Enforcing Hidden Access Policy for Supporting Write Access

in Cloud Storage Systems

Somchart Fugkeaw and Hiroyuki Sato

Department of Electrical Engineering and Information Systems, University of Tokyo, Tokyo, Japan

Keywords: CP-ABE, Access Control, Cloud Computing, Access Policy, Privacy, Policy Hiding.

Abstract: Ciphertext Policy Attribute-based Encryption (CP-ABE) is recognized as one of the most effective

approaches for data access control solution in cloud computing. This is because it provides efficient key

management based on user attributes of multiple users in accessing shared data. However, one of the major

drawbacks of CP-ABE is the privacy of policy content. Furthermore, the communication and computation

cost at data owner would be very expensive if there are frequent updates of data as those updated data need

to be re-encrypted and uploaded back to the cloud. For the policy privacy perspective in CP-ABE based

access control, access policy is usually applied to encrypt the plain data and is carried with the ciphertext. In

a real-world system, policies may contain sensitive information that must be hidden from untrusted parties

or even the users of the system. This paper proposes a flexible and secure policy hiding scheme that is

capable to support policy content privacy preserving and secure policy sharing in multi-authority cloud

storage systems. To address the policy privacy issue, we introduce randomized hash-based public attribute

key validation to cryptographically protect the content of access policy and dynamically enforce hidden

policies to collaborative users. In addition, we propose a write access enforcement mechanism based the

proxy re-encryption method to enable optimized and secure file re-encryption. Finally, we present the

security analysis and compare the access control and policy hiding features of our scheme and related

works. The analysis shows that our proposed scheme is secure and efficient in practice and it also provides

less complexity of cryptographic formulation for policy hiding compared to the related works.

1 INTRODUCTION

Traditional cryptography methods such as symmetric

or public key encryption is considered to be

inappropriate to be directly applied as an access

control solution for cloud computing. Public key

encryption solution provides multiple copies of

ciphertext that will be shared among multiple users,

while symmetric key encryption introduces the key

distribution problem for a large scale of users.

In 2007, the Ciphertext Policy Attribute Based

Encryption (CP-ABE) was proposed in [Bethencourt

et al. 2017]. It is regarded as an effective solution for

formulating a flexible access control enforcement to

outsourced data and multiple decrypting parties. In

CP-ABE, a set of attributes is assigned to each user

in the way they are embedded into the user’s secret

key. The ciphertext is associated with the access

policy structure in which encryptors or data owners

can define the access policy by their own control.

Users are able to decrypt a ciphertext if their

attributes satisfy the ciphertext access structure.

To date, several works adopting CP-ABE (e.g.,

Chase, 2009, Wang et al., 2010, Wan et al., 2012, Li

et al., 2012, Zhao et al., 2012, Yang et. al., 2014) for

access control solutions generally concentrate on

minimizing key management cost, optimizing

computing cost of interaction between data owner

and outsourced data storage, improving scalability

and efficient user or attribute revocation. However,

there are two additional significant requirements

regarding policy privacy and policy sharing for

supporting data encryption to the users having write

privilege. First, access control policy privacy

becomes more crucial for the attribute-based

encryption model. This is because access policies

may contain sensitive information which needs to be

protected from unauthorized users. For example, in

hospital information systems, a policy used to

encrypt patient treatment files usually reveals the

attributes of diseases, symptoms, or specialized

530

Fugkeaw, S. and Sato, H.

Enforcing Hidden Access Policy for Supporting Write Access in Cloud Storage Systems.

DOI: 10.5220/0006349605580564

In Proceedings of the 7th International Conference on Cloud Computing and Services Science (CLOSER 2017), pages 530-536

ISBN: 978-989-758-243-1

treatment. Therefore, a way to hide policies must be

introduced to solve this problem. Second, policy

sharing for supporting users with write permission is

required in CP-ABE data sharing scheme. In the CP-

ABE scheme, the policy is used by data owners for

data encryption. Nevertheless, in some cases, users

may have permission to update files. To complete the

updating, the file needs to be encrypted and loaded

back to the cloud. This incurs both communication

and computation cost at data owner side.

The aforementioned problems become more

serious when the shared data is considered as

sensitive and shared to a large number of

collaborative users. In addition, advanced data

sharing services require efficient and real-time data

encryption and policy management. Existing

solutions have not yet addressed these two problems

consecutively.

In this paper, we extend the capability of our

access control model called Collaborative-Ciphertext

Policy-Attribute Role-based Encryption (C-CP-

ARBE) proposed in (Fugkeaw et al., 2015) to

achieve access policy privacy preserving and

optimized cost of file re-encryption. To this end, we

apply hash-based conversion and random encryption

to protect the content of the policy and employ proxy

re-encryption to optimize the performance of file re-

encryption due to the data update.

This paper encompasses two major contributions as

follows:

 We develop a new policy hiding scheme based

on randomized hash-based public attribute key

validation. Our proposed scheme is a lightweight

crypto mechanism for policy hiding and policy

enforcement.

 We introduce a write access enforcement

mechanism based on the proxy re-encryption to

optimize the cost of file re-encryption.

2 RELATED WORK

The originally proposed CP-ABE (Bethencourt et al.

2007) formulates the access in monotone tree-based

structure. To support a more complex environment of

cloud computing where there is multi-owner and

multi-authority, multi-authority attribute based

encryption (MA-ABE) solution have been proposed

by several works [M.Chase, 2007, M. Chase et al.,

2009, Yang et al., 2014, Fugkeaw et al., 2015].

Nevertheless, none of these approaches have taken

policy privacy and write access enforcement into

consideration.

Zhao et al. (Zhao et al., 2011) proposed a secure

data sharing scheme based on a combination of CP-

ABE and attribute-based signature (ABS). Their

approach supports read and write access control. A

signature-based called Tsign contains the ABS’

verification attributes specifying the write privilege

attributes. Users who need to update the file and load

it back to the cloud need to be verified with the cloud

server.

Ruj et al. (Ruj et al., 2012) proposed a privacy

preserving authenticated access control for data

outsourced in clouds. They propose distributed

architecture for authentication and access control

which supports multiple reads and writes on shared

data. ABE and ABS are used to encrypt and sign the

data respectively. Signatures of both read and write

are verified in order to validate privileges.

Nevertheless, the write privilege is treated separately

from access policy specification. In addition, the

access policy taken by the write access user is not

hidden.

The attempt in dealing with the attribute-based

policy hiding was introduced by (Katz et al., 2008).

They proposed an inner predicate encryption.

Nevertheless, the proposed scheme is based on the

bilinear groups that require a product of three large

primes leading the overhead.

The notion of CP-ABE policy hiding is

introduced by (Nishide et al., 2008). They proposed

two constructions of ciphertext-policy hiding

supporting restricted access structures that can be

expressed as AND gates on multi-valued attributes

with wildcards. However, the scheme is found to be

limited in given CP-ABE construction and

selectively secure.

In (S. Yu et al., 2008), the access control

approach with hidden policy for content distribution

networks (CDNs) was proposed. The proposed

scheme is based on the symmetric external Diffie-

Hellman (SXDH) assumption between paired elliptic

curve groups. Nevertheless, the designed scheme is

not designed to support the multi-authority cloud

systems.

In (J. Lai et al., 2011), they proposed a ciphertext

policy hiding CP-ABE which is proven to be fully

secure. Their approach is based on the inner-product

predicate encryption (PE), which makes use of

composite order bilinear groups. With the PE

scheme, there is a fixed anonymity set applied to all

ciphertext while the predicate is also encoded in the

user keys. However, the complexity comes both

composite order groups and the size of predicates.

Enforcing Hidden Access Policy for Supporting Write Access in Cloud Storage Systems

531

3 BACKGROUND

This section describes the concept of CP-ABE and

our proposed access control policy (ACP).

3.1 Cp-Abe

Basically, the concept of cryptographic construction

and key generation used in CP-ABE is based on the

bilinear map.

Definition1: Bilinear Map

Let G

and G

be two multiplicative cyclic groups of

prime order p and e be a bilinear map, e : G

×G

→

. Let g be a generator of G

. Let H: {0,1}* → G

be a hash function that the security model is in

random oracle.

The bilinear map e has the following properties:

1. Bilinearity: for all u, v ∊G

and a, b ∊ Z

, e(u

, v

) =

e(u, v)

2. Non-degeneracy: e(g, g) ≠1.

3.2 Access Control Policy (ACP)

Definition 2: Access Control Policy (ACP) ACP is

an access tree-based structure. Let ACP T be a tree

representing the access structure. Each non-leaf node

of T represents the Role node where threshold gate is

associated with. We denote by parent(x) the parent of

the child node x in the tree. The function attr(x) is

defined only for x in a leaf node of the tree as the set

of attributes associated with x.

We also introduce a special attribute “privilege” as

an extended leaf (EL) node of the ACP T in order to

identify the read or write privilege.

Figure 1: Access Policy Structure.

Figure 1 shows an example of the collaborative

access control policy tree expressed in a patient

treatment system. In the policy, there are three major

roles pharmacist, MD, and nurse who are all allowed

to access the prescription records with different

privileges.

4 C-CP-ARBE CONSTRUCT

This section describes the major construction of our

access control model [Fugkeaw et al., 2015].

Table 1 presents the notations and its description

used in our proposed algorithms.

Table 1: Notations used in our access control.

Notation

Description

uid.aid

Set of all attributes issued to user uid and

managed by authority aid.

aid

a secret key which belongs to authority

aid.

aid

Public key which belongs to authority aid.

GSK

uid

A global secret key of a user uid. GSK is

a private key issued by the certification

authority CA.

Cert

uid

A public key certificate containing user’s

public key issued by a CA.

x.aid

A public attribute key of attribute x issued

by the authority aid

UDK

uid.aid

User Decryption key issued by authority

aid

EDK

uid.aid

EDK is an encrypted form of a UDK

which is encrypted by a user public key.

GRP

Group role parameter is a seed numbers

computed from a set of user ids of the

roles.

Secret seal which is a symmetric key

created from the AES algorithm together

with the GRP.

ACP

An access control policy used to encrypt

the data files.

SCT

A sealed ciphertext which is a ciphertext

encrypted with the SS

Here, three major operational phases including

System Setup, Key generation, Encryption, and

Decryption.

Phase 1: System Setup

In the Setup phase, we selectively describe three

algorithms as follows:

1. Create Attribute Authority(AA

)PK

aid

, PK

x.aid

The algorithm is based on the

bilinear map. It takes the attribute authority ID

(AA

) as input. It outputs the authority public

key (public parameter) PK

aid

, Secret key SK

aid

and public attribute keys PK

x.aid

for all attributes

issued by the AA

aid

2. Create GroupRole parameter GRP (GSK

uid

set of RID  GRP) The Create GroupRole

parameter algorithm takes input as a set of R

CLOSER 2017 - 7th International Conference on Cloud Computing and Services Science

532

and returns the GRP. Then, the GRP is signed

(encrypted) by AA’s private key GSK

uid

3. Create H-ACP(ACP, PK

x.aid

, H(f)), R, GSK

uid

)

 H-ACP. The algorithm consists of the

following steps.

(1) The algorithm takes the public attribute key

x.aid

to be hashed in the hash function H(f).

(2) These hash values are randomly applied with

cryptographically pseudo random R, Then, set of

randomized hash attribute value R(h(PK

x.aid

) is

obtained.

(3) These hash attribute values R(h(PK

x.aid

) are

used to construct the H-ACP as specified in the

original ACP.

(4) H-ACP is signed by the data owner’s private

key, GSK

uid

Phase 2: Key Generation

This phase consists of two algorithms as follows:

4. UserKeyGen(h(S

uid,aid

),SK

aid

,Cert

uid

)EDK

uid,aid

This algorithm takes continuous two steps as

follows:

(1) takes input as set of randomized hash values

of attributes Suid,aid associated to each hash-

based access control policy (H-ACP), attribute

authority’s secret key SK

aid

, and public key

certificate of users Cert

uid

issued by the CA

listed in the trust list, then it returns the set of

user decryption key UDK

uid.aid

(2) a UDK

uid.aid

is encrypted with the global

public key of the user and outputs the set of

encrypted decryption key EDK

uid,aid

. The

EDKuid.aid is stored in the cloud and it will be

retrieved by the user upon the access request.

Phase 3: Encryption

This phase runs our two encryption layer protocol

which accommodates data encryption and ciphertext

encryption.

5. ENC(PK

aid

{SS, GRP} M, H-ACP) CT. The

encryption algorithm performs two consecutive steps

as followings:

- Inner layer: the algorithm takes as inputs

authority public key PK

aid

, H-ACP, and data

M. Then it returns a ciphertext CT.

- Outer Layer: the algorithm takes GRP and

generates AES session key as a secret seal SS to

encrypt the ciphertext CT. It returns sealed

ciphertext SCT. Finally, a SS is encrypted with

user’s public key Cert

uid

, and stored in a cloud

server.

Phase 4: Decryption

6. DEC(PK

aid

, SCT, GSK

uid

, EDK

uid.aid

,)  M. The

decryption algorithm performs two steps as follows:

(1) Decrypt the secret seal SS. The algorithm

takes user’s global secret key GSK

uid

and

then obtains the session key to decrypt the

SCT and gets the CT.

(2) Decrypt the encrypted decryption key

(EDK

uid.aid

). The algorithm takes GSK

uid

decrypt EDK

uid.aid

. Then UDK

uid.aid

obtained to decrypt message M.

5 OUR PROPOSED POLICY

HIDING SCHEME

5.1 Overview

In this section, we propose a policy hiding scheme

that achieves policy privacy preservation and secure

policy sharing to support encryption service to users

having write permission. With our scheme, the policy

can be enforced in unreadable format; the policy

elements attached to the ciphertext is thus

anonymized. Also, hidden policies can be

conveniently shared to users having write privilege

for data encryption without the availability of data

owners. Figure 2 presents the system overview for

policy outsourcing model in Hospital Information

Systems (HIS).

Figure 2: Policy outsourcing model in HIS.

As can be seen in Figure 2, data owners initially

encrypt both data and access policies and then

outsource to a cloud. Data files are encrypted based

on the CP-ABE scheme while policy content is

annoymized by the hash-based function and random

encryption before it is encrypted with the CP-ABE.

Both encrypted files and policies are then sent to be

stored in the cloud servers. A server storing the

encrypted policies can be only accessed by the data

owners or users having write privilege. In our model,

Enforcing Hidden Access Policy for Supporting Write Access in Cloud Storage Systems

533

users are grouped by their role, such as doctor, nurse,

physicians, etc. These groups of user have different

access privileges (read or write) for different files.

Write privilege entails retrieval of encrypted policy

for data encryption and re-uploading of files to the

cloud server.

5.2 Policy Hiding

We introduce a randomized hash-based public

attribute key validation (RH-PAKV) technique to

conceal access control policy (ACP) content and

validate policy rules. Our technique combines

cryptographically pseudorandom number R (NIST,

2010) and cryptographic hash function to enable a

digest to be more secure and resistant for attacks

compared to normal hash. Generally, each attribute

authority (AA) maintains the attributes issued

together with their associated public attribute keys,

and randomized hash value. Attributes with constant

values (such as role name) will have their hashes

generated from the hash-value-pair. Computable

values (such as numeric, date/time) will only have

their attribute names hashed. Table 2 shows the

profile of hashed-based value of public attribute key.

This profile table is locally maintained by the data

owner.

Table 2: Public Attribute Keys and their hash values.

AttrID

Attribute

x.aid

h(PK

x.aid

)

0001

Medical Doctor

b2xdx4512s

hash value1

0002

Dept: General

y245xdss03

hash value2

0003

Dept:Specialized

K45xds7ws

hash value3

0004

Level

S24512sdf

hash value4

…

….

…

When the hash value of each public attribute key

(PK

x.aid

) is obtained, a random is generated by the

data owner and it will be applied to the individual

hash result of PK

x.aid

. Then, the randomized hash

value R(h(PK

x.aid

)) of each PK

x.aid

is obtained. In

essence, we combine hash-based scheme and random

number encryption for preserving the policy privacy

because of two reasons as follows.

1. Compared to other encryption algorithms,

hashing produces a lightweight output with fixed

and small size of digest.

2. Since there is no key required, it is

computationally feasible to verify the integrity

of the policy content based on the hashed data.

Figure 3 shows the hash-based policy tree equivalent

to the policy shown in Figure 1.

Figure 3: Hash-based ACP (H-ACP).

To enable a hash valued of attributes to be more

secure and highly resistant for the attack. We apply

secureRandom Java class (Java Platform

Specification:SecureRandom Number Generator,

2016) as a cryptographically secure random number

(R ) to encrypt the hash of attribute value.

5.3 Write Access Enforcement

Our solution allows the users having write privilege

can update the data file and retrieve the policy to re-

encrypt the updated file. This avoids expensive

communication cost for returning the file to be re-

encrypted by the data owners and the encrypted files

will be loaded back to the cloud server. With our

solution, data owners encrypt a set of hashed policies

H-ACPs with a simple CP-ABE policy where data

owner’s ID and a set of user IDs having write

permission are included. Then the H-ACPs are sent

to be stored in the cloud. Therefore, the permitted

users with write privilege can retrieve the policy by

using their current UDK

uid.aid

as it is used to re-

encrypt the data. In our scheme, we delegate a proxy

to perform file re-encryption initiated by the clients

(users having write privilege) as a part of our write

privilege enforcement mechanism. A proxy is a

semi-trusted server located in the cloud environment.

Even though the proxy is delegated to perform file

re-encryption, it cannot access to the plaintext as the

decryption key is not given. In our model, X.509

certificate and PKI key pair are issued to the proxy

and users for the authentication purpose. The

procedure of write privilege enforcement is shown

below.

CLOSER 2017 - 7th International Conference on Cloud Computing and Services Science

534

Write Privilege Enforcement

Input: UDK, ACP, Random R

Output: Updated file M’

Step 1: Data Access

1. A user requests to access the file.

2. A user is authenticated with her public key certificate. If

authentication is successful, the capability list

consisting data files she is able to access is presented.

3. A user downloads the chosen file and uses her

UDK

uid.aid

to decrypt and update the file.

4. After the file is updated, user signs the updated content

with her private key.

Step 2: Proxy Re-Encryption Initialization

5. User retrieves the encrypted ACP

and decrypts it by

using her UDK

uid.aid

to get the ACP.

6. The algorithm generates Random R which is used to

encrypt the updated file, and ACP. Then, it returns M

and ACP

which are the random encryption result of file

and ACP respectively.

7. User sets a passphrase to protect (encrypt) R and then

the protected R’ is generated.

8. User submits updated file M’ and ACP

, R’ to the

proxy.

Step 3: Re-Encryption

9. The proxy takes M’, R’, and ACP

for file re-encryption.

10. In the PRE process, the algorithm asks the user to enter

the passphrase to decrypt R’ and performs file re-

encryption.

11. The proxy sends the updated file M’ to be stored in the

cloud.

With our write enforcement mechanism, the users

having write privilege can update the data and

request for file re-encryption to the proxy without the

intervention of data owners.

6 ANALYSIS OF OUR PROPOSED

SCHEME

6.1 Security Analysis

For the security proof of our cryptographic construct,

we refer to the proof of our inner encryption layer

based on the CP-ABE encryption [Bethencourt et al,

2007]. Also, all access policies located in the cloud

server are also encrypted based on the CP-ABE

model.

Regarding the security protection in the re-

encryption process done by a proxy, the plain data

cannot be learned by the proxy or any parties as the

re-encryption key is encrypted with the random

number additionally protected by the passphrase

defined by the user. The statistical details of

cryptographic secure random number are given in

(NIST, 2010).

In our model, the PRE process is executed

instantly when the proxy gets the request for file re-

encryption. Therefore, our core access control model,

policy hiding and outsourcing scheme, and proxy re-

encryption method are secure.

6.2 Comparative Analysis of Access

Control Features of Existing Works

In this section, we give a comparative analysis of

policy hiding features provided by J. Lai et al.

scheme, Asghar et al. scheme, and our scheme. Both

Lai and Asghar access control scheme are based on

CP-ABE and RBAC model respectively. Thus, they

are suitable to be compared with our model based on

the integration of CP-ABE and RBAC. Table 3

presents the comparison of existing CP-ABE access

control schemes supporting policy hiding.

Table 3: Comparison of Policy Hiding Schemes.

Importantly, our scheme generally supports

multi-authority access policy enforcement and

provides write access control, while Lai et al.’s

scheme does not support these features. Regarding

the computation analysis, our scheme significantly

provides less cost for encryption than both compared

schemes since our scheme uses symmetric encryption

and general bilinear map based on CP-ABE. In

addition, our hidden policy is pre-computed based on

the hashing technique and possesses less complexity

compared to the inner predicate computation.

In Lai scheme (Lai et al., 2011), the composite

bilinear groups with the inner predicate encryption

could introduce the performance problem for

encryption, decryption, and policy update when a

policy consists of a high number of attributes and

various logical operations.

In Asghar method (Asghar et al, 2013),

encrypting the RBAC policy element could make the

communication and computation overhead high if the

policy size is big and there is frequent update on the

policy. The client-side workload is thus an issue for

this scheme.

In our scheme, an equality-based hashing scheme

yields less computation cost which is similar to the

original CP-ABE. With our scheme, data owners also

do not need to stay online to support encryption

service. Therefore, in addition to achieving policy

privacy, our solution offers secure policy retrieval for

Scheme

Write

Access

Cost of

encryption

Policy Hiding

Lai et al.

Composite Bilinear

map

CP-ABE

Inner Predicate

Asghar et al.

Yes

2 round of

encryptions + re-

encryption

RBAC encryption

Our Scheme

Yes

AES, and bilinear

maps

CP-ABE attribute

key hashing

Enforcing Hidden Access Policy for Supporting Write Access in Cloud Storage Systems

535

data encryption. Users with write permission can

retrieve the policy but they cannot learn the policy

content.

7 CONCLUSION AND FUTURE

WORK

We have presented a privacy-preserving access

control model in collaborative cloud data storage

systems. We introduce the policy hiding scheme as an

integrative solution for enhancing the capability of

our access control scheme C-CP-ARBE. The

proposed hash-based policy enforcement

compliments limitation of the traditional CP-ABE in

terms of policy privacy. Significantly, our scheme

does not require the process of de-anonymization of

policy and the encryption is done as the same as plain

policy encryption. Finally, we analyze the access

control features of related works and present the

comparative analysis of our method and two related

works.

For future works, we will conduct a larger scale

of experiments and evaluate the performance of the

proposed system in the real cloud environment such

as CloudStack. We will also investigate the cloud

forensics and auditing techniques to guarantee the

accountability of user access and integrity of the data

and policy outsourced.

REFERENCES

Bethencourt, J., Sahai, A., and Waters B., 2007.

Ciphertext-policy Attribute-based Encryption, In IEEE

Symposium of Security and privacy, SP’07, IEEE,

pages 321.-334.

Chase, M., 2007. Multi-authority attribute based

encryption, In Proceedings of the 4th Theory of

Cryptography Conference on Theory of Cryptography

(TCC’07), Springer, pages 525-534.

Nishide, T., Yoneyama, K., and Ohta, K., 2008. Attribute-

based encryption with partially hidden encryptor-

specified access structures. In Proceedings of Applied

Cryptography and Network Security, ACNS’08.

LCNS, Vol.5037, pages 111-129. Springer.

Katz, J., Sahai, A., and Waters, B., 2008. Predicate

encryption supporting disjunctions, polynomial

equations, and inner products. In Proceedings of

International Conference on the Theory and

Applications of Cryptographic Techniques, Eurocrypt

2008. LNCS, Vol 4965. pages 146-162, Springer.

Yu, S., Ren, K., and Lou., W., 2008. Attribute-based

content distribution with hidden policy. In Proceedings

of 4

Workshop on Secure Network Protocols, NPSEC

2008. IEEE.

Chase, M. and Chow, M. 2009. Improving privacy and

security in multi-authority attribute-based encryption,

In Proceedings of the 16th ACM Conference on

Computer and Communications Security, CCS’09,

pages 121-130, ACM.

NIST. "A Statistical Test Suite for Random and

Pseudorandom Number Generators for Cryptographic

Applications". NIST, Special Publication April 2010.

Wang, G., Liu, Q., and Wu, J., 2010. Hierarchical attribute-

based encryption for fine-grained access control in

cloud storage services. In Proceedings of the 17th

ACM Conference on Computer and Communications

Security,CCS '10, pages 735-737, ACM.

Zhao, F., Nishide, T., and Sakurai, K., 2011. Realizing

Fine-Grained and Flexible Access Control to

Outsourced Data with Attribute-Based Cryptosystems,

In Proceedings of 7th International Conference of

Information Security Practice and Experience,

ISPEC’11, papges 83-97, Springer.

Lai, J., Deng, R. H., and Li, Y., 2011. Fully Secure

Cipertext-Policy Hiding CP-ABE, In Proceedings of

the 7th International Conference on Information

Security Practice and Experience, ISPEC’11, pages 24-

39, Springer.

Wan, Z., Liu, J., and Deng, R. H., 2012. HASBE: A

Hierarchical Attribute-Based Solution for Flexible and

Scalable Access Control in Cloud Computing. In IEEE

Transactions on Information Forensics and Security

7(2): pages 743-754, IEEE.

Li, M., Yu, S., Zheng, Y., Ren, K, and Lou, W., 2012.

Scalable and Secure Sharing of Personal Health

Records in Cloud Computing using Attribute-based

Encryption, In IEEE Transactions on Parallel and

Distributed Systems. Volume 24, Issue: 1, pages 131-

143, IEEE.

Ruj, S., Stojmenovic, M., and Nayak, A., 2012. Privacy

Preserving Access Control with Authentication for

Securing Data in Clouds, In Proceedings of 12th

IEEE/ACM International Symposium on Cluster,

Cloud and Grid Computing, CCGrid 2012, pages 556-

563, IEEE.

Asghar, M. R., Ion, M., Russello, G., and Crispo, B., 2013.

ESPOONERBAC: Enforcing Security Policies in

Outsourced Environments, In Elsevier Jounral of

Computers & Security (COSE), Volume 35, pages 2-

24. Elsevier Advanced Technology Publications.

Yang, K., Jia, X., Ren, K., Zhang, B., Xie, R., 2014.

Expressive, Efficient, and Revocable Data Access

Control for Multi-Authority Cloud Storage. IEEE

Transastions on Parallel Distributed Systems, Vol.

25(7), pages 1735-1744, IEEE.

Fugkeaw, S. and H. Sato, 2015. An extended CP-ABE

based Access control model for data outsourced in the

cloud, In Proceedings of IEEE International Workshop

on Middleware for Cyber Security, Cloud Computing

and Internetworking, MidCCI 2015, IEEE.

Java™ Platform, Standard Edition 7

API Specification, 2016. : Secure random number

generator Java library, https://docs.oracle.com/javase/

7/docs/api/java/security/ SecureRandom.html

CLOSER 2017 - 7th International Conference on Cloud Computing and Services Science

536