Enhancing Data Security in Cloud Computation Using Addition-Composition Fully Homomorphic Encryption Scheme

IOSR Journal of Engineering (IOSRJEN) www.iosrjen.org
ISSN (e): 2250-3021, ISSN (p): 2278-8719
Vol. 07, Issue 10 (October. 2017), ||V2|| PP 20-29
International organization of Scientific Research 20 | P a g e
Enhancing Data Security in Cloud Computation Using Addition-
Composition Fully Homomorphic Encryption Scheme
Omollo R.O.1
, Raburu G.O.2
, Omolo-Ongati N.3
, Okelo B.N.4
1,2
Department of Computer Science and Software Engineering
3,4
Department of Pure and Applied Mathematics
Jaramogi Oginga Odinga University of Science and Technology
P.O. Box 210-40601, Bondo-Kenya.
Abstract: -This paper focusses on the use of homomorphic encryption schemes in solving data security in cloud
computation. We started by acknowledging Gentry’s contribution on partial homomorphic encryption schemes
where he constructed a somewhat homomorphic encryption based on ideal lattices using both additive and
multiplicative Homomorphisms. We also noted improvement made on his construction by Dijkwherehe used
integers instead of ideal lattices. Again we appreciated efforts by Braskerski based on Learning with Errors
(LWE) and Ring Learning with Errors (Ring LWE) problems. Besides all these efforts, the schemes still had
strain on the computing storage and processing resources. We discovered that there is need to constructan
encryption scheme which lessens the computational strain on the computing assets. Here, we found remedy in
the use of both addition and composition operations to implement a fully homomorphic encryption scheme. A
mathematical theory was translated into an algorithm implementable JAVA. We testediton a single hardware
with minimum specifications: 1.5GHz, 4GB RAM and Windows 10 OS. The results showed that the new
scheme supports faster encryption process, larger ciphertext sizes and is versatile. This new addition-
composition fully homomorphic scheme enhances data security in cloud computation thus boosts consumer
confidence in the cloud services and applications.
Keywords: -Addition-Composition, Cloud Computing, Data Security, Fully Homomorphic Encryption,
Homomorphic Encryption Scheme.
---------------------------------------------------------------------------------------------------------------------------------------
Date of Submission: 16-10-2017 Date of acceptance: 02-11-2017
---------------------------------------------------------------------------------------------------------------------------------------
I. INTRODUCTION
1.1 Cloud Computing Definition
Cloud Computing has become the emergent technological term common with computing data storage
and offsite computing resource management. A recent study on its definition reviewed twenty-two (22) different
definitions that saw many of them overlapping in meaning [1]. However, its definition by the United States
National Institute of Standards and Technology (US-NIST) has been viewed as more refined since it captures its
practical meaning [2].
In order to arrive at a more functional understanding of Cloud Computing, there is need to understand
the concept of the term CLOUD. Weinmann [3] derived five logical tautology and concludes that a cloud
structure is a six-tuple (space, time, directed graph, set of states, transition function, and initial state) that
satisfies the five axioms namely; Common, Location-independent, Online, Utility, and on-Demand
(C.L.O.U.D).
The definitions of the five axioms were captured as follows;
 Common: It is common if and only if there are at least two nonzero flows that are commonly sourced.
 Location-independent: It is location-independent if there still exists some feasible allocation at any other
possible locations given that at least one feasible allocation at time t given a particular configuration of
locations.
 Online: It is online when there still exists a path between any two vertices on a network graph without
emphasizing on the direction of each edge.
 Utility: It is utility since its pricing is proportional to quantity.
 On-Demand: it is on-Demand if and only if it is completely feasible.

Enhancing Data Security in Cloud Computation Using Addition-Composition Fully Homomorphic ..
1.2 Cloud Computing Architecture
From the definition of cloud computing [4], cloud computing is delivered by server-based applications
accessible via the Internet. There are five evident qualities synonymous with cloud computing, namely, on
demand self-service, broad network access, resource pooling, rapid elastic or expansion, and measured services.
Fig.1 Cloud computing architecture
The architecture is designed in such a way that it has the front-end which is the visible interface that
cloud users encounter through web-enabled client devices, and the back-end which comprises resources required
to deliver cloud computing services. The back-end comprises of the bare metal servers, data storage
compartments, virtual machines, implemented security mechanism, and cloud service(s) with conformity with a
particular deployment model. It is the primary responsibility of the back-end to integrate built-in security
mechanisms, traffic management control, and protocols. In the bare metal has the operating system referred to as
hypervisor which makes use of well-defined protocols enabling efficient concurrent access onto the virtual
machines.
The cloud computing architecture is functionalized in such a way that all applications are controlled
and served by a central cloud server with replicated data remotely in the cloud configuration to create limitless
efficiency and possibilities. The figure below shows a diagrammatic summary of cloud computing architecture.
Fig.2 Cloud Computing Structure
Building on the five qualities of cloud computing mentioned above, the type of cloud computing can be
categorized based on two models: cloud computing Service Delivery Models and Cloud Computing Deployment
Models[5].
1.3 Cloud Service Delivery Models
The delivery of services within the cloud environment can be implemented through the following service
delivery models;

1.3.1 Software as a Service (SaaS)
This is a service delivery model where applications are available to the cloud users over the network.
The cloud providers takes the responsibility of hosting these applications used by the consumers under software
distribution model. This model is characterized by application delivery from 1:M model, that is, single-instance,
multitenant architecture as opposed to traditional 1:1 model, and also associated with pay-as-you-go
subscription licensing model. The distinguishing bit of SaaS from other service delivery model is that it was
designed to work with web browsers.
The main advantages with this model includes automated updates and patch management services, data
compatibility across the enterprise and global accessibility of software of interest by the cloud users.
1.3.2 Platform as a Service (PaaS)
This is service delivery model where all facilities required to support the complete life cycle of
building and delivery of web applications and services entirely available on the Internet. It is an outgrowth of
SaaS model but concerns with web-based development unlike with SaaS where specific operating systems
instance is favoured. The advantages with this service delivery model is that it allows users to focus on
innovation and not the complex infrastructure as it provides all the infrastructure needed.
1.3.3 Infrastructure as a Service (IaaS)
This is service delivery model where cloud service providers manage the transition and hosting of
selected applications on their infrastructure. In this model, the computing services are provided on a virtualized
environment with cloud users maintaining ownership and management of the applications while hosting
operations and infrastructure management are off-loaded to the cloud provider. The benefits that comes with this
service delivery model includes availability of resources on demand from any location 24x7when required by
the cloud users. The physical security of cloud users’ data and any chance of system failure is smoothly handled
by the cloud provider.
1.3.4 Communication as a Service (CaaS)
This is service delivery model where communication is outsourced and the provider is responsible for
management of VOIP services, IM services, video conferencing et cetera consumed by their user base. This
model allows a CaaS provider’s business customer to selectively deploy communication feature and services
throughout their company on a pay-as-you-go basis on the services used.
The advantages that comes with this model is that every component of this solution comes with 24/7
management by CaaS provider with no capital expenses needed. The customers has the benefit of multiple
carrier-grade data centers with full redundancy built on to the system with minimized point of failure.
1.3.5 Monitoring as a Service (MaaS)
This is service delivery model the provision of security is outsourced especially on business platforms
depends on the Internet to conduct business. This kind of protection benefits enterprise or single client assets
from cyber threats since there is constant vigilance over security of infrastructure and critical organizational
assets. The advantage that comes with this model is the early detection and protection against internal and
external threats since the platform is under constant control and monitoring.
1.3.6 Network as a Service (NaaS)
This is a service delivery model where the provision of a virtual network service is by the owners of the
network infrastructure to the third party. It is sometimes referred to as Telco As A Service (TaaS) since it
involves optimization of resources allocation by considering network and computing resources over a
connectivity. It is the business model that guarantees delivery of network infrastructure in a data center virtually
over the Internet. The advantages of this service delivery model is the on-demand network resource usage as the
network is maintained by the cloud provider and the ability to deliver network capabilities and services without
the requirements of configuring individual devices.
1.4Cloud Deployment Models
Due to the principle of outsourcing computing resources through cloud computing, their deployment can be
viewed under the following models.
1.4.1 Private Cloud
This is type of cloud infrastructure that is provisioned exclusive use by a single organization with multi-tenants
resource access. It may be owned, managed and operated by the organization, a third party or some combination
of them.

1.4.2 Public Cloud
This is type of cloud infrastructure that is provisioned for open use of resources dynamically by the general
public. It may be owned, managed and operated by one or more organizations who sells the cloud services.
1.4.3 Community Cloud
This is type of cloud infrastructure that is provisioned for exclusive use by a specific community of clients from
organizations that share common interests like mission, security requirements, policy or compliance
considerations. It is owned, managed and operated by one or more organization in a community.
1.4.4 Hybrid Cloud
This is type of cloud infrastructure composed of two or more distinct cloud infrastructures
characteristics (private, public or community) that remain unique entities, but are bounded together by
standardized or proprietary technology that enables data and application portability. It is owned, managed and
operated by both cloud infrastructures incorporated.
1.4.5 Virtual Private Cloud
This is a type of cloud infrastructure allocated within a public cloud environment provisioned for a
certain level of isolation between the different organizations sharing the cloud resources. It is viewed as a hybrid
model of cloud infrastructure in which a private cloud solution is provided within a public cloud infrastructure.
Its ownership, management and operations is assumed by the public cloud infrastructure.
1.5Cloud Computing Security
In consideration of cloud architecture, its security components and services must be transparent and
generic: transparent since there is need to be applied automatically without much human intervention, and
generic to ensure adjustability on the part of clients, requirements, applications and required services. A
functional cloud computing security architecture is composed of Security Access Points that provides front-end
security services, Security infrastructure servers that manage all cloud stored data registered in the cloud, and
secure client for cloud standard clients stations extended with some security components.
Fig.3 Cloud Computing Security Infrastructure
The multi-tenancy model and the pooled computing resources has introduced new security challenges
such as shared resources on the same physical machines inviting unexpected side channels between machine
resources and a regular resource [6]. Security being an important component of cloud computing deployment,
there is need to address it to enable cloud consumers enjoy its benefits uninterrupted [7].
Research has pointed out that the most important security issue in cloud computing is data security.
The problem is that the cloud service being a web application, it contains data remembrance or persistence
remains an issue due to the replication and distribution of data even after user has left a cloud provider. This can
be corrected by use of homomorphic encryption schemes to protect data on cloud [8].
1.6 Homomorphic Encryption Schemes and Cloud Data Security
The term homomorphism is borrowed from algebra meaning a structure-preserving map between two
algebraic structures of the same type. The concept of homomorphism has been employed in explaining
homomorphic, which means in literal sense same effect for different expression.
Encryption is one secure way of ensuring that cloud data is shielded from confidentiality abuse. The
fact that cloud user has off-loaded his data onto the cloud environment despite uncertainty and hostility that
comes with it, makes encryption a better candidate in security enhancement. A better proposal to ensure that

confidentiality of data is maintained while it is offloaded in the cloud environment is to find a way in which any
operation on the said data can be effected without a third party being privy to its contents.
Homomorphic encryption enables computation to be performed on the encrypted data without
requiring the secret key. This makes it a better approach to enhance security of sensitive data stored and
manipulated on untrusted storage systems [9]. Therefore an efficient and fully homomorphic cryptosystem have
great practical benefits to outsourcing private computations.
The concept of homomorphic encryption was introduced in 1978 by Rivest, Adleman and Dertouzous
after the development of RSA algorithm [10], where they highlighted four possible encryption functions which
include RSA as additive and multiplicative privacy homomorphism.
Since then, many encryption schemes have been developed but they were either using addition or
multiplication homomorphic computations [11]. According to [12], an encryption scheme was developed that
perform unlimited number of addition operations but single multiplication.
The realization of fully homomorphic encryption scheme appeared elusive for more than thirty years
until in 2009 when Craig Gentry from IBM implemented the first version that could perform many additions and
multiplications using ideal lattices and bootstrapping technique [13].
Even though Gentry’s scheme was able to demonstrate the possibility of fully homomorphic encryption
scheme, there was still need to improve on its complexity, efficiency and performance. For instance, he
estimated that building a circuit to execute an encrypted Google search with encrypted keywords would multiply
the current computing time by one trillion [11].
1.6.1 Gentry Fully Homomorphic Encryption Scheme
Gentry’s scheme [13] based his construction on somewhat homomorphic encryption scheme, that is, it
supported both addition and multiplication operations on ciphertexts. His idea was based on ideal lattices to
provide additive and multiplicative homomorphism. His construction started from a somewhat homomorphic
encryption scheme limited to evaluating low-degree polynomials over encrypted data. He then proceeded to
show how to slightly modify the scheme to make it bootstrappable. Then finally, he proved that any
bootstrappable somewhat encryption scheme can be converted to a fully homomorphic encryption scheme
through a recursive self-embedding.
The bootstrapping technique “refreshes” the ciphertext to reduce noise factor responsible for making
the ciphertext indecipherable, which was increasing with every increase in addition and multiplication of
ciphertexts. With the introduction of periodical refreshing by the Boolean circuit whenever the noise increases,
computation of arbitrary number of additions and multiplications was made possible, qualifying the scheme to
be fully homomorphic.
He based the security of his scheme on assumed hardness of two problems:
 Certain worst-case problems over ideal lattices, and
 The sparse subset sum problem.
In general scenario, homomorphic encryption schemes are described by the following functions:
 KeyGen (λ): this is responsible for generating encryption keys where it takes the security parameter λ as an
input, and generates the secret key sk and the public key pk.
 Enc (pk, m): this is responsible for encrypting the plaintext m with the public key pk and create ciphertext
c.
 Dec (sk, c): this is responsible for decrypting the ciphertext c using the secret key sk to retrieve the
plaintext m.
In Gentry’s construction, he added another new function “Evaluate” as follows:
 Eval (pk, C, c1, c2, …….,cl): this uses a Boolean circuit C to output a ciphertext c’ of f(m) such that Decrypt
(sk, m) = f(m).
There are three main elements evident in Gentry’s scheme [11]:
 A Somewhat homomorphic encryption scheme that evaluates low degree polynomials
 A technique to “squash the decryption circuit” to obtain “bootstrappable” scheme, and
 A method of transferring “bootstrapping” the scheme into fully homomorphic scheme.
The goal of his scheme was to get a scheme that can evaluate high degree polynomials while the
decryption procedure can still be expressed as low degree polynomial. Provided the scheme could evaluate its
own decryption function plus an additional operation then it is called “bootstrappable” scheme and can be
converted into a fully homomorphic scheme.
1.6.2 Fully Homomorphic Encryption Scheme over Integers
In June 2010, a new scheme was proposed based on the hardness of the approximate integer greatest
common divisors (approximate GCD) problem [14]. It borrowed heavily from Gentry’s construction though

didn’t use ideal lattices. They considered a simpler mathematical approach where they replaced the Gentry’s
somewhat homomorphic scheme that uses ideal lattices with a simpler somewhat homomorphic scheme that
uses integer over polynomial ring.
The main goal of this scheme was to reduce the complexity of Gentry’s scheme. This scheme borrowed
from [15] which supported only additions though can be improved to accept a limited number of multiplications,
and also from [16] which was not even additively homomorphic.
In this Dijk’s somewhat homomorphic encryption scheme, it has the following functions:
 KeyGen (λ): The secret key here is a random -bit integer: p(2Z+1)Ո(2-1
, 2
). Where Z is a real number
and  is a bit-length of the secret key,  is the bit-length of the integers in public key,  is the bit-length of
the noise and  is the number of integers in public-key. For the public key sample a random number xi from
the distribution D,(p) for 0i subject to the condition that the largest xi is odd. The public key
pk=<x0,x1,……x>.
 Enc (pk, m):This encrypts the plaintext message m{1,2,…….,} and a random integer r in (-2
, 2
) the
output ciphertext c is calculated as c[m+2r+2iSXi]x0.
 Eval (pk, C, c1, c2,……,cl):The evaluate function work as per Gentry’s scheme, however all operations are
performed over integers.
 Dec (sk, c):The outputs m’ is calculated as m’(c mod p) mod2.
The greatest challenge with this scheme is that it can only be fully homomorphic scheme if the noise stays
sufficiently smaller than the secret key p, meaning if the noise grows bigger then the original plaintext cannot be
revealed well.
1.6.3 Fully Homomorphic Encryption without Bootstrapping
This proposed scheme was based on Learning with Errors (LWE) and Ring Learning with Errors (Ring
LWE) problems [17]. The bootstrapping technique was not employed here since there was need of having to
“squash the decryption circuit” in order to realize a fully homomorphic encryption scheme.
The Braskerski’s somewhat homomorphic algorithms consists of the following functions:
 KeyGen (lλ
): Sample sk:=sRq to be polynomial with small coefficients chosen from the error distribution
.
 Enc (sk, , R2): Sample aRq at random, and a polynomial e with small coefficient from error distribution
. The output ciphertext c is calculated as c:=(a,as+2e+).
 Dec (sk, c=(a,b)): Sample bRq, at random, :=b-as over Rq. The output :=’(mod2).
Unlike in the earlier Fully Homomorphic Encryption schemes, the breakthrough in this scheme is the use of
modulus switching in the management of noise so that it increases linearly with the multiplicative level instead
of exponentially in precious cases. This allowed the scheme to increase the number of homomorphic operations
without squashing or bootstrapping.
1.7 Analysis of the Fully Homomorphic Schemes
The fully homomorphic encryption schemes are compared based on three parameters, namely, security,
efficiency and complexity factors.
1.7.1 Gentry’s Fully Homomorphic Encryption Scheme
Complexity: This scheme is based on somewhat homomorphic encryption scheme with assumed hardness
of certain worst-case problems over ideal lattices, and the sparse subset sum problem.
Performance: The ciphertext size and computation time increases with increase in security level making the
scheme impractical [18].
Security:The security of this scheme is based on assumed hardness of certain worst-case problems over ideal
lattices, and the sparse subset sum problem [19].
1.7.2 Fully Homomorphic Encryption over Integers Scheme
Complexity: This scheme was motivated by the need to reduce the complexity of Gentry’s scheme. It
significantly proved that different mathematical approaches and theories can be applied to construct a fully
homomorphic encryption scheme using Gentry’s blueprint.
Performance: The noise factor increases here with addition and multiplication operations performed, that is,
it doubles with addition and squares with multiplication. For instance, in order to improve the security of the
scheme, the ciphertext was selected to have a large value n6
, which increases with multiplication hence impact
negatively on the efficiency.
Security: The security of this scheme is based on the hardness of the approximate-gcd problem. With
appropriate selection of parameters the scheme has proved to resist different types of attacks to recover the
secret key including brute-force attack with at least 2λ time. It has also been proved that the scheme can be
attacked to recover the plaintext from ciphertext using lattice reduction algorithm [20].

1.7.3 Fully Homomorphic Encryption without Bootstrapping Scheme
Complexity: In comparison with Dijk’s scheme, Brakerski’s scheme uses more complex mathematical
algorithms and notations as a result of Ring Learning With Errors instead of working with integers. Compared
to Gentry’s scheme, removal of bootstrapping technique has simplified the decryption functions and
calculations.
Performance: This scheme made milestone in noise management technique under three considerations. First
the technique controlled the noise level by constraining it to increasing linearly with multiplication instead of
exponentially. Secondly it had allowed the L-level arithmetic circuit to be evaluated with Ō(λ.L3
) per-gate
computation or instead of Ω’(λ4
) which is large polynomial in the security parameter. Thirdly the removal of
bootstrapping technique resulted on real cost reduction as the cost of bootstrapping in only Ɵ(λ) time was
Ω’(λ4
), hence allowing evaluation of deeper circuits at a lower cost.
Security:The security of this scheme is based on the hardness of lattice problems with quasi-polynomial
approximation factors. The security level has not improved from original Fully Homomorphic Encryption
Scheme as it remains 2λ time against known lattice attacks.
II. METHODOLOGY
In order to design the Addition-Composition fully homomorphic scheme, there is need to define and prove
mathematical concept around it. It is the developed mathematical concept that was used to generate an
algorithm, which was later tested. We started by deriving mathematical theory to support the new scheme.
2.1 Mathematical Background
Let R = Z[x]/ )
(x
 where )
(x
 is irreducible over Q[x] but factor modulo t. If  splits exactly into r distinct
irreducible factors of degree g = n/r = r
d /
)
(
 i.e. 
)
(x
 

r
j
j x
j
1
)
( .
then by Chinese Remainder Theorem, the following R t  Z t [x]/ (f1 (x)  ……  Z t [x]/f r (x)) is a natural
isomorphism which is obtained.
In particular this helps to add and multiply in parallel. In our new technique we obtain the addition-composition
via tensor products.
Definition 2.1
Let f(x) and g(x) be two polynomials such that f(x):A  B and g(x):B  C where A, B, C are arbitrary rings.
Then the composition of f(x) and g(x) is well defined and given by:
g(x)  f(x) = (g  f)(x) = g(f(x))  x A.
Example
Let f(x) = 3x+1 and g(x) =x-2. Clearly, (g  f)(x) = 3x-1.
Definition 2.2
Consider two ciphers  and .

We define addition-composition homomorphic encryption by:
)
( 

  = )
(
)
( 


  where  is a composite function and  is an additive function.
Lemma 2.3
Generally,




m
i
i
i
1
)
(
:
)
( 


 




m
Proof
Definition is well posed and defined. To see this consider  and  and z


 where z=Z.
So  =
2
1 modn
y
x

 
 =  mod
2
y
k

 2
n

  =
k
y
x


 
1 mod
2
 n2
= mod
)
( 2
1
y
k
x


 

n
2
So 
 2
1 
 …………..  m
 =
)
.........
( 2
1 m
x
x
x 


  ( mod
)
.
..........
( 2
1
y
m


 m
n and this completes proof.
The addition-multiplication homomorphic encryption is to the advantage of being usable in real time. The
composite function element comes in handy since the result of the composite function f(x) is encrypted and

malicious hosts cannot know the results of the function. For instance in the mobile phone technology, the owner
of the function gets the encrypted result through the function g(x). For instance Richard is the owner of function
h(x) he wants to calculate the input c of Bernard but he doesn’t want to expose himself as the owner of the
function. So he chooses g(x) and creates f(x) then sends it to Bernard. Bernard hence calculates the result
through f(x) using his input x and sends the result to Richard. Bernard cannot calculate h(x) because what he can
see is just f(x). Only Richard can get the real result of h(x) through adding f(x) into inverse function i.e. h(x) = g
1

(f(x)). For this reason, we develop some theoretic results based on the addition-composition homomorphic
encryption[21].
III. SCHEME TESTING AND RESULTS
To run the program run
> javac acfhes.java
> java acfhes 10 20
The above runs the program specifying values of 10 and 20 for m1 and m2 a brief explanation of the outputs, this
can be found in the code
$ java acfches 10 20
# Specify 10 and 20 as inputs
10 composed to 34 # (10 + 7) * 2
20 composed to 54 # (20 + 7) * 2
34 reverse composed to 10 # (34 / 2) - 7
54 reverse composed to 20 # (54 / 2) - 7
Input 10 was encrypted to
914230298129780563011715681158340346654543602699977594450413796356184635162358093323230424
681696171637738663970932013303240313090330397882701683852919774558090980556153806745895351
887029329584210683602643550613641944502069410962257148973647432913024456315152143971912392
18922096715229579116578586841359618827
Input 20 was encrypted to
195871832380398390712064736903081228410982010530497359491304890546381955178006401516941397
235860351701415195879993524751311621904227470186880716468621690370012492329637488592526514
419657178063328558900523651203645487805449824562590341630887921851927793128324256451209785
33016963729009633333582640628244441680
Composed 34 was encrypted to
435979835190856069575731053492627477693122994147643478666481631966106300080928605072941717
361105448847316024008073488076464116457390131962519865940279770694936674882194902603300653
368128233844131406381093790852058588239611074466185351965911407413382081930993606892729547
2518192502771130916516605558783858404
Composed 54 was encrypted to
562234489019579959780500888969326350546292803602895504092061574348688071282624629945470971
345564002577314785691081471268095671752902222216274678842140014299124832954267865103647902
544118836854787492471316016565033986749963234007657859580660106797755236895691903430025641
55127322014572152909088866952530229958
First 10 input decrypted to 10
Second 20 input decrypted to 20
First composed 34 decrypted to 34
Second composed 54 decrypted to 54
# Above shows acfches successful encryption and decryption
Original sum: 30
# Manual sum of 10 + 20

Composed sum: 88
# Manual sum of 34 + 54
# Below are now homomorphic
/* test homomorphic properties -> D(E(m1)*E(m2) mod n^2) = (m1 + m2) mod n */
Decrypted original sum: 30
# Decrypt(E(10) * E(20) mod n2
)
Decrypted composed sum: 88
# Decrypt(E(34) * E(54) mod n2
)
Reversed decrypted composed sum: 37
# Applying ReverseComposition on Sum
/* test homomorphic properties -> D(E(m1)^m2 mod n^2) = (m1*m2) mod n */
Original product: 200
# Manual product 10 * 20
Composed product: 1836
# Manual product 34 * 54
Decrypted product: 200
# (E(10) to power 20) mod n
Decrypted composed product: 1836
# Sum as above
Reversed decrypted composed product: 911
# Result from reverse composing theproduct.
IV. CONCLUSIONS
The study focused on the design and analysis of addition-composition fully homomorphic encryption
technique that enhances data security on cloud computing. We have demonstrated that this is a possibility. A
new scheme was developed that has foundations on mathematical proofs on validity and reliability. The study
also demonstrated that on the tested metrics, addition-composition fully homomorphic scheme performs better
than the previous schemes. The scheme is complex for easy adversarial attack, the speed of encryption faster
and the resource consumption are manageable.
REFERENCES
[1] Luis M. Vaquero, Luis Rodero-Merino, Juan Caceres and Maik Lindner. A Break in the Clouds: Towards
a Cloud Definition. ACM SIGCOMM Computer Communications Review, Vol. 39, No. 1, 2009.
[2] Peter Mell and Tim Grance. The NIST Definition of Cloud Computing. NIST Special Publication 800-
145, Vol. 15, 2011.
[3] Joe Weinmann. Axiomatic Cloud Theory. Communications, Media and Entertainment Industry for
Hewlett-Packard, 2011.
[4] NIST. US National Institute of Standards and Technology. Special Publications 800 – 145, 2011.
[5] Rittinghouse J. W., and Ransome J. F. Cloud Computing: Implementation, Management and Security.
(CRC Press, 2010)
[6] Gnanavelu D. and Gunasekaran G. Survey on Security Issues and Solutions in Cloud Computing.
International Journal of Computer Trends and Technology. Vol. 8 No. 3. Pp 126 – 130. ISSN: 2231-
2803, 2014.
[7] Tiwari P. K. and Mishra B. Cloud Computing Security Issues, Challenges and Solution. International
Journal of Emerging Technology and Advanced Engineering. Vol. 2 Issue 8 ISSN: 2250-2459, 2012.
[8] Kumar K. S., Anjaneyulu N. and Venkanna. Security Issues in Cloud Computing and Study on
Encryption Method. International Journal of Emerging Trends and Technology in Computer Science.
Vol. 2 Issues 3, pp 442 – 446, ISSN: 2278-6856, 2013.
[9] Chakraborty Nilotpal. Cloud Security using Homomorphic Encryption. National Conference on
Advances in Computing, Networking and Security, 2013.
[10] Rivest R., Adleman L., and Dertouzos M. On Data Banks and Privacy Homomorphisms. Foundations of
Secure Computation, Academic Press, pp. 169 – 177, 1978.

[11] Amna Ahmed Ali. A Comparative Study of Fully Homomorphic Schemes for Cloud Computing.
International Journal for Emerging Technology and Advanced Engineering. Vol. 3 Special Issue 4. 2013.
[12] Boneh D., Goh E., and Nissim K. Evaluating 2-DNF Formulas on Ciphertexts. In Proceedings of Theory
of Cryptography (TCC) ’05, LNCS 3378, pp. 325-341, 2005.
[13] Gentry C. A Fully Homomorphic Encryption Scheme. PhD Thesis in Computer Science, Stanford
University, California, USA, 2009.
[14] Marten Van Dijk, C. Gentry, S. Halevi and V. Vaikuntanathan. Fully Homomorphic Encryption Over
Integers. Advance Cryptology – EUROCRYPT’10, Vol. 6110. Springer, 2010, pp. 24- 43.
[15] Leviel E., and Naccache D. Cryptographic Test Correction. 2008.
[16] Cohen Bram. Simple Public Key Encryption. 1998.
[17] Brakerski Z., Gentry C., and Vaikuntanathan. Fully Homomorphic Encryption without Bootstrapping.
Proceedings of the 3rd
Innovations in Theoritical Computer Science Conference (ITCS’12), pp. 309 –
325, ACM Press, New York, NY, USA, 2011.
[18] Lauter K., Naehrig M., and Vaikuntanathan V. Can Homomorphic Encryption be Practical? CCSW’11
Proceedings of the 3rd
ACM Workshop Cloud Computing Security Workshop, pp 113 – 124, 2011, ISBN:
978-1-4503-1004-8, ACM New York, NY, USA. 978-1-4503-1004-8, ACM New York, NY, USA.
[19] Hu Y. and Wang F. An Attack on a Fully Homomorphic Encryption Scheme. Proceedings PKC2010,
International Association for Cryptologic Research. 2010.
[20] ChungSheng Gu. Attack on Fully Homomorphic Encryption over Integers. Cryptology ePrint Archive
Report 2012/157. 2012.
[21] Omollo R. O. and Okelo B. N. On Theory and Applications of Mathematics to Security in Cloud
Computing: A Case of Addition-Composition Fully Homomorphic Encryption Scheme. International
Journal of Academic Studies. Vol. 2 No. 4, 2016.
Omollo R.O Enhancing Data Security in Cloud Computation Using Addition-Composition
Fully Homomorphic Encryption Scheme.” IOSR Journal of Engineering (IOSRJEN) , vol. 7,
no. 10, 2017, pp. 20-29.

Enhancing Data Security in Cloud Computation Using Addition-Composition Fully Homomorphic Encryption Scheme

Recommended

More Related Content

Similar to Enhancing Data Security in Cloud Computation Using Addition-Composition Fully Homomorphic Encryption Scheme (20)

Recently uploaded (20)

Enhancing Data Security in Cloud Computation Using Addition-Composition Fully Homomorphic Encryption Scheme