International Journal of Computer Applications (0975 – 8887)

Volume 96– No.16, June 2014

A Comparative Analysis of Encryption Techniques and

Data Security Issues in Cloud Computing

N. Hemalatha A. Jenis A. Cecil Donald L. Arockiam

M. Phil Scholar M. Phil Scholar Research Scholar Associate Professor
St. Joseph’s College St. Joseph’s College St. Joseph’s College St. Joseph’s College
Tiruchirappalli, India Tiruchirappalli, India Tiruchirappalli, India Tiruchirappalli, India

ABSTRACT two characters, namely weak confidentiality and strong

Cloud computing has been seen as the next generation confidentiality. Weak confidentiality means only authorized
architecture of IT enterprise. The cloud paradigm advantages users and cloud providers get the meaningful data from cloud
and its potential for decreasing costs and reducing the time for storage. Strong confidentiality means cloud providers cannot
a service that favours towards security issues. Cloud access the data. For example confidential business
Computing is an aggregation of IT services that offered to the information, government secret information, etc. Cloud
customer based on leasing. Though a large number of security computing provides an efficient storage for computer storage
issues are addressed, still some are not addressed and several and call back the cloud users sensitive data. Using symmetric
algorithms are proposed for security issues. This paper encryption techniques, the user’s data are encrypted before
presents a perspective of cloud computing technologies, storing into the cloud storage.
essential characteristics, classifications, delivery models and This paper is organized as follows: Section 2 explains the
various encryption mechanisms. A comparative study made Essential Characteristics of Cloud Computing and Section 3
on several encryption techniques are used for maintaining the presents the Classifications of Cloud Computing. Section 4
confidentiality in the cloud. Finally, the major data security presents Delivery Models of Cloud Computing. Section 5
issues present in cloud computing are discussed. elaborates various Encryption Techniques for Data Security in
Cloud Computing. Section 6 lists out the current Data
Keywords Security Issues in cloud computing and finally, Section 7
Cloud Computing, Confidentiality, Encryption, Data Security concludes the paper.
Life Cycle
2. ESSENTIAL CHARACTERISTICS OF
1. INTRODUCTION
Cloud Computing is a set of IT services provided by various CLOUD COMPUTING
service providers. The term cloud has originated from the  On-demand Self Service: Cloud consumers expect on
internet and cloud computing is a platform that gives people demand services from the cloud environment. The
the opportunity for sharing resources, services and service provider must allow self-service access
information among the people globally. In general, cloud according to their request. So that customers can
computing has different definitions offered by several request customized, pay and use services without the
important organizations. One of the standard definition intervention of human operators.
National Institute of Standard and Technology (NIST) [1]  Broad Network Access: Cloud resources are globally
defined cloud computing as “a model for enabling available and utilized through a standard mechanism
convenient, on demand network access to a shared pool of that promotes the user by heterogeneous platform.
configurable computing resources (e.g., Network, servers,  Resource Pooling: The providers of cloud pooled the
storage, applications and services) that can be rapidly resources together and serve multiple clients using
provisioned and released with minimal management effort or multiple tenant models.
service provider interaction”. Security and privacy are always  Rapid Elasticity: The cloud will be flexible and
the major concerns in cloud computing. According to scalable to suit consumer business needs. Consumers
Gartner’s survey, 70% of respondents think that the recent can easily add or remove users, software features,
CTO of cloud computing without the utilization of the resources, etc.
primary reasons is that there is data security and privacy  Measured Service: Cloud computing resource usage
concerns [2]. Google Apps is the best example of cloud can be measured, controlled and reported providing
computing where everyone can access their applications on transparency for both provider and consumer of the
Google data centers through a web server [3]. Security is utilized service.
categorized into two shells such as protecting the assets and 3. CLASSIFICATION OF CLOUD
protecting the data. Storing data in the cloud have many
benefits which reduce hardware cost and improve storage
COMPUTING
Based on the customer service and their usage, the cloud can
reliability. In cloud computing, data security problem was not
be classified as public, private, community, and hybrid.
resolved entirely. Some of the security parameters are
availability, authentication, confidentiality and integrity.
Users don’t know where the information will be stored in the
cloud. Most users care about the confidentiality of their
information. Confidentiality is the important parameter in all
cases of deployment. Confidentiality can be categorized into

1
 International Journal of Computer Applications (0975 – 8887)
Volume 96– No.16, June 2014

3.1 Public Cloud 4.2 Platform as a Service (Paas)

Public or external cloud infrastructure is available to the Platform as a Service means a way to rent hardware, operating
general owned by an organization for selling cloud service on system and network capacity over the internet. The platform is
pay-per-use basis. Some examples of public cloud are Google provided to the cloud customer as a service. Google App
App Engine and Windows Azure services platform. Engine [8], Microsoft Windows Azure [9] is the popular
example of paas. Advantage of paas is Google minimizing
3.2 Private Cloud developers, maintained customization and configuration.
Private or internal cloud infrastructure is maintained and
controlled by a single organization. Some examples of private 4.3 Infrastructure as a Service (IaaS)
cloud are Sun Cloud, Google App Engine, IBM Blue clouds, Infrastructure as a Service means just building a home from
etc. beginning, it consists “on demand” to internet connection,
space with the flexibility to increase or decrease server
3.3 Community cloud capacity [7]. The IaaS cloud example is Amazon EC2 [10],
Community cloud is another type of private clouds. Cloud Go grid, 3tera, etc... The advantage of IaaS is a Cloud solution
infrastructure is shared by several organizations, typically cost scales according to use.
with shared concerns. It may be managed by a group of
organizations or third party on/ off premises. 5. ENCRYPTION TECHNIQUES FOR
DATA SECURITY IN CLOUD
3.4 Hybrid Cloud Encryption Techniques are used for securing the sensitive
Cloud infrastructure consists of two or more public, private or information [11]. There are two different encryption
community clouds. The purpose of hybrid cloud is to provide techniques are used commonly.
extra services and resources to customers to serve their
demands. Symmetric Key Encryption uses Single key is involved in
data security. Both sender and receiver use the same key to
4. DEPLOYMENT MODELS IN CLOUD encrypt and decrypt.
COMPUTING Asymmetric Key Encryption uses two keys are involved.
Services are offered by cloud providers can be grouped into The receiver has a secret key which is private and another key
three categories. Figure 1 explains the cloud delivery models. public which is published to everyone.
 Software as a Service (Saas) Generally, Classical Encryption algorithms categorized into
 Platform as a Service (Paas) two principles.
 Infrastructure as a Service (IaaS)
 Substitution Cipher (replaces one character with
another)
 Transposition Cipher (transpose or reorder the
characters).
Homomorphic encryption concept origin from esoteric world
of abstract algebra. Homomorphic means same shape or same
effect on two different sets of objects can be transformed. The
fully homomorphic encryption means (FHE), that there are no
limitations on what manipulations can be performed [12].
The following section analyses the various symmetric
encryption techniques, which effectively handle large amount
of data.

5.1 Classical Substitution Cipher

Algorithm
Padmapriya et al [13] have proposed a method for providing
the inverse of Caesar Cipher that supports more security for
the data compared with earliest Caesar Cipher. The main
scope of their paper is a new level of data security solution
with encryption using ASCII full characters, which is
important for designing the complete security solution. Caesar
Cipher is replacing the letter three places down in the
alphabet. For example “COME” is a plain text which will be
converted into “FQOH” as cipher text. But this cipher can be
Figure 1. Cloud Delivery Models broken by brute force attack because at the end there are only
25 possible available options of key.
4.1 Software as a Service (SaaS)
Software is provided to the cloud customer as a service. Over Sastry et al [14] proposed playfair cipher which has encrypt
the web the SaaS applications are delivered to end users [4]. multiple letters. They used a square matrix of 5x5 alphabetical
SaaS is offered by Google Apps [5], Salesforce.com [6], letters are arranged in an appropriate manner. The user can
Social media, Gmail is software as a service of Google, etc. select the key and place it in the matrix. The user can select a
Advantages of Saas is scalable, flexible pricing, portability, key and place it in the matrix. The remaining letters of the
user-friendly. English alphabet from the key are then one by one placed in
the matrix of playfair cipher. The plain text is broken into

2
 International Journal of Computer Applications (0975 – 8887)
Volume 96– No.16, June 2014

pairs and if a pair ha same alphabet is separated by  Diffie Hellman keys in exchange.
introducing a filler letter with ‘x’.  Digital Signature used for authentication
 AES encryption algorithm is used to encrypt or
5.2 Fully Homomorphic Encryption decrypt user’s data file.
Maha tebaa et al [15] proposed an application of a method to
execute operations on encrypted data without decrypting In cloud storage environment to avoid data modification, they
them, which will provide us with the same results after used two separate servers are maintained, one for encryption
calculations as if the authors have worked directly on the raw process known as computing platform and another one known
data. Homomorphic Encryption method is able to perform as a storage server for storing user data.
operations on encrypted data without decrypting.
Homomorphic Encryption systems are used to perform
5.4 Cloud DES Algorithm
operations on encrypted data without knowing the private key Neha Jain and Gurpreet et al [19] have proposed Data
(without decryption), the client is the only holder of the secret security in cloud computing using the DES algorithm. This
key. In their paper, a new concept of security which enables Cipher Block Chaining system is to be secure for clients and
providing results of calculations on encrypted data without server. The security architecture of the system is designed by
knowing the raw data on which the calculation was carried using DES cipher block chaining, which eliminates the fraud
out, in respect of the data confidentiality. The improvement of that’s occurring today with stolen data. The data which are
the complexity of the Homomorphic encryption algorithms sent, being intercepted and replaced has no danger. The
and compares the response time of the requests to the public system with encryption is acceptably secure, but that the level
key. of encryption has to be stepped up, as computing power
increases. Results in order to be secured the communication
Huda et al [16] proposed fully homomorphic encryption system between the modules are encrypted using a symmetric
(FHE) allows a user that does not have the secret decryption key. The author proposed that the cloud data security must be
key to compute any result of the data. The author focused considered to analyze the data security risk, the data security
technique is based on a FHE algorithm with key delegation to requirements, deployment of security functions and the data
ensure data confidentiality, authentication, integrity and security process through encryption. The main contribution of
availability of multi-level hierarchical order. Their proposed their paper is the new view of data security solutions with
framework solution is the using of homomophic cryptography encryption, which is important and can be used as reference
with Attribute Based Encryption. for designing the complete security solution.

5.3 Cryptography, Encryption Technique Monikandan et al [20] have described an encryption

Sugumaran et al [17] discussed techniques that are algorithm to address the security and privacy issue in cloud
implemented to protect data and propose an architecture to storage to protect the data from unauthorized access. Data can
protect the data in the cloud. Their architecture was developed be attacked in two ways. An insider attack is an administrator
to store data in the cloud in encrypted data format using having the possibility to hack the user’s data. Outsider attack
cryptography technique which is based on a block cipher. is third party can access the user’s data. The Author proposed
Their proposed architecture is based on cryptography a symmetric encryption algorithm to protect the data stored in
algorithm, which is secure for data storage. This architecture cloud storage from the unauthorized access. Their proposed
has been designed for data security using block based technique is converting plain text into the corresponding
symmetric cryptography, having better speed for storing the ASCII code value of each alphabet and the key value ranges
data. Their algorithm improved encryption of data secured by between 1 to 256. This technique improves classical
interesting the symmetric layer and using this technology is encryption techniques by integrating substitution cipher and
efficient for storing the user’s data in the cloud. transposition cipher. Symmetric encryption has the speed and
computational efficiently to handle encryption of large
Prashant et al [18] have proposed a “Three way mechanism” volumes of data in cloud storage. Their proposed algorithm is
ensures all the three protection scheme of authentication, data used to encrypt the user data in cloud storage and it can’t be
security and verification at the same time. Their proposed accessed by administrators or attackers. The main contribution
architecture to make use of digital signature and Diffie of this paper is a comparison between various Symmetric
Hellman key exchange with AES algorithm to protect Encryption algorithms are compared based on its techniques,
confidentiality of data stored in the cloud. Their proposed description, concepts, security, issues addressed.
mechanism makes it tough for hackers to crack the security
system, thereby protecting data stored in the cloud.

Table 1. Comparison Techniques of Existing Encryption Algorithms

Author Techniques Used Description Concepts Used Security Issues

Applied On Addressed

Padmapriya et al Inverse Caesar Classical Substitution ASCII full characters Cloud Data Security
Cipher Cipher Same key used customer and and Privacy
for Encryption & (256 characters) Cloud provider
Decryption side

Sastry et al Playfair Cipher Classical substitution 5x5 matrix and Cloud Data Security
Cipher. Same key used Alphabetic characters customer and and Privacy
for Encryption & used Cloud provider

3
 International Journal of Computer Applications (0975 – 8887)
Volume 96– No.16, June 2014

Decryption side

Maha et al Fully The private key is used Cryptosystem based Cloud Data Security
Homomorphic for Encryption on fully provider side
Encryption (without Decryption) Homomorphic only
Encryption

Huda et al Fully The private key is used Electronic Health Cloud Data
Homomorphic for Encryption Records classify Customer and confidentiality,
Encryption (without Decryption) based on PI Cloud Provider Authentication,
(Personally side. Availability
Identifiable and Integrity.
information)

Sugumaran et al Block based Symmetric layer The private key Cloud Data Security
Symmetric inserted for encrypting concept is used Customer side and Privacy
Cryptography the secure data using a between sender and
symmetric algorithm receiver

Monikandan et al Classical Both Substitution and Palin text is Customer's Data Security
Encryption Transposition. Same converted to ASCII side only and Privacy
key used for code value, key range
encryption and between 1 to 256
Decryption

Neha Jain et al DES Algorithm The same key is used Cipher Block Both Cloud Data Security
for Encryption and Chaining mode customer and
Decryption Cloud provider

6. DATA SECURITY ISSUES IN CLOUD  Use: This phase uses the data in the repository for
various processes.
COMPUTING
Data Security means protecting data from unauthorized  Share: In this phase, the transmission of data occurs
access, modification or destruction. In the cloud model, between customers and the partners.
service provider is responsible for maintaining the data
security. Recently, technology has focused on only one stage  Archive: In this phase, stored data is utilized for
for a data security life cycle, which is process manage from future use.
creation to destroy. Figure 2 explains the data security life  Destroy: This phase will permanently remove the
cycle and the phases in it. data in the repository.
 Create: This phase is the generation of new digital Data security threats can be classified into internal
content in client or server in the cloud. threats and external threats. Internal threats, mainly come
 Store: This phase occurs simultaneously after the from an insider attack because cloud service providers
creation process, therefore storage of data occurs in and users, are the main reason for these threats. External
the repository or stored in multiple nodes. threats, mainly come from outside attack because data
can be accessed from third party. The attacker can steal
the user’s personal data. [21]. There are Six types of
major issues [22], while discussing data security in the
cloud.
 Data Authentication
 Data Privacy and Confidentiality
 Data Integrity
 Data Location
 Data Availability
 Data Storage, Backup and Recovery

6.1 Data Authentication

A user may gain access within a LAN by entering a cloud
identification and password, which may be affirmed by a
cloud authentication mechanism. If the authentication
mechanism validates the certification, the user identification
and password are stored locally for subsequent authentication
Figure 2. Data Security Life Cycle requests. The authentication mechanism may be applied in
both domain and Workgroup LAN and may function in

4
 International Journal of Computer Applications (0975 – 8887)
Volume 96– No.16, June 2014

parallel with other users who may have a LAN or client [3] Donald, A. Cecil, S. Arul Oli, and L. Arockiam. "Mobile
credentials which may not be authenticated from the cloud. Cloud Security Issues and Challenges: A
Perspective." International Journal of Electronics and
6.2 Data Privacy and Confidentiality Information Technology (IJEIT), ISSN (2013): 2277-
Once the clients outsource data to the cloud there should be 3754.
assurance that data is accessible to only authorized users. The
cloud computing service provider should make secure the [4] Maneesha Sharma, Himani Bansal, Amit Kumar Sharma,
customer personal data is well protected from other service “Cloud Computing: Different Approach & Security
provider’s and user. Authentication is the best solution for Challenges”, International Journal of Soft Computing
data privacy because service provider must make sure who is and Engineering (IJSCE) ISSN: 2231-2307, Volume-2,
accessing the data and who is maintaining the server; so that Issue-1, March 2012.
the customer’s personal data is protected. The cloud customer [5] Sales force. Salesforce CRM applications and software
must be guaranteed that data stored in the cloud will be solutions. http://www.salesforce.com/eu/crm/products.jsp
confidential [22].
[6] Google, Google Apps. <http://www.google.com/apps/>.
6.3 Data Integrity [7] Sarvesh Kumar, Jahangeer Ali, Ashish Bhagat, Jinendran
Data Integrity means data is complete and consistent. The data
P.K, “An Approach to Creating a Private Cloud for
stored in the cloud may suffer from damage during integration
Universities and Security Issues in Private Cloud”
operations. The cloud provider must make the client aware of
International Journal of Advanced Computing, Vol. 36,
what particular data are outsourced to the cloud, the native
Issue 1, ISSN: 2051- 0845, 2013.
and the integrity mechanisms put in place [22].
[8] Cecil A Donald and L Arockiam. Article: Securing Data
6.4 Data Location with Authentication in Mobile Cloud Environment:
The cloud users did not know where the data will be hosted Methods, Models and Issues. International Journal of
and in fact, their users want to know the location exactly. It Computer Applications 94(1):25-29, May 2014.
requires a contractual agreement between the users that data Published by Foundation of Computer Science, New
should stay in a particular location. York, USA.
6.5 Data Availability [9] Microsoft. Microsoft Windows Azure.
Data provided by the customer is normally stored in different <http://www.microsoft.com/windowsazure/>.
servers often placing in different locations or in different [10] Amazon. Amazon Elastic Compute Cloud (EC2).
clouds. Data availability becomes a major legitimate issue as <http://aws.amazon.com/ec2/>.
the availability of corrupted and relatively difficult servers.
[11] Padmapriya and Subhasri,”Cloud Computing: Security
6.6 Data Storage, Backup and Recovery Challenges & Encryption Practices”, International
The cloud users decide to move their data to the cloud Journal of Advanced Research in Computer Science and
provider should ensure adequate resilience storage systems. Software Engineering, Volume 3, Issue 3, ISSN: 2277
The process of recovering and backing up data is simplified. 128X, March 2013.
The cloud providers will store the data in several places across
many independent servers. [12] C. Gentry, "Computing Arbitrary Functions of Encrypted
Data", ACM, Vol. 53, Issue 3, March 2010, pp. 97-105.
7. CONCLUSION [13] Padmapriya, Subhasri,” Cloud Computing: Reverse
Cloud computing is a versatile technology, widely studied in
Caesar Cipher Algorithm to Increase Data Security”,
recent years. The providers and the clients must make sure
International Journal of Engineering Trends and
that the cloud is safe from all the internal threats, external
Technology (IJETT), Volume 4, Issue 4, 2013.
threats and mutual understanding between the customer and
provider when it comes to the security of cloud. The major [14] V.U.K. Sastry, N. Ravi Shankar and S. Durga Bhavani,
issues in cloud computing is data security and it has many “A Generalized Playfair Cipher involving Intertwining,
aspects like confidentiality, Integrity, surveillance, reliability, Interweaving and Iteration”, International Journal of
availability, Security, anonymity, telecommunications Network and Mobile Technologies, pp 45-53, 2010.
capacity, government and backup & recovery. But the most
important issue in data security is security and privacy for [15] Maha TEBAA, Saïd EL HAJJI, Abdellatif EL GHAZI
protecting the data in cloud storage. This paper analyses the “Homomorphic Encryption Applied to the Cloud
importance of the data security in the cloud. Reason for Computing Security” Proceedings of the World Congress
choosing symmetric encryption algorithms are effective to on Engineering, London, U.K. ISBN: 978-988-19251-3-8
handle encryption for large amount of data, and effective ISSN: 2078-0958 (Print); ISSN: 2078-0966 (Online), Vol
speed of storing data in the cloud. In this paper, table 1 1, July 4 - 6, 2012.
explains the comparison among various encryption techniques [16] Huda Elmogazy, Omaima Bamasak,” Towards
used in the cloud. Healthcare Data Security in Cloud Computing”, IEEE
8th International Conference for Internet Technology and
8. REFERENCES Secured Transactions (ICITST-2013).
[1] P. Mell and T. Grance, “The NIST Definition of Cloud
Computing”, September 2011. [17] Sugumaran, BalaMurugan. B, D. Kamalraj,” An
Architecture for Data Security in Cloud Computing”,
[2] Anwar J. Alzaid, Eng. Jassim M. Albazzaz, “CLOUD IEEE World Congress on Computing and
COMPUTING: AN OVERVIEW”, International Journal Communication Technologies 2014.
of Advanced Research in Computer and Communication
Engineering, Vol. 2, Issue 9, September 2013.

5
 International Journal of Computer Applications (0975 – 8887)
Volume 96– No.16, June 2014

[18] Prashant Rewagad, Yogita Pawar,” Use of Digital Tiruchirappalli, Tamil Nadu, India. She had attended several
Signature with Diffie Hellman Key Exchange and AES National and International Workshops and Conferences. Her
Encryption Algorithm to Enhance Data Security in Cloud area of research is Cloud Computing. Her areas of interest are
Computing”, 2013 IEEE International Conference on Networks Security, Software Programming and Web
Communication Systems and Network Technologies. Developing Technologies.
[19] Neha Jain and Gurpreet Kaur,” Implementing DES A. Cecil Donald received his Masters in Software
Algorithm in Cloud for Data Security”VSRD-IJCSIT, Engineering from Anna University, Chennai, India. He has
Vol. 2 (4), 2012, 316-321. one year experience in IT industry as a Software Developer.
Currently, he is a Ph.D. research scholar in the department of
[20] L. Arockiam, S. Monikandan,” Data Security and Computer Science at St. Joseph’s College (Autonomous),
Privacy in Cloud Storage using Hybrid Symmetric Tiruchirappalli affiliated to Bharathidasan University, India.
Encryption Algorithm”, International Journal of His main area of research is Mobile Cloud Computing. He has
Advanced Research in Computer and Communication published several papers in the International Journals and also
Engineering Vol. 2, Issue 8, August 2013. he has atteneded several national and international
[21] Xiaojun Y, Qiaoyan Wen,” A View about Cloud Data conferences and workshops.
Security from Data Life Cycle”, IEEE 2010. Dr. L. Arockiam is working as Associate Professor in the
[22] Parsi Kalpana and Sudha Singaraju, “Data Security in Department of Computer Science, St.Joseph’s College
Cloud Computing using RSA Algorithm”, International (Autonomous), Tiruchirappalli, Tamil Nadu, India. He has 25
Journal of Research in Computer and Communication years of experience in teaching and 18 years of experience in
technology, IJRCCT, ISSN 2278-5841, Vol 1, Issue 4, research. He has published more than 187 research articles in
September 2012. the International & National Conferences and Journals. He
has also presented 2 research articles in the Software
9. AUTHORS BIOGRAPHY Measurement European Forum in Rome. He has chaired many
N. Hemalatha is doing M. Phil research in the Department of technical sessions and delivered invited talks in National and
Computer Science, St. Joseph’s College (Autonomous), International Conferences. He has authored 3 books. His
Tiruchirappalli, Tamil Nadu, India. She had attended several research interests are: Cloud Computing, Big Data, Cognitive
National and International Workshops and Conferences. Her Aspects in Programming, Data Mining and Mobile Networks.
area of research is Cloud Computing. Her area of interest is He has been awarded “Best Research Publications in Science”
Networks Security, Software Engineering and Web for 2009, 2010 & 2011 and ASDF Global “Best Academic
Technologies. Researcher” Award from ASDF, Pondicherry for the
academic year 2012-13 and also the “Best Teacher in college”
A. Jenis is doing MPhil research in the Department of award for the year 2013 & 2014.
Computer Science, St. Joseph’s College (Autonomous),

IJCATM : www.ijcaonline.org 6