International Journal on [600948]

International Journal on

“Technical and Physical Problems of Engineering”

(IJTPE)

Published by International Organization of IOTPE
ISSN 2077 -3528

IJTPE Journal

www.iotpe.com

[anonimizat]
September 2015 Issue 24 Volume 7 Number 3 Pages 47 -52

47
SOFTWARE ENGINEERING AND APPLIED CRYPTOGRAPHY IN CLOUD
COMPUTING AND BIG DATA

M.I. Mihailescu 1 S.L. Nita 2

1. Information Technology and Communications Department, University of South -East Europe LUMINA , Bucharest,
Romania , [anonimizat]
2. Computer Science Department , University of Bucharest , Bucharest , Romania , [anonimizat]

Abstract – In this article we have decided to cover the
most important trends and challenges from softw are
engineering and cryptography fields by presenting some
important results achieved already (as state -of-the-art).
We will propose some new research directions, which are
absolutely necessary to be covered and focused on if we
want to move to cloud computing environment and using
big data . Due to these research directions , we have figure
out that if we don’t have a multi -disciplinary research
activity based on these two directions, the actual
technologies, such as cloud computing and big data, will
be in deep bucket and they will become useless . The
mentioned technologies, represents the peak of the spear,
in which we can actually say that we have a research
work to do.

Keywords: Cryptography , Chaos -Based Cryptography ,
Fully Homomorphic Encryption , Searchable Encryption ,
Cloud Computing , Big Data.

I. INTRODUCTION
Cloud computing represents the delivery process of
computing services over the Internet. Cloud services give
the possibility to individuals and businesses to use
software and hardware which are managed by third
parties at locations that can be accessed remotely .
Examples of cloud services online file storage, social
networking sites, webmail, and on line business
applications [10].
The cloud computing give the possibility to access
information and computer resources from any location
where a network connection is available. Cloud
computing provides a shared pool of resources, including
data storage space, networks, computer processing power,
and specialized corporate and user applications [10]. The
following definition of cloud computing has been
developed by the U.S. National Institute of Standards and
Technology (NIST):
Cloud computing is a model for enabling convenient,
on-demand network access to a shared pool of
configurable computing resources (e.g., networks,
servers, storage, applications, and services) that can be rapidly provisioned and released with minimal
management effort or service pro vider interaction. This
cloud model promotes availability and is composed of
five essential characteristics, three service models, and
four deployment models [11].
We are facing to one of the most problem from
nowadays and which will have a strong impact on the
future of encrypted search. Encrypted search is a very
sensitive subject from many aspects and it becomes an
important problem in security and cryptography. From
my point of view, this aspect is happening because of a
combination of three things:
– Searching represent the main way of accessing the data.
– Outsourcing is winning more and more faith to third
parties.
– The trust in these third parties is limited, and the reasons
are obvious.
Now, searching over encrypted data is of interest for
many ot her sub -fields of computer science, such as
cryptography, privacy, databases) but also we can find
applications in governments and industry.
A-priori searching on encrypted data could sound
impossible and even contradictory; there are several ways
of doing it. Some methods are practical, some of them are
more secure, and some of them are more functional
and/or flexible.
In 2001, for the first time the problem of searching on
encrypted data was proposed and taken into consideration
by Song, Wagner, and Perri g [1]. The authors describe a
set of schemes for the mentioned problem and provide
proofs of security for the proposed cryp to systems. The
schemes proposed by them are focused on:
– The scheme is provably secure [];
– Query isolation for searches is provided;
– Controlled searching for the untrusted server is
provided;
– Hidden queries.
Beside the advantages provided above, their schemes
are very efficient and practical. The algorithms developed
by them are simple to understand and to implement, and
also are very fast.

International Journal on “Technical and Physical Problems of Engineering” (IJTPE), Iss. 24, Vol. 7, No. 3, Sep. 201 5
48 A. Characteristics
The characteristics of cloud computing are based on-
demand self-service , broad network access, resource
pooling, rapid elasticity and measured service. On –
demand self-service is represented by the customers
(usually organi zations) that can request and administrate
their own resources used in computing . Broad network
access give the possibility for the services to be offered
over the Internet or private networks. Pooled reso urces
means that customers are able to draw from a pool of
computing resources, usually in data centers place in
other location (remotely) . Services are scaled larger or
smaller; and use of a service which is measured and
customers are billed according with the plans .

B. Service Models [10]
The cloud comput ing service models are Software -as-
a-Service (SaaS), P latform -as-a-Service (PaaS) and
Infrastructure -as-a-Service (IaaS). In a Software -as-a-
Service model, a pre -made application, along with any
required software, operating system, hardware, and
network are provided. In PaaS, an operating system,
hardware, and network are provided, and the customer
installs or develops its own software and applications.
The IaaS model provides only the hardware and network;
the customer installs or develops its own operating
systems, software and applications.

C. Deployment of Cloud Services [10]
Cloud serv ices are typically made available via a
private cloud, community cloud, public cloud or hybrid
cloud. Generally speaking, services provided by a public
cloud are offered over the Internet and are owned and
operated by a cloud provider. Some examples include
services aimed at the general public, such as online photo
storage services, e -mail services, or social networking
sites. However, services for enterprises can also be
offered in a public cloud.
In a private cloud , the cloud infrastructure is oper ated
solely for a specific organization, and is managed by the
organization or a third party.
In a community cloud , the service is shared by several
organizations and made available only to those groups.
The infrastructure may be owned and operated by the
organizations or by a cloud service provider .

II. CLOUD COMPUTING AND ITS IMPORTANCE
As we can observe, the needs of the market are very
important when we try to fix some patterns regarding the
necessities and to find the right patterns and solutions for
enterprises and not only.
The market around Google, AWS, and Microsoft will
become more and more convergence. The enterprises that
wish to move in public clouds, they have to deal with
these three companies. In the foreseeable future, the
mentioned compan ies are placing their technology bets
on cloud computing.
Anybody, besides of Google, AWS, and Microsoft,
will have to obtain a small piece from the market and
hold it as long as possible. The private cloud only, will face a fall if they will not emerge with public clouds. In
this way, after the emerging process is done, a new
hybrid or multi -cloud(s) is borne . In the next few years,
the companies that will focus on private clouds will fall
substantially.
Another important aspect is that the new providers
which offer public access on cloud will have to find very
fast a niche. The market is highly dominated by some
players who are willing to spend billions on research and
development, and marketing. Startups are focusing on
new and emerging cloud technology . The same pattern
will be observed around data storage services, analytics
services, Internet of Things (IoT), and other new
technologies. Regarding the Internet of Things (IoT) [8],
we have covered some important aspects which are
important to take into consideration when we are facing
with this concept. The paper is case study for biometrics
technologies and eLearning applications.
In Figure 1 we can observe that 58% is characterized
by a top -line growth and the collaboration among
employees, and there is also 56% for the supply chain,
three areas on which the enterprises expect cloud
computing to have a huge impact in the following three
years. Also we can see the evolution in three years (light –
blue line), and the top -line growth will have the most
powerful evolution, around 58%, and also an important
role will be played by collaboration with employees,
around 58%.

Figure 1. Evolution of the impact of cloud on business, from today and
in three years [12]

A. Software Developing in Cloud Computing
Agile Development is the best way (method) to use
when we want to develop software for cloud computing.
Cloud computing and virtualization technologies create
an easy way for agile development teams to combine

International Journal on “Technical and Physical Problems of Engineering” (IJTPE), Iss. 24, Vol. 7, No. 3, Sep. 201 5
49 multiple environments based on development, test and
production. These environments are combined with other
cloud services, a very important aspect which need to be
taken into consideration.
In the following s, we present six important ways in
which cloud computing and virtualization technologies
enhance agile software development.
1. Cloud computing gives an unlimited number of
testing and staging servers.
2. The Agile development becomes a truly parallel
activity.
3. Innovation and experimentation are encouraged.
4. The Delivery and continuous integration is enhances.
5. We have more development platforms and external
services become more and more available.
6. Code branching and merging becomes very easy to use.

B. Security in Cloud Computing
This topic is a very important one . When we wish to
have a secure cloud, security becomes a sensible
discussion which needs to give the right focusing.
The favorite topic in this area is homomorphic
encryption or fully homomorphic encryption (FHE) . FHE
is a form of encryption which give the possibility for
computations to be carried out on the ciphertext, in this
way by generating an encrypted result which, when is
decrypted , matches the result of operations perfor med on
the plaintext [1].
Because of the design, the homomorphic schemes are
malleable [2], a property of some specific cryptographic
algorithms. We say that an algorithm is malleable is it is
possible for an adversary to transform the cipher text into
another cipher text which decrypts to a related plaintext.
We suppose, given an encryption of a specific plaintext
message, it is obvious to generate another cipher text
which decrypts to f (message) , for a known function f,
without necessarily knowing or learning message.
Some examples of malleable cryptosystems need to
be listed because represents the foundation and the main
entry point of the homomorphic encryption schemes and
research that can be used in cloud computing in order to
assure the security of the data.
In a stream cipher, the ciphertext is generated by
using
 (exclusive or) on the plaintext and a
pseudorandom stream which is based on a secret key k, as
   E message message S k 
. The adversary is able to
construct the encryption of
message t for any t, as
  . E message t message t S k E message t      
In RSA cryptosystem, the plaintext message is encrypted
as
   mod eE message message n  , where
, en
represent the public key. Starting from ciphertext, the
adversary is able to construct the encryption of
message .t for any t, as
   . mod  mod   .e eE message t n mt n E message t 
.
In ElGamal cryptosystem, the plaintext message is
encrypted using
  ,  .bbE message g message A  , where
, gA
represent the public key. Starting from this ciphertext
 12,cc , the adversary is able to compute
 12,.c t c
which represent a valid encryption of
.t message
, for any t.
In 2005, we are facing with a new concept, called
non-malleable, introduced by Mare Fisclin [3], an ability
characteristic to the system to keep their non -malleable
property while giving some extra power to the adversary
in order to choose a new public key , a key that is
represented by a function based on the original public
key. So, in this case, the adversary shouldn’t be able to
come up with a cipher text whose underlying plaintext is
in connection with the original message which also take
into consideration the public keys.
Starting from the algorithms mentioned above, now
we are ready to go further with the evolution of the
security mechanisms used in cloud computing.
First modern techn ique that is used in cloud
computing is crypto cloud computing [5] which is used in
order to assure the security of the customer’s data which
relies on the security of service from the cloud computing
providers, however, the current structure of cloud
comp uting service that are provided by independent
operators.
Crypto Cloud Computing (CCC) represents one of the
newest secure cloud computing architecture. Using CCC
we have protection of information security at the system
level, and the users are allowed to access shared services
in a convenient way and also accurately. The architecture
is able to:
1. Protects individual’s connection with the rest of
world.
2. The personal privacy is protected without any delay of
the exchanging information.
The method, on which CCC is based, is known as
Quantum Direct Key (QDC) [4]. QDC represent a set of
asymmetric offline mechanism, everything regarding the
entities will go public and the private key pair is attached
to their ID. Each entity will have his o wn private key and
the entity will have also a public key generator in order to
generate any public key.
In the followings, we will present how DeTron Inc.
describes the algorithm behind the system.
QDK functions are used by generating a pair of public
and private key from two public and secret seed matrices
pM
(public) and
sM (secret). The Key Management
Center (KMC) is the single owner of
sM and it is always
offline. The private key is generated using the KMC
based on the use’s ID:
( , )cs s sKey g ID M
(1)
The public is generated any user which use a known ID:
( , )CP p pKey g ID M
(2)

C. Homomorphic Encrption
The second modern technique is represented by
homomorphic encryption [6]. Here are facing with two
categories of homomorphic cryptosystems, such as
partially homomorphic cryptosystems (PHC) and fully
homomorphic encryption (FHE) .

International Journal on “Technical and Physical Problems of Engineering” (IJTPE), Iss. 24, Vol. 7, No. 3, Sep. 201 5
50 In PHC, we have some strong examples which are
absolutely necessary to know and to understand. They
represent the main core of frameworks, such as cloud
computing architecture and QDK. In the following
examples, the E (message) denote the encryption function
of the messag e.

C.1. First Example – Unpadded RSA
In case that the RSA public key is modulus message
and exponent e, then the encryption of the message is
given by
     mod eE message message m  . The
homomorphic property is then
 1 2 1 2   .   mod eeE message E message message message m 
    1 2 1 2   mod   . .emessage message m E message message

C.2. Second Example – ElGamal
When we deal with ElGamal cryptosystem, in a cyclic
group
G of order q with generator g, if the public key is
represented as
   , , ,G q g h , where
xhg , and x
represent the secret key, then the encryption of the
message is
  ,  .rrE message g message h  , for some
random
 0, , 1 .rq   The homomorphic property
becomes
 11
1 2 1. ,  . .rrE message E message g message h 
   2 2 1 2 1 2
2 1 2 . ,  . ,  .r r r r r rg m h g message message h

  12. E message message

C.3. Third Example – Goldwasser -Micali
In Goldwasser -Micali cryptosystem, if the public key
is represented by the modulus m and a quadratic non –
residue x, then the encryption of a bit is
2  bitE bit x r mod m
, where
 0, , 1 .rm   The
homomorphic property is then
 2 1 2 2 2 1 2
1 2 1 2 1 2 1 2. . ,b b b bE bit E bit x r x r x r r E bit bit  
where . denotes addition modulo 2.
In FHE the computation are done on some operations
over the ciphertexts, such as additions, multiplications,
quadratic functions, etc). If a cryptosystem which
supports arbitrary computation on the cipher texts is
known as fully homomorphic encryption (FHE) and is
very powerful.
One of the most important examples in this direction
is represented by the Gentry’s cryptosystem, proposed by
Craig Gentry, which is using lattice -based cryptography.
The author described the first plausible construction for a
fully homomorphic encryption scheme. The scheme
proposed by Gentry, supports both addition and
multiplication operations on cipher texts , from which is
possible to develop and build circuits for performing
arbitrary computation. Another example is represented by the cryptosystem
over the integers, proposed in 2010 by Marten van Dijk,
Craig Gentry, Shai Halevi and Vinod Vaikuntanathan.
The scheme uses many of the tools of Gentry’s scheme,
but does not require ideal lattices. This scheme uses
integers.

III. BIG DATA
Big data is a compo und term for data sets so large and
complex where the classic applications are not the right
ones that are able to process the data. New challenges
include analysis, capture, data curation, search, sharing,
storage, transfer, visualization, and information privacy.
Big data can be described by the following
characteristics (we will not go into deep for each
characteristic, because is not the purpose of the article):
– Volume
– Variety
– Velocity
– Variability
– Veracity
– Complexity
In [7] is presented a survey for securing data analytics
in the cloud.
One of the most important challenge on which will
stop in this work is represented by security, information
privacy, by focusing on searchable symmetric encryption
(SSE) .
A SSE scheme is correct if the protocol used for
search returns the (current) correct results for the
keyword that is begin searched
 ()DB w , except the
negligible probability. In order to simplify the formalism
we will ignore the case when a client will attempt to add
a file with an existing identifier or delete/edit with an
identifier which is not present in DB. The protocols that
already exist can handle these situations in a clean
manner.
Let’s see an example of algorithm used in this
direction. Definition: Suppose we have
  Π ,  ,  Setup Search Update
, which represent a dynamic
SSE scheme and
represent a leakage functions. Let’s
start from two algorithms A and S, we define some
games
Π
AReal , where
  represe nt the security
parameter, and
Π
,AS Ideal (
) as following:

A.
Π
ARealλ
1A
choose DB. The game will run
 , K EDB Setup DB
and will give EDB to A. So, A
will repeat requests to engage in the protocols for Search
and/or Update. This is happening when A picks a client
input as entry. In order to respond, the game will runs the
Search or Update protocols with the input from the client
, K in
and the server input EDB and will receive a
transcript to A (the server we supposed to be
deterministic). So, A will return a bit that the game uses
as its own output.

International Journal on “Technical and Physical Problems of Engineering” (IJTPE), Iss. 24, Vol. 7, No. 3, Sep. 201 5
51 B.
 ,Π
A  S Ideal λ
1A
will choose DB. The game will run
 EDB S L DB
and will give EDB to A. In this case
the A will repeat the requests to engage in the Search or
Update protocols. This will take place when A picks a as
client input in. In order to respond, the game will give as
output of
 L in to
S . This is taken place when
S will
outputs a simulated transcript which is passed to A. So, A
will return a bit which is used by the fame as his own
output.
In order to demonstrate that
Π is L-secure against
adaptive attacks for any adversaries for A, there will be an
algorithm,
, S that exist in such way that
  ΠΠ
, Pr 1 Pr 1A A S Real Ideal neg            
(3)
In [9] we find more details about the mentioned
algorithm. In [13] the authors discuss about a set of
algorithms such as Zernike and support vector machine
(SVM) in order to detect the moving targets. This
example represents a very interesting case study for data
which c an be stored and made them available to
subscribers through a big data environment. Another
interesting aspect presented in the mentioned article
refers to reconstructing binary image, a nice example of
using Zernike moments in order to find the most suita ble
parameters when they are compared with the original
image. This idea can be used in combination with [16] in
order to create a strong authentication mechanism. The
authentication mechanism is based on face recognition
and an analysis is strongly recommended before the
authentication process is executed.

IV. CONCLUSIONS
By presenting the main trends and challenges in cloud
computing and big data, we can say that we fulfill our
goal to cover the most important technologies and
algorithms from software engineering and cryptography
that can be applied .
Our incursion in cloud computing and big data we
will not stop here. Our research will be conducted in
order to optimize and to find the best ways to develop
software and cryptographic algorithms for the mentioned
environments.
Cloud computing offers benefits for organizations and
individ uals. There are also privacy and security concerns.
If you are considering a cloud service, you should think
about how your personal information, and that of your
customers, can best be protected. Carefully review the
terms of service or contracts, and cha llenge the provider
to meet your needs [10].
The presented article addresses some important
aspects which are necessary to be take into consideration
when we will have software applications which will work
for technical and physical problems in electrical
engineering. Such examples we have seen in [13] and
[14]. ACKNOWLEDGEMENT S
We would like to show our gratitude to Assoc. Prof.
Adrian Beteringhe and Prof. Ion Sima for their support
and courage that they give it to us for writing this paper.
In the end we would like to thank Dr. Cristian
Kevorchian for his advices that he gives us in these
directions , for opening our minds and encouraging us to
explore the new challenges and research directions in
areas of cloud computing and big data.

REFERENCES
[1] “Homomorphic Encryption ”, https://en.wikipedia.org/
wiki/Homomorphic_encryption , 17 August 2015.
[2] “Malleability ( Cryptography )”, https://en.wikipedia.
org/wiki/Malleability_%28cryptography%29 , 10 February
2015.
[3] “Completely Non -malleable Schemes, Marc
Fischlin ”, Automata, Languages and Programming,
Lecture Notes in Computer Science, Vol . 3580, pp. 779 –
790, 2005 .
[4] F . Ladinois, DeTron , “Introduces its QDK
Cryptosystem to Enable True Trusted Identity for the
Cloud Era”, Posted on September 26, 2012,
https://en.wikipedia.org/wiki/Crypto_cloud_computing .
[5] “Crypto Cloud Computing ”, https://en.wikipedia.org/
wiki/Crypto_cloud_computing , 27 December 2014 .
[6] V. Vaikuntanathan, “Computing Blindfolded: New
Developments in Fully Homomorphic Encryption ”,
http://www.cs.toronto.edu/~vinodv/FHE -focs-survey.pdf .
[7] S . Yakoubov, V . Gadepally, N . Schear, E . Shen, A .
Yerukhimovich, “A Survey of Cryptographic Approaches
to Securing Big -Data Analytics in the Cloud ”,
http://www.ieee -hpec.org/2014/CD/index_htm_files/Fina
lPapers/28.pdf .
[8] P .V. Corneliu, M .M. Iulian, “Internet of Things and
its Role in Biometrics Technologies and E learning
Applications ”, 13th Internationa l Conference on
Engineering of Modern Electric Systems (EMES), pp. 1-
4, ISBN 978 -1-4799 -7649 -2, INSPEC Accession
Number: 15295927 , 11-12 June 2015 .
[9] D . Cash , et al., “Dynamic Searchable Encryption in
Very -Large Databases: Data Structures and
Implementation ”, http://www.internetsociety.org/sites/
default/files/07_4_1.pdf .
[10] V.C. Pau, M.I. Mihailescu , “Introduction to Cloud
Computing”, https://www. priv.gc.ca/resource/fs fi/02_05_
d_51_cc_e.pdf
[11] “NIST Cloud Definition ”, Version 15 , http://csrc.
nist.gov/groups/SNS/cloud -computing/ .
[12] “ 55% of Enterprises Predict Cloud Computing Will
Enable New Business Models in Three Years ”,
http://www.forbes.com/sites/louiscolumbus/2015/06/08/5
5-of-enterprises -predict -cloud -computing -will-enable –
new-business -models -in-three -years/ .
[13] E. Yakhti Fard, A. Amiri, “Finding Specific Targets
Based on Fuzzy Logic Using Zernike Moments and
Support Vector Machine ”, International Journal on
Technical and Physical Problems of Engineering (IJTPE),
ISSN: 2077 -3528 , Issue 23, Vol . 7, No. 2, pp. 60-64, June
2015.

International Journal on “Technical and Physical Problems of Engineering” (IJTPE), Iss. 24, Vol. 7, No. 3, Sep. 201 5
52 [14] S.H.R. Alemohammad, M. Saniei, E. Mashhour,
“Reconfiguration of a Distribution Network in a
Restructured Power Industry for Minimizing the Cost of
Energy Loss ” International Journal on Technical and
Physical Problems of Engineering (IJTPE) , ISSN: 2077 –
3528 , Issue 23, Vol . 7, No. 2, pp. 90 -95, June 2015.
[15] M.I. Mihailescu , “New Enrollment Scheme for
Biometric Template Using Hash Chaos -Based
Cryptography ”, Elsevier – Procedia Engineering , ISSN:
1877 -7058 , Vol. 69, pp. 1459 -1468, 2014 .
[16] M.I. Mihailescu , “Research on Biometric Synthetic
Faces ”, Indian Journal of Research (PIJ R), ISSN: 2250 –
1991 , Vol. 2, Issues 9, pp. 38 -40, September 2013 .

BIOGRAPHIES

Marius Iu lian Mihailescu was born
in Bucharest, Romania, 1985. He
received two B.Sc., first B.Sc. degree
from University of Titu Maiorescu
(Bucharest, Romania), second B.Sc.
from University of Southern
Denmark (Odense, Denmark) and the
M.Sc. degrees from University of

Bucharest (Bucharest, Romania) and Military Technica l
Academy (Bucharest, Romania). He has Ph.D. degree
from University of Bucharest (Bucharest, Romania), all
in Information Communication Technology, Computer
Science, Computer Engineering, and Information
Security, in 2008, 2009, 2010, 2011 and 2014,
respectively. Currently, he is Lecturer of Information
Technology and Communications at University of South –
East Europe LUMINA (Bucharest, Romania).

Stefania Loredana Nita was born in
Bucharest, Romania, on January 10,
1991. He received the B.Sc. in
Mathematics from University of
Bucharest (Bucharest, Romania) in
2013 and M.S.E. degrees in Software
Engineering in 2016 (expected).
Currently, he is a n Assistant Lecturer
(associate) at University of Bucharest, Romania. Her
research interests are in the application of cryptography,
programming l anguages and databases, cloud computing
and big data, applied mathematics, software development
management.