netw Comput Appl
2014;43(August):121–41. http://dx.doi.org/10.1016/j.jnca.2014.04.011.
Sue-Chen H, Jing-Yan L, Ming-Yen L. Secure cloud storage for convenient data
archive of smart phones. In: 2011 IEEE 15th international symposium on con-
sumer electronics. Singapore: IEEE; 2011. p. 156–61. http://dx.doi.org/10.1109/
ISCE.2011.5973804.
Thorwat MPD, Shetty MS. Implementation of multilevel authentication scheme for
multicloud environment. In: International conference on information and
communication technologies. Karnataka, India; 2014. p. 247–52.
Tolia N, Andersen DG, Satyanarayanan M. Quantifying interactive user experience
on thin clients. Computer 2006;39(3):46–52.
Truong T-T, Tran M, Duong A-D. Robust mobile device integration of a fingerprint
biometric remote authentication scheme. In: 26th IEEE international con-
ference on advanced information networking and applications. Fukuoka, Japan:
Institute of Electrical and Electronics Engineers Inc.; 2012. p. 678–5. http://dx.
doi.org/10.1109/AINA.2012.47.
Verbelen T, Simoens P, De Turck F, Dhoedt B. Adaptive deployment and config-
uration for mobile augmented reality in the cloudlet. J Netw Comput Appl
2014;41(May):206–16. http://dx.doi.org/10.1016/j.jnca.2013.12.002.
Wang JK, Jia X. Data security and authentication in hybrid cloud computing model.
In: 2012 IEEE global high tech congress on electronics. Shenzhen, China: IEEE;
2012. p. 117–20. doi: http://dx.doi.org/10.1109/GHTCE.2012.6490136.
Wang Y, Streff K, Raman S. Smartphone security challenges. Computer 2012;
45(12):52–8. http://dx.doi.org/10.1109/MC.2012.288.
Wang S-C, Liao W-P, Yan K-Q, Wang S-S, Tsai S-H. Security of cloud computing
lightweight authentication protocol. In: 2nd international conference on engi-
neering and technology innovation 2012. Kaohsiung, Taiwan: Trans Tech Pub-
lications; 2013. p. 3502–6. doi: http://dx.doi.org/10.4028/www.scientific.net/
AMM.284-287.3502.
Weiwei J, Haojin Z, Zhenfu C, Lifei W, Xiaodong L. SDSM: a secure data service
mechanism in mobile cloud computing. In: 2011 IEEE conference on computer
communications workshops. Shanghai, China: IEEE; 2011. p. 1060–5. http://dx.
doi.org/10.1109/INFCOMW.2011.5928784.
Xia F, Ding F, Li J, Kong X, Yang L, Ma J. Phone2Cloud: exploiting computation
offloading for energy saving on smartphones in mobile cloud computing. Inf
Syst Front 2014;16(1):95–111. http://dx.doi.org/10.1007/s10796-013-9458-1.
Xiao Z, Xiao Y. Security and privacy in cloud computing. IEEE Commun Surv Tutor
2013;15(2):843–59.
Xuanxia Y, Xiaoguang H, Xiaojiang D. A lightweight access control mechanism for
mobile cloud computing. In: 2014 IEEE conference on computer communica-
tions workshops. Toronto, Canada: IEEE; 2014. p. 380–5. http://dx.doi.org/10.
1109/INFCOMW.2014.6849262.
Xu L, Zheng X, Guo W, Chen G. A cloud-based monitoring framework for smart
home. In: 4th IEEE international conference on cloud computing technology
80 M. Alizadeh et al. / Journal of Network and Computer Applications 61 (2016) 59–80
and science. Taipei, Taiwan: IEEE Computer Society; 2012. p. 805–10. http://dx.
doi.org/10.1109/CloudCom.2012.6427534 .
Xu L, Cao X, Zhang Y, Wu W. Software service signature (S3) for authentication in
cloud computing. Clust Comput 2013;16(4):1–10. http://dx.doi.org/10.1007/
s10586-013-0262-y.
Yang SY, Hsu CL, Lee DL. An ontology-supported ubiquitous interface agent for
cloud computing-example on Bluetooth wireless technique with java pro-
gramming. In: Ninth international conference on machine learning and
cybernetics. Qingdao, China: IEEE; 2010. p. 2971–8.
Yang S, Kwon Y, Cho Y, Yi H, Kwon D, Youn J, et al. Fast dynamic execution off-
loading for efficient mobile cloud computing. In: 2013 IEEE international con-
ference on pervasive computing and communications (PerCom). San Diego,
USA: IEEE; 2013. p. 20–8. http://dx.doi.org/10.1109/PerCom.2013.6526710.
Yang S, Kwon D, Yi H, Cho Y, Kwon Y, Paek Y. Techniques to minimize state transfer costs
for dynamic execution offloading in mobile cloud computing. IEEE Trans Mob
Comput 2014;13(11):2648–60. http://dx.doi.org/10.1109/TMC.2014.2307293.
Yao D, Yu C, Jin H, Zhou J. Energy efficient task scheduling in mobile cloud com-
puting. In: Hsu C-H, Li X, Shi X, Zheng R, editors. Network and parallel com-
puting. Lecture notes in computer science, vol. 8147. Berlin, Heidelberg:
Springer; 2013. p. 344–55. http://dx.doi.org/10.1007/978-3-642-40820-5_29.
Yassin AA, Jin H, Ibrahim A, Zou D. Anonymous password authentication scheme by
using digital signature and fingerprint in cloud computing. In: Second inter-
national conference on cloud and green computing. Xiangtan, China: IEEE;
2012. p. 282–9. doi: http://dx.doi.org/10.1109/CGC.2012.91.
Yongqing S, Xiang Z. Desktop cloud-based research on unified authentication
architecture. In: 2012 spring congress on engineering and technology. Xian,
China: IEEE; 2012. p. 1–4. doi: http://dx.doi.org/10.1109/SCET.2012.6342101.
Yoon E-J, Choi S-B, Yoo K-Y. A secure and efficiency ID-based authenticated key
agreement scheme based on elliptic curve cryptosystem for mobile devices.
Comput Sci Eng 2012;8(4):2637–53.
Yu X, Wen Q. Design of security solution to mobile cloud storage. In: Tan H, editor. Knowledge
discovery and data mining- advances in intelligent and soft computing. Berlin, Heidel-
berg: Springer; 2012. p. 255–63. http://dx.doi.org/10.1007/978-3-642-27708-5_34.
X. Yu, Q. Wen, A protect solution for data security in mobile cloud storage. In: 5th
international conference on machine vision: algorithms, pattern recognition
and basic technologies. Wuhan, China: SPIE; 2013. http://dx.doi.org/10.1117/12.
2014030.
Yu-Jia C, Li-Chun W. A security framework of group location-based mobile appli-
cations in cloud computing. In: 40th international conference on parallel pro-
cessing workshops. Taipei City: IEEE; 2011. p. 184–90. http://dx.doi.org/10.1109/
ICPPW.2011.6.
Zeng P, Cao Z, Choo K-k, Wang S. On the anonymity of some authentication
schemes for wireless communications. IEEE Commun Lett 2009;13(3):170–1.
http://dx.doi.org/10.1109/LCOMM.2009.081821.
Zhang Z-h, Li J-j, Jiang W, Zhao Y, Gong B. An new anonymous authentication
scheme for cloud computing. In: 7th international conference on computer
science and education. Melbourne, Australia: IEEE; 2012. p. 896–8. doi: http://
dx.doi.org/10.1109/ICCSE.2012.6295212.
Zhang J, Liu YS, Xue L. A remote monitoring system based on measuring instrument
cloud. In: 4th international conference on industry, information system and
material engineering. Switzerland, Nanjing, China: Trans Tech Publications;
2014. p. 525–9. http://dx.doi.org/10.4028/www.scientific.net/AMR.1014.525.
Zhi-Hua Z, Jiang X-F, Jian-Jun L, Wei J. An identity-based authentication scheme in
cloud computing. In: 2012 international conference on industrial control and
electronics engineering. Xi'an, China: IEEE; 2012. p. 984–6. doi: http://dx.doi.
org/10.1109/ICICEE.2012.261.
Zhou Z, Huang D. Efficient and secure data storage operations for mobile cloud
computing. In: Proceedings of the 8th international conference on network and
service management. International Federation for Information Processing, Las
Vegas, USA; 2012. p. 37–45.
Zhu H, Lin X, Shi M, Ho P-h, Shen XS. PPAB: a privacy-preserving authentication and
billing architecture for metropolitan area sharing networks. IEEE Trans Veh
Technol 2009;58(5):2529–43. http://dx.doi.org/10.1109/TVT.2008.2007983.
Zhu X, Yang LT, Chen H, Wang J, Yin S, Liu X. Real-time tasks oriented energy-aware
scheduling in virtualized clouds. IEEE Trans Cloud Comput 2014;2(2):168–80.
http://dx.doi.org/10.1109/TCC.2014.2310452.
Zissis D, Lekkas D. Addressing cloud computing security issues. Future Gener Comput Syst
2012;28(3):583–92. http://dx.doi.org/10.1016/j.future.2010.12.006.