Journal of Software Engineering and Applications, 2013, 6, 23-28
doi:10.4236/jsea.2013.63b006 Published Online March 2013 (http://www.scirp.org/journal/jsea)
23
TCLOUD: A New Model of Data Storage Providing
Public Verifiability and Dynamic Data Recovery for
Cloud Computing
Sultan Ullah, Xuefeng Zheng, Feng Zhou
School of Computer and Communication Engineering, University of Science and Technology Beijing, Beijing, People Republic of
China.
Received 2013
ABSTRACT
Cloud computing is a new paradigm of computing and is considered to be the next generation of information technol-
ogy infrastructure for an enterprise. The distributed architecture of cloud data storage facilitates the customer to get
benefits from the greater quality of storage and minimized the operating cost. This technology also brought numerous
possible threats including data confidentiality, integrity and availability. A homomorphic based model of storage is
proposed, which enable the customer and a third party auditor to perform the authentication of data stored on the cloud
storage. This model performs the verification of huge file’s integrity and availability with less consumption of computa-
tion, storage and communication resources. The proposed model also supports public verifiability and dynamic data
recovery.
Keywords: Data Storage; Integrity; Confidentiality; Cloud Computing
1. Introduction
This The advancement of network technology and ampli-
fication in the requirement for computing wherewithal
have provoked countless business firms to subcontract
for storage requirements [1]. This innovative model of
storage is generally known to be cloud storage, having
identical attributes like cloud computing in form of dex-
terity, scalability and flexibility. The services of distrib-
uted and large scale system can be materialized in an
easy way due to the development of cloud system. This
system offer a uncomplicated and integrated interface
connecting vendor and customer, permitting the custom-
ers to focus more on the services itself relatively than the
essential underlying structure. There are numerous cloud
applications, ranging from hardware platform to software
system and multitenant databases. The allocation of
computational resources takes place dynamically, as per
the request of the customers demand and in conformity
with predesigned service quality by the customer [2].
Cloud model of storage has brought a lot of benefits
for the customers. By availing the services of this tech-
nology the customers are no more required to purchase
high cost equipment for their personal offices or data
centers. The customer can acquire numerous virtual
storage devices just using an ordinary web browser, and
there is no need to have an entire awareness about the
installation and communication. Additionally, the main-
tenance tasks of the storage, for instance backing up the
data, data duplication, and obtaining extra storage
equipments, are delegate to the liability of a supplier,
permitting the business firms to only care for their main
business. It gives the impression that, the user only have
to perform one task, to pay for the services they avail and
it is normally very economical [2,3].
The tools that have been proposed for the assurance of
data by the cryptographic community are known to be
the proofs of retrievability (PORs) and proofs of data
possession (PDPs). The POR represent a mechanism of a
challenge and reaction protocol which facilitates the user
to confirm whether a data file can be retrieved from
cloud server or not. The advantage of POR as compared
to simple file transmissions is its effectiveness. The reply
will possibly be very much compacted, so the client only
needs a small portion of the file for completing the proof.
The value of POR is limited if used for verifying the re-
trievability in a single server environment and as a sepa-
rate means. It is not important to detect that the file a
customer is looking for is corrupted, rather than if the file
is not retrievable and the customer has no alternative
[4-7].

TCLOUD: A New Model of Data Storage Providing Public Verifiability and
Dynamic Data Recovery for Cloud Computing
24
It is important to have a system for open verifiability,
when the files are scattered athwart the several comput-
ing systems. Herein, the data files are stored in superflu-
ous shape across various servers. The availability can be
checked on every server using POR by a customer. If a
data loss is encountered inside a specific server, the cus-
tomer is able to have a request for the recovery of the file
to other server.
A single data file is shared transversely through server
redundantly with the help of some reassure code in an
environment with distributed file system. The use of this
code method helps in the recovery of file during server
failures. This will also facilitate a user to verify the reli-
ability of data files by reclaiming the small pieces from
each server and again checking its uniformity. The man-
agement of availability and integrity is maintained by the
system in the group of servers. The PORs tools are al-
lowed by these systems to test and modify the storage
resources wherever any failure is encountered [8,9].
As thy advantages of the cloud storage are very prom-
ising, but still as it is said, there is no opportunity with
taking risks, the issue of data confidentiality and integrity
becomes a bottleneck in cloud computing and one of the
main hurdle in the fast adoption of this technology [2,[3].
The rest of the paper is organized as; in section 2 re-
lated literature is presented, section 3 contains the se-
cured scheme of data storage, section 4 describe the pro-
posed model of integrity checking and section 5 presents
the evaluation of proposed model and conclusion is pre-
sented in section 6.
2. Review of Related Literature
A lot of researchers have worked on the issue of remote
verification of integrity in the storage mechanism
adopted by cloud, all them have focused on different
scenario of applications and tried to attain various objec-
tives [4-14]. A model to ensure the control of data files in
an environment which is not trustworthy is proposed by
researcher and is known is provable data possession
[6,10]. The proposed technique make use of the homo-
morphic authenticator based on RSA for inspecting the
data which is outsourced and it also suggest arbitrarily
sampling a small number of block in the file. But, the
proposed methods have no support for public auditability
and a specific limit is imposed on the quantity of audits
to be performed.
A more thorough verification of the techniques/model
presented called the proof of retrievability [11]. The
model utilizes a unique combinations of codes called the
error – correcting and spot – checking in order to make
sure the availability and control of the outsourced data.
However, the number of operations and challenges is
fixed in advance, which put a limit, if there is any effec-
tive expansion which will support the public auditability
and updates. The dynamic proof of data passion was first
proposed which is based on the method of skip lists [10].
This method provide a means for controlling the out-
sourced data and having the dynamic support as well, yet
this model suffer from efficiency issues.
A framework which is provides a dynamic mechanism
of for open verification, working on the method of chal-
lenge and response protocol, capable of locating any
possible errors and also verifies the data correctness is
proposed in [13]. On the other hand, the ineffective per-
formance to a great extent has an effect the realistic
claim of the framework.
The models and methods presented above, each one
presented some mechanism to provide guarantee on the
accuracy of outsourced data, but none of it can have ef-
fective verification, dynamic changes in the contents of
the data, preservation of privacy and public auditability
of storage in the cloud. These are some of the problem
that will be under consideration in this paper.
3. Problem Statement and Solution
Requirement
In this paper the problem which is under consideration is
the cloud storage, where a customer / the owner of the
data need to store its personal / confidential data on the
cloud storage. These cloud storage service providers
named storage server supposed to store the outsourced
data, till the owner of the data request for the retrieval of
it.
The basic requirements which a storage system must
ensure to be present are confidentiality, integrity and
availability.
4. Illustration of the Proposed Model
In this part an innovative model for the data storage and
preservation is presented for cloud data storage. This
model is based on the concept of erasure coding instead
of replication in order to be more efficient.
4.1. Encryption Requirement
This protocol is made secure by using the full homo-
morphic encryption technique. The function used to real-
ize the homomorphism is represented by C
H
F. The
C
H
F is utilized to create a relation between the gener-
ated block of data to a piece of metadata, which is used
for the proof of the reliability of the data block. Such as,
a block of data represented by
j
BD then the related
metadata is ()
Cj
H
FBD which also a block digests and
required to be accumulated at a chosen auditor.

TCLOUD: A New Model of Data Storage Providing Public Verifiability and
Dynamic Data Recovery for Cloud Computing
25
The linear combination is preserved through the use of
homomorphism, and each set


1
j
BD wherejn and
for every coefficient

1,
j
CO wherejn then can be
simplified as;

11
nn
Cjj jCj
jj
H
F COBDCOHFBD


 


 (1)
The homomorphic function C
H
Fcan also be devel-
oped on the basis elliptic curve method or may be on the
discrete logarithm method [14,15].
4.2. The Detail Working of the Proposed Model
The propose model is comprises of four phases, and each
phase is described in this section in detail.
a) Data Storage Phase
In order to ensure the availability, consistency and due
to the nature of the cloud storage environment, the data
needs to be stored on different servers. The proposed
model used the erasure coding method which needs
fewer amounts of data storage space as compared to rep-
lication method [14].
The above algorithm is used to store n blocks of
data. First the data DaT is divided into ,n blocks

1
j
BD wherejn. The blocks are encode to make
nm coded blocks {}1
jjnm


 , one of these
blocks facilitate the recovery of the original data. The
random erasure codes create a matrix M in Z as follows:
n
I
A
M
A
A



(2)
Algorithm for Data Storage
START:
Step 1:


,
j
OwRDaTn BlockBD

jk
jk
1
1
2: Ζα,
1,1
3: Repeat for 1
α,
4: Repeat for 1
5: Re
n
jk
j
j
wherej n
StepOwR GNRTRND
wherekn mjn
Stepknm
OwR
Stepkn m
OwRjth Server
Step





 

 



peat for 1
()
:
j
knm
jth Serverkeep
END



In the above equation n
I
A represents and identity
matrix with nn entries, and AA is used to denote a
mn matrix in Z and the values are selected randomly.
Every coded block
j
B is produced in Z as follows:
jk
1
α,
n
j
k
j





 (3)
In the above equation jk
α, is any entry of M at the
location of
J
th rowandKth Column The coded block
of data
j

is sent to the storage server finally by the
owner.
b) Allocation of Auditor
In this phase the owner of the data assign the task of
data verification to either a third party auditor TPA or
numerous data user.
It is quite possible that multiple coded blocks will be
verified by a single TPA or user. The TPA which is as-
signed the task of verification will receive metadata in-
formation
j

from the owner such that ()
j
Cj
HF

 .
The
j

represent a numerical value which maps the
coded block stored at the specified storage server.
Algorithm for Assignment of Auditor

:
1: Repeat for 1
()
2: Repeat for 1
1
, :
3: Repeat for 1
1
jCj
j
START
Stepkn m
OwRH F
Stepkn m
Repeatj TP
OwRjkthe TPA
Stepkn m
Repeatj TP










, ()
:
j
jKthe TPAkeep
END


c) Public Auditability
The third party auditor or any user capable of verifying
data can perform the operation of verification on the ba-
sis of the metadata. The method only guarantees the con-
firmation of one block and is considered to be limited.
The cloud storage server is challenge with a message
()
M
CN
CHHF R

from TPA, where
N
R is a random
number. The cloud server computers


M
Mj
RCH


in response to the challenge from TPA, then
M
R is send
back to. The TPA then verifies the response as under;




M
Mj jCN
NC jNj
RCH HFR
RHFR



 (4)
The TPA checks equality of the expression, if the
equality is not preserved then the TP A considers the data
block to be either damaged or corrupted. In response to
solve the problem of data loss or replace the corrupted
block, a new block will be created.
The algorithm to follow these entire steps is given as
under:

TCLOUD: A New Model of Data Storage Providing Public Verifiability and
Dynamic Data Recovery for Cloud Computing
26
Algorithm for Data Verification
:
1:
2: ()
3:
4: ().
5:
6:
:
N
MCN
M
M
Mj
M
MNj
START
StepTPA GNTR
StepTPACHHF R
StepTPACH CSC
StepCSC RCH
StepCSCR TPA
StepIf RRthen
InitiateRepair Phase
END





 


d) Recover Lost Data
In the recovery phase, TPA detects the corruption of
data in a block at a specific cloud storage server. The
data block can be regenerated from coding operation
over n blocks. The operation of coding is performed by
the new cloud storage server. The new cloud storage
server will receives n blocks of verified data, which is
selected form the set of left over cloud storage servers.
The random coefficients are also generated and the
new block l

 is computed in Z as follows:
Oj
1
C
n
lj
l





 (5)
The recently created data block can be inscribed as a
linear arrangement of the old data blocks. This can have
the following form;
Ol
1
ln
1
C
α,
l
n
ll
l
n
On
j
C







 




(6)
Then,
ln
11
α,
l
nn
lOn
jl
C








 (7)
Therefore, the random linear method of erasure coding
is used for the creation of the new data block. For in-
stance, if only thel

 is destroyed or corrupted then only
the row containing the data block will be updated in the
data matrix M The new record can be entered by the fol-
lowing equation:
Ol ,
1
C
n
llj
l





 (8)
The TPA will work as the auditor for the new data on
the new cloud server, but these authorities will need the
fresh copy of the metadata and will use the same set of
coefficients.
Ol
1
C
j
n
TT
l
M
M


 (9)
The creation of the new data block and the metadata
conform to the homomorphic encryption of the function
C
H
F.

 
OlOl ,
11
Ol ,
1
CC
C
j
nn
TT Clj
ll
n
CCljCl
l
MM HF
HF HFHF













(10)
The TPA will inform the owner of the data about the
update and the other entities interested in the data. The
cloud storage service is a distributed environment, so the
information about the update process should be distrib-
uted among the entities interested the data.
Algorithm for Data Recovery
Ol
1
:
1: ()
2: {}1
3: {}1
4: C
5:
6: {}
7: {}1
8:
N
Nl
NOl
n
Nl l
l
Nl
NNlj
NOl
START
StepTPASLCTRAND CSC
StepTPACSCl n
StepCSCGNT RND Cln
Step CSC
Step CSC
StepTPACSClj n
StepTPAGNT RAND Cln
StepTP










 




Ol
1
C
9:
:
n
Nl lj
l
Nl
A
Step TPA
END





 


5. Evaluation of the Proposed Scheme
Av The proposed model is evaluated for security con-
cerns and performance in this section.
a) Security Analysis
The proposed model provides secure mechanism of the
data storage. It helps the cloud storage provider to pre-
vent any DoS ,
M
an in theMiddle

 and attacks at
the recovery phase of the model. In order to reduces the
chances of DoS / flooding attack, the system will pro-
vide only quota system for the regulation of test / verifi-
cation messages. The TPA or another entity involved in
the verification process, will be given a timeframe for
sending integrity checking messages to cloud storage
server. This possibly reduces the chances of flooding
attack by a malevolent user.
A cloud storage server or a third party auditor may
possibly send some fake information to the entity con-
cerned. This issue can be resolved by introducing signa-
ture which provide the proofs of origin and increase the

TCLOUD: A New Model of Data Storage Providing Public Verifiability and
Dynamic Data Recovery for Cloud Computing
27
reliability of information for the recipient.
  

{, 1
Cj OwRCj
H
FSiGHF jn   (11)
In order to make things easier and reduces the over-
head of the signature, the model is based on the signature
structure encrypted by homomorphic means. It has the
following shape;

Ol Ol
11
CC
nn
OwRllOwR l
ll
SiG SiG
 






 (12)
The man in the middle attack can be prevented, when
the reply to a challenge for verification is sent back to the
auditor in the hash code generated from the metadata and
ID of the cloud server in the form of
 
,
MCj
RHsHHF ID  .
This is oneway pseudorandom function. The ID is coun-
terfeiting, but it is assumed that there is no attack on the
routing. This can possibly reduces the chances of the
malicious server to be the one on which the original data
was stored due the difference in the ID.
b) General Performance
The two phase’s i.e the storage and the authentication
is a specific case of elliptic curve and is already evalu-
ated [14]. As the metadata is computed on the basis of a
single block rather than the entire data, so the perform-
ance is improved. The method proposed for the recovery
of lost or corrupted data required the exact amount of
coded block one time only, and offer modified created
block to the fresh cloud data server. The communication
bandwidth needed for the recovery is divided among the
cloud storage server. The over head can be further opti-
mized by introducing the hierarchical codes presented in
[15].
The data can be prevented from the destruction by
means of oneway homomorphic function C
H
F. The cloud
server cannot change the content of ()
M
CN
CHHFR
from the auditor challenge for the integrity. The cloud
storage server has to response correctly and keeps the
whole block of the data.
6. Conclusion
Av Herein, the problem of data integrity is investigated
in cloud data storage. In order to solve the issue, an effi-
cient model of public variability and dynamic data re-
covery is proposed. On the basis of security and effi-
ciency analysis, it believed that this model is acquiescent
for the cloud computing applications.
The auditing method consumes less communication
and computation resources, which is the reason for
higher improvement in efficiency and is most suitable for
cloud data storage system.
REFERENCES
[1] Kamara, Seny, and Kristin Lauter. "Cryptographic cloud
storage." Financial Cryptography and Data Security
(2010): 136-149.
[2] Ullah, Sultan, and Xuefeng Zheng. "Cloud Computing
Research Challenges." In Biomedical Enginnering and
Informatics, 2012. BMEI12. fifth International Confer-
ence on, pp. 1397-1401. IEEE, 2012.
[3] Ullah, Sultan, Zheng Xuefeng, Zhou Feng, and Zhao
Haichun. "Tcloud: Challenges And Best Practices For
Cloud Computing." International Journal of Engineering
1, no. 9 (2012).
[4] Kaufman, Lori M. "Data security in the world of cloud
computing." Security & Privacy, IEEE 7, no. 4 (2009):
61-64.
[5] Mykletun, Einar, Maithili Narasimha, and Gene Tsudik.
"Authentication and integrity in outsourced databases."
ACM Transactions on Storage (TOS) 2, no. 2 (2006):
107-138.
[6] Shacham, Hovav, and Brent Waters. "Compact proofs of
retrievability." Advances in Cryptology-ASIACRYPT 2008
(2008): 90-107.
[7] Ateniese, Giuseppe, Roberto Di Pietro, Luigi V. Mancini,
and Gene Tsudik. "Scalable and efficient provable data
possession." In Proceedings of the 4th international con-
ference on Security and privacy in communication ne-
towrks, p. 9. ACM, 2008.
[8] Wang, Qian, Cong Wang, Jin Li, Kui Ren, and Wenjing
Lou. "Enabling public verifiability and data dynamics for
storage security in cloud computing." Computer Secu-
rity–ESORICS 2009 (2009): 355-370.
[9] Hao, Zhuo, and Nenghai Yu. "A multiple-replica remote
data possession checking protocol with public verifiabil-
ity." In Data, Privacy and E-Commerce (ISDPE), 2010
Second International Symposium on, pp. 84-89. IEEE,
2010.
[10] Ateniese, Giuseppe, Randal Burns, Reza Curtmola, Jo-
seph Herring, Lea Kissner, Zachary Peterson, and Dawn
Song. "Provable data possession at untrusted stores." In
Proceedings of the 14th ACM conference on Computer
and communications security, pp. 598-609. ACM, 2007.
[11] Juels, Ari, and Burton S. Kaliski Jr. "PORs: Proofs of
retrievability for large files." In Proceedings of the 14th
ACM conference on Computer and communications secu-
rity, pp. 584-597. ACM, 2007.
[12] Sebé, Francesc, Josep Domingo-Ferrer, Antoni Marti-
nez-Balleste, Yves Deswarte, and J-J. Quisquater. "Effi-
cient remote data possession checking in critical informa-
tion infrastructures." Knowledge and Data Engineering,
IEEE Transactions on 20, no. 8 (2008): 1034-1038.
[13] Erway, Chris, Alptekin Küpçü, Charalampos Pa-
pamanthou, and Roberto Tamassia. "Dynamic provable
data possession." In Proceedings of the 16th ACM con-
ference on Computer and communications security, pp.

TCLOUD: A New Model of Data Storage Providing Public Verifiability and
Dynamic Data Recovery for Cloud Computing
28
213-222. ACM, 2009.
[14] Oualha, Nouha, Melek Önen, and Yves Roudier. "A se-
curity protocol for self-organizing data storage." In Pro-
ceedings of The Ifip Tc 11 23 rd International Informa-
tion Security Conference, pp. 675-679. Springer Boston,
2008. (6)
[15] Duminuco, Alessandro, and Ernst Biersack. "Hierarchical
codes: How to make erasure codes attractive for
peer-to-peer storage systems." In Peer-to-Peer Comput-
ing, 2008. P2P'08. Eighth International Conference on,
pp. 89-98. IEEE, 2008.