This article has been accepted for publication in a future issue of this journal, but has not been

fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TCC.2015.2513388, IEEE
Transactions on Cloud Computing
JOURNAL OF LATEX CLASS FILES, VOL. 13, NO. 9, SEPTEMBER 2014 1

A Modified Hierarchical Attribute-Based Encryption Access Control

Method for Mobile Cloud Computing
Yuanpeng Xie, Hong Wen, Bin Wu, Yixin Jiang and Jiaxiao Meng
Abstract—Cloud computing is an Internet-based computing pattern through which shared resources are provided to devices on-
demand. Its an emerging but promising paradigm to integrating mobile devices into cloud computing, and the integration performs
in the cloud based hierarchical multi-user data-shared environment. With integrating into cloud computing, security issues such as
data confidentiality and user authority may arise in the mobile cloud computing system, and it is concerned as the main constraints
to the developments of mobile cloud computing. In order to provide safe and secure operation, a hierarchical access control method
using modified hierarchical attribute-based encryption (M-HABE) and a modified three-layer structure is proposed in this paper.
In a specific mobile cloud computing model, enormous data which may be from all kinds of mobile devices, such as smart phones,
functioned phones and PDAs and so on can be controlled and monitored by the system, and the data can be sensitive to unauthorized
third party and constraint to legal users as well. The novel scheme mainly focuses on the data processing, storing and accessing,
which is designed to ensure the users with legal authorities to get corresponding classified data and to restrict illegal users and
unauthorized legal users get access to the data, which makes it extremely suitable for the mobile cloud computing paradigms.

Index Terms—Mobile cloud computing, M-HABE, access control.

I. I NTRODUCTION considered as the most popular definition of mobile cloud

computing [4].

W ith explosive growth of mobile devices including smart

phones, PDAs, and tablet computers and the applica-
tions installed in them, the mobile-Internet will maintain the
2) Taking advantages of leisure resources such as CPU,
memory, and storing disks, another model of mobile
cloud computing exploits the mobile devices themselves
development growth trend as 4G communication network is as resources providers of cloud [5]. And the scheme
extensively promoted to our lives. What users of the mobile supports user mobility, and recognizes the potential of
devices and applications need is that mobile-Internet can mobile clouds to do collective sensing as well.
provide them with the service which is user-friendly, high-
In this paper, we mainly use the first paradigm mentioned
speed, and steady. In addition, the security issues of mobile
above, but the second one inspires us to assume that what
terminals and the Internet access are attached importance to.
if the mobile devices do not provide computing resources or
And as a combination of cloud computing, mobile devices and
storing resources but sensing data instead?
wireless networks, mobile cloud computing is an emerging In fact, most mobile devices are capable to capture some
but very promising paradigm which brings rich computational data from the environment nowadays, for example, almost
resources to mobile users, network operators, as well as cloud every smart phone are equipped with sensors of proximity,
computing providers [1] [2] [3]. The flaws of data storing accelerometer, gyroscope, compass, barometer, camera, GPS,
and data computing in mobile-Internet applications can be microphone [6], etc. Combining the concept of WSN, mobile
overcome by mobile cloud computing while the new paradigm devices can be regarded as mobile sensors that are able to
can also accomplish cloud based multi-user data sharing, end provide other mobile devices who are users of the mobile cloud
geographical service limitation, and process real-time tasks services with some sensing information including environment
efficiently at the same time. monitoring data, health monitoring data, and so on.
There is no accurate definition of mobile cloud comput- We take a weather monitor application as an example in
ing, several concepts were proposed, and two most popular this paper.
schemes can be described as follows: Assuming that a company develops a weather monitor
1) Mobile cloud computing is a kind of scheme which could application which aims to share real-time weather information
run an application such as a weather monitor application such as temperature, humidity, pictures, and precise location
on remote cloud servers as displayed in Figure 1, while information and so on to other users of the application. And the
the mobile devices just act like normal PCs except that application utilizes the user-cloud-user model instead of peer-
the mobile devices connect to cloud servers via 3G or to-peer model so that the users can get classified and demanded
4G while PCs through Internet. And this concept is information. Another feature of the application is that the users
are divided into different hierarchies, depending on which
Yuanpeng Xie and Hong Wen is with the National Key Laboratory of users can get different sensing data, and users with higher
Science and Technology on Communications, University of Electronic Science privilege level can, of course, get access to more specific and
and Technology, Chengdu 610054, China. E-mail: [email protected]
and [email protected]. more frequently updated information.
Bin Wu is with the School of Computer Science and Technology, Tianjin In order to meet what the application requires, security
University, Tianjin, P. R. China. E-mail: [email protected] issues of the whole system should not be ignored, among all
Yixin Jiang and Jiaxiao Meng are with the EPRI, China Southern Power
Grid Co. Ltd., Guangzhou, China. security issues the most important two security issues in such
model can be divided into two parts: authority of application

2168-7161 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TCC.2015.2513388, IEEE
Transactions on Cloud Computing
JOURNAL OF LATEX CLASS FILES, VOL. 13, NO. 9, SEPTEMBER 2014 2

In this paper, a hierarchical access control method using

a modified hierarchical attribute-based encryption (M-HABE)
Functioned Phone PDA and a modified three-layer structure [16] is proposed. Differing
Smart Phone Cloud Servers from the existing paradigms such as the HABE algorithm and
the original three-layer structure, the novel scheme mainly
focuses on the data processing, storing and accessing, which
Cloud is designed to ensure the application users with legal access
authorities to get corresponding sensing data and to restrict
illegal users and unauthorized legal users get access to the
D
data, the proposed promising paradigm makes it extremely
suitable for the mobile cloud computing based paradigm. What
D should be emphasized is that the most important highlight of
D all in the proposed paper can be described as that the modified
three-layer structure is designed for solving the security issues
D
illustrated above.
D
The rest of the paper can be organized as follows. Section
D 2 points out the security issues in mobile cloud computing
D paradigm, which can be divided into 2 portions including se-
curity issues of cloud computing and security issues of mobile
cloud computing. The proposed modified hierarchical attribute
based encryption access control method is described in section
3. Section 4 demonstrates how the proposed access control
Fig. 1. A mobile cloud computing model method based on M-HABE applies in a weather application
scenario specifically. Conclusions are given in section 5.

users and the confidentiality of sensing data. Those issues can

be solved by providing methods of access control [7]. Attribute
II. M OBILE C LOUD C OMPUTING S ECURITY I SSUES
Based Encryption (ABE) is a recent cryptographic primitive
which has been used for access control [8]–[11]. Access
More and more users are starting to use mobile cloud com-
control issue deals with providing access to authorized users
puting services such as iCloud and OneDrive services because
and preventing unauthorized users to access data. Attaching a
of the poor storage and computation capability of current
list of authorized users to each data is the simplest solution
mobile devices. However, these kind of mobile cloud services
to achieve access control. However, this solution is difficult
are considered to be vulnerable in security and users may lose
in the scenario with a large number of users, such as the
their stored files or messages such as pictures, documents,
application mentioned above within the environment of cloud.
contacts, and calendars, whats worse, those information may
Public cryptographic scheme is another solution, in which
be stolen by third parties. In September, 2014, Apple admitted
a public/secret key pair is given to each user and encrypt
that iCloud was compromised by hackers and many pictures
each message with public key of the authorized user, so that
of celebrities leaked out [17].
only the specific users are able to decrypt it. In the proposed
scenario, users with different privilege levels have different Such leakage event alarmed us that the security issues of
rights to access the part of sensing data coming from the mobile cloud should be taken seriously. For solving such
mobile devices. Therefore, one same data has to be encrypted security challenges, data authority and data confidentiality
into ciphertext once, which ought to be able to be decrypted should be paid more attention.
multiple times by different authorized users. Authority of data users: Different authority-level system to
Based on such application demands, the concept of attribute get access to sensing data for application users should be
based encryption is introduced [11]. Senders encrypt message established since the paradigm is applied in the hierarchical
with certain attributes of the authorized receivers. The ABE multi-user shared environment, which also means that the
based access control method uses several tags to mark the users with higher authority level should get all the data that the
attributes that a specific authorized user needs to possess. users with lower privilege level could get access to, while the
The users with certain tag sets can get access to the spe- lower privilege users cant get the data beyond his/her authority.
cific encrypted data and decrypt it. Lots of paper [12]–[15] Confidentiality of data: Although the cloud services utilized
introduced the scheme about the attribute based encryption in the scenario are provided by private cloud which is supposed
access control method in the cloud computing. In the mobile to be secure, it is still necessary to ensure the sensing data
loud computing environment, there are tremendous data which protected from malicious third parties that do not belong to the
needs to be processed and marked with attributions for the mobile cloud system. Therefore it is important for the system
convenient attributing access before storing. At the same time, to bring in a secure and efficient encryption scheme.
the hierarchical structure of the application users need an In this section, we mainly discuss the general cloud com-
authentication center entity to control their attributes. puting security issues and mobile cloud computing issues.

2168-7161 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TCC.2015.2513388, IEEE
Transactions on Cloud Computing
JOURNAL OF LATEX CLASS FILES, VOL. 13, NO. 9, SEPTEMBER 2014 3

A. Security Issues for Cloud Computing TABLE I

SECURITY ISSUES FOR CLOUD COMPUTING
As long as the data is transmitted to cloud, it is utilizing
cloud services like IaaS or DaaS, security challenges of which Security Descriptions
must be overcome since then. There are plenty of research Challenges
results about cloud security, in conclusion, a secure cloud Availability Cloud providers are supposed to guarantee to
consumers that they can get and use their
should at least satisfy 4 basic urges of consumers [12], say data at any places and any time.
availability, confidentiality, data integrity, control. Confidentiality Consumers’ data should be kept secret in
1) Availability cloud systems.
Cloud providers should offer services that consumers could Data Integrity The data stored in cloud systems need a
mechanism to ensure their data not lost or
get and use at any places and any time. There are mainly modified by unauthorized users.
two methods to enhance availability in cloud, which are Control A secure control system distributes
virtualization and redundancy. Currently, cloud technology is appropriate resources to be utilized in
mainly based virtual machine [13], since cloud providers can different occasions.
provide separated virtualized memory, virtualized storage, and
virtualized CPU cycles, so that users can always get them.
Large cloud provider enterprises build data centers in mul- cloud services make it difficult to overcome defects in se-
tiple regions all over the world to protect files they store from curity control because of the insufficient control mechanisms
failing in one particular region and spreading to other regions. that virtualized networks offer. And poor key management
For example, Google set three replications for each object procedures of virtualized based cloud services make it worse
stored in it [14], all these redundancy strategies are enhancing [13]. Because virtual machines don’t have a fixed hardware
the availability for consumers to get whatever they want at any infrastructure and cloud-based content is often geographically
time and any place. distributed, it is a very tough task to ensure a secure control
Besides these concerns on availability, don’t trust HTTP in cloud.
protocol too much as it is a stateless protocol for attackers, We conclude the discussion in Table I.
which may cause unauthorized access to the management
interface of cloud infrastructures [13]. B. Security Issues for Mobile Cloud Computing
2) Confidentiality Mobile cloud computing model in this paper means that
Confidentiality has been a huge barrier for cloud providers mobile device users run applications on remote cloud servers
to popularize cloud to consumers since it comes out. It is instead of mobile devices themselves, the paradigm performs
understandable that consumers cannot trust the cloud services, almost the same as normal cloud computing with computers
after all, nobody knows what will happen to the files, espe- except that mobile cloud model connects mobile devices
cially important and confidential ones, once they are placed in and cloud servers through 3G or 4G while cloud computing
cloud vendors’ hosts. paradigm via Internet, therefore, mobile cloud computing
There basically exist two common approaches in current inherits the security threats of traditional cloud computing.
cloud infrastructures, say physical isolation and encryption. Whats more, the security issues that are specific to mobile
Physical isolation specifically means virtual physical isolation devices such as battery exhaustion attacks [20] mobile bonnets
as cloud services are transmitted via public networks. In this and targeted attacks [21] should be concerned as well [1] .
context, virtual physical isolation [15], [18] are using VPN
and firewalls to secure database [12]. Encrypting vital and III. M ODIFIED H IERARCHICAL ATTRIBUTE - BASED
confidential data before placing it in cloud infrastructures is E NCRYPTION (M-HABE)
another method to enhance confidentiality of cloud. But do We take a weather application on mobile devices as a
not count on that approach too much because novel methods scenario.
of breaking cryptographic algorithms are discovered [13]. As the mobile cloud computing defines [1] [22], there would
3) Data integrity be so much sensing data from the mobile devices inbursting
Data integrity ensures consumers that their storing data is into the cloud infrastructures to process and store the data.
not modified by others or collapsing owing to system failure. The sensing data belonging to a mobile cloud computing
An easy method is making plenty of copies of consumers’ model can contain information of different hierarchies such
files, which is a good but highly-cost way. Besides the method, as temperature and humanity numbers, the weather changing
a ”cloud security capture application” [19] could be in use to trend, information update frequency and so on. It is important
show consumers when and where their data was modified or that the users with lower privilege cannot get access to
transmitted. some information that the higher privilege user can get to,
4) Control while the higher authority user can get access to all the
It is a sophisticated work to control a cloud system, a data that is obtainable for users in lower hierarchical position
controlling work mainly includes deciding what resource could since different users of the mobile cloud computing system
be utilized in what occasions. constitute a hierarchical authority system. At the same, all the
In order to own a secure control system, cloud vendors information should be encrypted appropriately since the data
may need a specialized operating system. Virtualization based is not supposed to be available for a third party which doesnt

2168-7161 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TCC.2015.2513388, IEEE
Transactions on Cloud Computing
JOURNAL OF LATEX CLASS FILES, VOL. 13, NO. 9, SEPTEMBER 2014 4

TABLE II A. Related Work

L IST OF THE ACCESS S TRUCTURE IN A M OBILE C LOUD C OMPUTING
M ODEL 1) Hierarchical identity-based encryption
The concept of Identity Based Encryption (IBE) was pro-
Information Access Structure
Type posed by Shamir [11] first in 1984, differing from traditional
Temperature All users
symmetrical encryption system, IBE took arbitrary character
Value strings that can represent the identities of users, such as ID
Weather Users of level 2, users in the cooperative numbers, e-mail addresses, as public keys to encrypt data. One
Horizontal company{××meteorology company, . . . } advantage of IBE is that the sender didnt have to search the
Comparison
public keys information on certificate authority (CA) online,
Update Users of level 3, users with titles {professor,
Frequency doctor, major journalist, . . . }
which solved the problem of poor CA performance. The
Weather Users of level 4, users with designated work
shortage of IBE system was that all users keys were generated
Vertical numbers {290381, 209378, 98302, . . . }, by the private key generation (PKG), which would become the
Comparison users doing jobs of {meteorology researcher, bottleneck in the system.
journalist, . . . }
Horwitz [23] proposed the idea of hierarchical IBE (HIBE)
Location Users of level 5, users in specific locations
Accuracy in 2002, a user in the higher hierarchical position of the
system could create private keys for lower position users with
his/her private keys. Which mean that only the first level
users private keys need be created by PKG, while lower-level
belong to the system. So a secure and hierarchical access users private keys could be generated and managed by their
control method should be proposed to apply in the mobile ancestors. This improved system relieved PKG of great burden
cloud computing system. An example of the access control and enhanced the system efficiency by authenticating identities
list in such circumstance is shown in Table II. and transporting keys within locality area instead of global
As the list suggests, the access structure should meet the area.
following requirements: The public key of a user is described by a set of IDs
composed of the public key of father node and the users own
a) One encrypted data can be received by several users.
ID in the method of G-HIBE [9], the most important feature of
b) Not only precise level descriptions, but users attributes are
the proposal is that the users public key could reflect precise
there in the access structure. For example, weather vertical
position of the user in the hierarchical structure.
comparison information can be attainable for users of level
2) Ciphertext-policy attribute-based encryption
4, some users with working numbers {290381, 209378,
98302, . . . }, and users doing jobs such as meteorology Attribute based encryption (ABE) [24] is regarded as the
researcher, journalist and so on, among which the users IBE method with an access structure bringing into the ci-
with specific numbers and users doing specific jobs are phertext or private key, the access structure determines what
described as the attributes while the other one is described ciphertext can be obtained by which users.
as an accurate privilege level. Two major branches of ABE system are key-policy ABE
c) The structure of encryption keys should performs just as the (KP-ABE) [24] and ciphertext-policy ABE (CP-ABE) [10], the
hierarchical structure of the mobile cloud computing users. later one is utilized in many paradigms including this proposed
For instance, the way the authentication center within the paper. The access structure mentioned above in CP-ABE is
mobile cloud computing application company distributes placed in ciphertext, which means that the data sender can be
the keys to users should just be as how the users privileges so initiative that he/she can determine the receiver. Users are
show. described by a set of attributes in CP-ABE, only when the
attribute set satisfies the access structure can the user obtains
In order to accomplish these requirements, a proposed the ciphertext.
access control method should contain the following features: The core of the proposed scheme is called modified hi-
erarchical attribute-based encryption (M-HABE), which is
a) One ciphertext can be decrypted by several keys.
different from the HABE scheme.
b) Both precise level description and user attribute should be
HABE was proposed based on G-HIBE [9] and CP-ABE
supported in the access structure of the method.
[10] by Wang [8] in 2010, it was designed mainly for the usage
c) The keys in the authentication center ought to have the
within an enterprise. We modified the proposal to adapt the
same hierarchical structure just as the structure of users
scenarios of mobile cloud computing system, which could be
privilege levels.
illustrated as figure 2, with the aim of making it accommodate
A modified hierarchical attribute-based encryption (M- to the system based on mobile cloud computing.
HABE) access control method applied in mobile cloud com- As the Figure 2 shows, the proposal consists of an au-
puting is proposed in this paper, which changes a proposed thentication center (AuC), Sub-AuCs, and application users.
scheme called hierarchical attribute-based encryption HABE The AuC is responsible for generating and publishing system
[8], M-HABE combines the hierarchical identity-based en- parameter and the system master key; Sub-AuCs can be
cryption [9] and the ciphertext-policy attribute-based encryp- divided into first-level Sub-AuC(Sub-AuCi ) and other Sub-
tion (CP-ABE) [10] to meet the conditions described above. AuCs, among which the AuC just need to be in charge of

2168-7161 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TCC.2015.2513388, IEEE
Transactions on Cloud Computing
JOURNAL OF LATEX CLASS FILES, VOL. 13, NO. 9, SEPTEMBER 2014 5

to manage, while they also create users’ secret identity keys

Authentication
AuC
Center SKi,u and data users’ secret attribute keys SKi,u,a at the
same time.
Sub-AuC 1 c) Each data consumer is described by one precise ID denot-
Sub-AuC 2 Sub-AuC 2 ed as IDu , and a set of data users’ attributes represented
as {a}. Besides these, each user also owns a user public
A A A A A A A A
Key
key P Ku denoted as {P K∗ , IDu } and a consumer secret
distribution
key SKu , a set of user secret identity key {SKi,u } and a
U U
Management
set of consumer secret attribute key {SKi,u,a }.
Authentication Center
d) Each attribute a is described by a precise ID denoted as
U U First-level
Sub-Authentication Center
IDa . And even an attribute owns a public key in the form
Second-level
U Sub-Authentication Center of (P Ki−1 , IDi ) where P Ki is the public key of Sub-AuC
U U U Data User
that takes charge of the attribute.
A Attribute
Data Users

C. M-HABE Definition
Fig. 2. M-HABE model
The M-HABE is composed by the following algorithm-
TABLE III
s:
L IST OF MAJOR KEYS IN HABE Setup: Given a security parameter K that is huge enough,
AUC will generate a system parameter params and a root
Key Name Meaning
master key M K0 .
M K0 Root key, owned by AuC
CreateMK: Using system parameter params and their own
M K∗ Master key, owned by Sub-AuC
master keys, AUC or Sub-AuCs can create master keys for
P K∗ Public key, owned by Sub-AuC1
lower-level Sub-AuCs.
P Ki Public key, owned by Sub-AuCs
CreateSK: With its own master key M K∗ and system param-
M Ki Master key, owned by Sub-AuCs
eter params, Sub-AuC1 creates secret key SKu for each
P Ku Public key, owned by users
consumer if it is sure that the public key of the user is P Ku ,
SKu Secret key, owned by users
or there would be no secret key for the user.
SKi,u Secret identity key, owned by users
CreateUser: Sub-AuCs will create users’ secret identity keys
SKi,u,a Secret attribute key, owned by users
SKi,u and secret attribute keys SKi,u,a for them if the Aub-
P Ku Public key, owned by attributes
AuC makes sure that the attribute a is in charge of it and
the user u satisfies a. And if not there would be no secret
identity keys or secret attribute keys.
users and create their private keys, while other Sub-AuCs take
Encrypt: With R denoting a set of users’ IDs, A representing
charge of users attributes and create their secret identity keys
the attribute-based access structure, the pubic keys of all the
and secret attribute keys for users.
users that are in R, and the public keys of all the attributes
Each data user shown in the figure possesses a unique ID
that are in A, the data provider, which is also a data user of
which is a character string designed to describe the features
the cloud computing in this case, can encrypt the sensing
of internal parties within the system, and so do AuC, Sub-
data D into ciphertext C.
AuCs, and users attributes, especially, the ID of each user
RDcrypt: Given the ciphertext C, a data user possessing the
contains an integer for describing the privilege level of the
precise ID that is in R can decrypt the ciphertext C into
user. Additionally, data users also own a set of attributes while
plaintext D with params and the user’s secret key SKu .
other internal parties do not.
ADcrypt: Given the ciphertext C, a data user possessing an
attribute set {a} that satisfies A, which means that the
B. Key Description consumer owns at least an attribute key SKi,u,a , can also
Public key encryption is utilized in the proposed system, decrypt the ciphertext C into plaintext D with system
the related keys are summarized in Table III. parameter params, the user’s secret identity key SKi,u ,
a) Root key M K0 possessed by AuC is used to create M K∗ and the secret attribute key SKi,u,a .
for Sub-AuC1 .
b) Each Sub-AuC owns a public key P Ki and a master key
D. M-HABE Construction
M Ki , among which P Ki is composed as (P Ki−1 , IDi )
where P Ki−1 is the public key of the Sub-AuC’s father Assuming that IG is a BDH parameter generator, the M-
node, and M Ki is also created by the father node. P K∗ is HABE scheme based on bilinear map [9] is constructed by
the public key of Sub-AuC1 , which can be demonstrated as following algorithms:
ID∗ meaning that it is composed by its own IDs. Unlike Step 1. Setup(K)→(params, M K0 ): The AUC
HABE proposed by Wang [8], Sub-AuC1 in this paper only firstly chooses the root master key mk0 ∈Zq∗ ,
needs to take charge of users, and create their secret keys and then outputs the system parameter
SKu for them. And other Sub-AuCs have a set of attributes params=hq, G1 , G2 , ê, n, P0 , Q0 , H1 , H2 i, among which

2168-7161 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TCC.2015.2513388, IEEE
Transactions on Cloud Computing
JOURNAL OF LATEX CLASS FILES, VOL. 13, NO. 9, SEPTEMBER 2014 6

(q, G1 , G2 , ê) is the output of IG, n is a positive integer, Uu1 =rPu1 , . . . , Mum =rPum ; (2)
P0 is a random generator of G1 , H1 : {0, 1}∗ →G1 and U12 =rP12 , . . . , U1t1 =rP1t1 ; (3)
H2 : G2 →{0, 1}n are two random oracles. In this step, the Xn1
system parameter is able to be obtained publicly while the U1 =r Pa1j ; . . . ; (4)
master key M K0 is kept secret. j=1
Step 2. CreateM K(params, M Ki , P Ki+1 )→(M Ki+1 ): UN 2 =rPN 2 , . . . , UN tN = rPN tN ; (5)
Assuming that Sub-AuCi is the father node of XnN
Sub-AuCi+1 . And the master key of Sub-AuCi+1 UN =r PaN j ; (6)
which is created by Sub-AuCi is in form of j=1
M Ki+1 =(mki+1 , SKi+1 , Q-tuplei+1 , HA ), among V =D⊕H2 (ê(Q0 , rnA P∗ )) (7)
which

• mki+1 is a random element belonging to Zq∗ .

• SKi+1 =SKi +mki Pi+1 ∈ G1 , where mki is part of • Let CD = [U0 , Uu1 , . . . , Uum , U12 , . . . , U1t1 , U1 , . . . ,
M Ki , Pi+1 =H1 (P Ki+1 )∈G1 . Especially, SK0 is the UN 2 , . . . , UN tN , UN , V ], and the ciphertext CT =
generator of G1 if the father node is AUC, which [R, A, CD ]
means that i=0.
• Q-tuplei+1 =(Q-tuplei , Qi+1 ), where Step 6. RDcrypt(params, IDu , SKu , CT )→D: The user
Qi+1 =mki+1 P0 ∈G1 . ui can get access to the ciphertext and decrypts it into
• HA : {0, 1}∗ →Zq∗ is a random oracle. the plaintext D if the ID of ui is in R, the verification is
shown in Eq.(8):
Step 3. CreateSK(params, M K∗ )→(SKu ):
System parameter params and master key
M K∗ are used by AuC to create managers 
ê(nA M0 , SK∗ +mk∗ Pui )

secret key SKu =(Q-tuple∗ , SK∗ +mk∗ Pu ), where V ⊕H2
ê(nA Q∗ , Mui )
Pu =H1 (P Ku )∈G1 .
Step 4. CretaeU ser(params, M Ki , P Ku , P Ka )→ 
ê(rnA P0 , mk0 P∗ +mk∗ Pui )

(SKi,u , SKi,u,a ): Suppose a user u asks for the = V ⊕H2
ê(nA Q∗ , rPui )
attribute key about attribute a, what the Sub-AuC
does firstly is to check if the attribute a is in its 
ê(rnA P0 , mk0 P∗ )ê(rnA P0 , mk∗ Pui )

charge, and outputs SKi,u =(Q-tuplei−1 , mki mku P0 ), = V ⊕H2
ê(nA Q∗ , rPui )
SKi,u,a =SKi +mki mku Pa if it is, where
mku =HA (P Km )∈Zq∗ , Pa =H1 (P Ka )∈G1 , or the 
ê(mk0 P0 , rnA P∗ )ê(nA mk∗ P0 , rPui )

algorithm will output ’null’. = V ⊕H2
ê(nA Q∗ , rPui )
Step 5. Encrypt(params, R, A, P Ka , P Ku , D)→CT :
Given a set of ID R= {IDu1 , . . . , IDum } and a DNF 
ê(mk0 P0 , rnA P∗ )ê(nA Q∗ , rPui )

access control structure A= ∨N N ni = V ⊕H2
i=1 (CCi )= ∨i=1 (∧j=1 aij ), ê(nA Q∗ , rPui )
where N is the number of conjunctive clause in A, ni
is the number of attribute in the i-th conjunctive clause = V ⊕H2 (ê(Q0 , rnA Pui ))
CCi , and aij is the j-th attribute in CCi . Assume that
all the attributes in CCi are managed by the ti -th level =D
Sub-AuC denoted as Sub-AuCiti whose public key is (8)
(IDit1 , . . . , IDiti ), where IDik , for 1 ≤ k ≤ ti , is the Step 7. ADcrypt(params, SKi,u , SKi,u,a , CT )→D: Given
ID of Sub-AuCiti ’s ancestor, and IDi1 is the ID of AuC, the ciphertext C, a user possessing an attribute set {a} that
which is ID∗ . The following steps demonstrate how to satisfies A, which means that the user owns at least an
encrypt sensing data D by senders: attribute key SKi,u,a , can also decrypts the ciphertext C
into plaintext D with params, the user’s secret identity key
• Compute P∗ =H1 (P K∗ )∈G1 . SKi,u , and the secret attribute key SKi,u,a , the verification
• For 1 ≤ i ≤ m, compute Pui =H1 (P Ku1 )∈G1 . is shown in Eq.(9):
• For 1 ≤ j ≤ ti , compute
Pij =H1 (IDi1 , . . . , IDij )∈G1 .
• For 1 ≤ i ≤ N and 1 ≤ j ≤ ni , compute
Paij =H1 (IDi1 , . . ., IDit1 , IDaij )∈G1 . V ⊕H2  
• Choose a random number r∈Zq∗ , assume that nA is the n A
ni
X
lowest common multiple (LCM) of n1 , . . . , nN , and ê U0 , SK(iti ,u,aij ) 
ni j=1 !
compute formula (1)-(7):
 ti
Y
nA
ê mku mkit1 , Ui ê(Mij , nA Qi(j−1) )
U0 =rP0 ; (1) ni j=2

2168-7161 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TCC.2015.2513388, IEEE
Transactions on Cloud Computing
JOURNAL OF LATEX CLASS FILES, VOL. 13, NO. 9, SEPTEMBER 2014 7

= V ⊕H2 PDA Authentication Center

Classified Stored Data A Sub-AuC
ê(U0 , nA SKi1 ) Functioned Phone

 Y ti
× Smart Phone
Cloud Servers First Level Sub-AuC U Data User

nA A Attribute AuC
Authentication
Center

ê SKiti ,u , Ui ê(Uij , nA Qi(j−1) ) Cloud

ni j=2
Sub-AuC 1
Sub-AuC 2
Sub-AuC 2
 ! A A
ti ni D
X n A
X D
A A A A A A A A

ê U0 , nA mki(k−1) Pik + mkitj mku Paij  D D

ni j=1
D
D D
U U

k=2 D
D
U U
D
D U

= V ⊕H2 Layer 1 Layer 2 Layer 3

U U
  Data Users
nA
ê(M0 , nA rPi1 )ê SKiti ,u , Ui
ni
ti
×

nA
Y Fig. 3. General structure of M-HABE access control method for mobile cloud
ê SKiti ,u , Ui ê(Uij , nA Qi(j − 1)) computing
ni j=2
ti
!
Y
ê(Qi(k−1) , nA Uik ) kinds of mobile devices, for example, JPEG, WMA,
k=2 TXT,PNG,WMV,etc. M obiledeviceID is the only sign
of the source mobile device where raw weather data
= V ⊕H2 (ê(Q0 , nA rP∗ )) comes from. The size of sensing weather data is defined
=D (9) by the raw weather data itself, which indicates the size
of one specific weather data. As for time, as long as a
Note 1: In order to improve the performance of the whole mobile device captures data from the environment where
scheme, data user can send Q-tuplei(ti −1) to cloud to explicit it is in, the time that the sending action occurs will be
the great Qcomputation ability with the aim of helping to regarded as the time attribute of the raw sensing data. A
ti
compute j=2 ê(Mij , nA Qi(j−1) ), so that the data user can value sign represents the most important characteristic
decrypt the ciphertext with computing bilinear maps only of sensing data, the meaning it stands for differs
constant times when running the ADcrypt algorithm. from format to format, and different kinds of mobile
Note 2: The public key of data consumer u (ID∗ , IDu ) devices have different meanings. For example, for a
can be combined with the public key of attributes temperature sensor, the value means specific numbers
(ID∗ , . . . , IDi , IDa ) in form of a character string as the input of the temperature, while the humidity sensors can only
of H1 and HA . produce the data with the value attributes that indicate
the specific numbers of humidity. A period identification
IV. M-HABE ACCESS C ONTROL M ETHOD A PPLIED IN is a time cycle of the sensing data, it is utilized to
C LOUD - BASED S MART G RID describe the life period of one specific sensing raw data,
Applying M-HABE, the proposed scheme is illustrated in and the data will be destroyed once the storing time in
Figure 3. The whole system works as following steps: cloud of it is beyond the period time.
• All kinds of mobile devices which are installed with the • The sensing weather data is encrypted into ciphertext
mobile cloud computing based weather application are in layer 2 by M-HABE encryption algorithm using the
distributed into different locations all over the country key in form of (R, A, P Ka |a∈A, IDu ∈R),Ł and the
with users. The applications can exploit the sensors cyphertext is sent to layer 3 which is also a kind of
installed in the mobile devices to capture the weather data IaaS cloud service in cloud. The encryption step can be
that the applications need, including temperature value, demonstrated as Figure 5.
humidity information, atmospheric pressure and so on. • The data users of the scheme are in charge of just
• The sensing weather data is transported to the layer1 like Figure 2 indicates. The users can get access to
which is a kind of IaaS cloud service provided by the the ciphertexts only if he/she satisfies the requirements
cloud provider [25]. of RDcrypt algorithm or ADcrypt algorithm that are
• Before sent to layer 2, the sensing weather data is described in part III. The decryption procedure is shown
classified by its data model [16] in layer 1 with its in Figure 6.
excellent ability of computing and storing, the step
can be illustrated by figure 4. The data model we V. C ONCLUSION
present is inspired by the data model proposed in The paper proposed a modified HABE scheme by tak-
[26], based on which our data model is composed ing advantages of attributes based encryption (ABE) and
by format, device ID, size, time, value and period. hierarchical identity based encryption (HIBE) access control
Therefore, a raw data can be expressed as a vector Data processing. The proposed access control method using M-
hf ormat; mobiledeviceID; sizeltimelvalue; periodi. HABE is designed to be utilized within a hierarchical multi-
F ormat stands for the basic format of the raw weather user data-shared environment, which is extremely suitable for
data that a specific mobile device produces, there a mobile cloud computing model to protect the data privacy
are different kinds of formats depended on different and defend unauthorized access. Compared with the original

2168-7161 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TCC.2015.2513388, IEEE
Transactions on Cloud Computing
JOURNAL OF LATEX CLASS FILES, VOL. 13, NO. 9, SEPTEMBER 2014 8

R EFERENCES
Cloud
[1] N. Fernando, S. W. Loke, and W. Rahayu, “Mobile cloud computing:
A survey,” Future Generation Computer Systems, vol. 29, no. 1, pp.
Data Model 84–106, 2013.
D D
[2] S. Abolfazli, Z. Sanaei, E. Ahmed, A. Gani, and R. Buyya, “Cloud-
Data Model
D D based augmentation for mobile devices: motivation, taxonomies, and
D
Data Model
D open challenges,” Communications Surveys & Tutorials, IEEE, vol. 16,
D
Data Model
D
no. 1, pp. 337–368, 2014.
[3] R. Kumar and S. Rajalakshmi, “Mobile cloud computing: Standard
approach to protecting and securing of mobile cloud ecosystems,” in
Layer 1 Layer 2 Classified Stored Data Computer Sciences and Applications (CSA), 2013 International Confer-
Cloud Servers
ence on. IEEE, 2013, pp. 663–669.
[4] J. Carolan, S. Gaede, J. Baty, G. Brunette, A. Licht, J. Remmell,
L. Tucker, and J. Weise, “Introduction to cloud computing architecture,”
White Paper, 1st edn. Sun Micro Systems Inc, 2009.
Fig. 4. Data model
[5] E. E. Marinelli, “Hyrax: cloud computing on mobile devices using
mapreduce,” DTIC Document, Tech. Rep., 2009.
[6] Q. Han, S. Liang, and H. Zhang, “Mobile cloud sensing, big data, and 5g
Cloud
networks make an intelligent and smart world,” Network, IEEE, vol. 29,
no. 2, pp. 40–45, 2015.
[7] I. Stojmenovic, “Access control in distributed systems: Merging theory
D D with practice,” in Trust, Security and Privacy in Computing and Com-
D
munications (TrustCom), 2011 IEEE 10th International Conference on.
D
D IEEE, 2011, pp. 1–2.
D D
D
[8] G. Wang, Q. Liu, and J. Wu, “Hierarchical attribute-based encryption for
D D fine-grained access control in cloud storage services,” in Proceedings of
D
the 17th ACM conference on Computer and communications security.
Layer 1 Layer 2
ACM, 2010, pp. 735–737.
Classified Stored Data
[9] C. Gentry and A. Silverberg, “Hierarchical id-based cryptography,” in
Cloud Servers
Advances in cryptologyASIACRYPT 2002. Springer, 2002, pp. 548–566.
[10] J. Bethencourt, A. Sahai, and B. Waters, “Ciphertext-policy attribute-
based encryption,” in Security and Privacy, 2007. SP’07. IEEE Sympo-
sium on. IEEE, 2007, pp. 321–334.
Fig. 5. Encryption procedure [11] A. Shamir, “Identity-based cryptosystems and signature schemes,” in
Advances in cryptology. Springer, 1985, pp. 47–53.
[12] M. Zhou, R. Zhang, W. Xie, W. Qian, and A. Zhou, “Security and
Cloud Servers
privacy in cloud computing: A survey,” in Semantics Knowledge and
M Data User Grid (SKG), 2010 Sixth International Conference on. IEEE, 2010, pp.
105–112.
[13] B. Grobauer, T. Walloschek, and E. Stöcker, “Understanding cloud
U U computing vulnerabilities,” Security & privacy, IEEE, vol. 9, no. 2, pp.
50–57, 2011.
U U [14] S. Ghemawat, H. Gobioff, and S.-T. Leung, “The google file system,”
in ACM SIGOPS operating systems review, vol. 37, no. 5. ACM, 2003,
U pp. 29–43.
[15] M. Zhou, R. Zhang, W. Xie, W. Qian, and A. Zhou, “Security and
U U privacy in cloud computing: A survey,” in Semantics Knowledge and
Grid (SKG), 2010 Sixth International Conference on. IEEE, 2010, pp.
Data Users
105–112.
Layer 3
[16] Y. Xie, J. Zhang, G. Fu, H. Wen, Q. Han, X. Zhu, Y. Jiang, and
X. Guo, “The security issue of wsns based on cloud computing,” in
Fig. 6. Decryption procedure Communications and Network Security (CNS), 2013 IEEE Conference
on. IEEE, 2013, pp. 383–384.
[17] R. Walters, “Cyber attacks on us companies in 2014,” Heritage Foun-
dation Issue Brief, no. 4289, 2014.
HABE scheme, the novel scheme can be more adaptive for [18] A. Fox, R. Griffith, A. Joseph, R. Katz, A. Konwinski, G. Lee,
mobile cloud computing environment to process, store and D. Patterson, A. Rabkin, and I. Stoica, “Above the clouds: A berkeley
access the enormous data and files while the novel system view of cloud computing,” Dept. Electrical Eng. and Comput. Sciences,
University of California, Berkeley, Rep. UCB/EECS, vol. 28, p. 13, 2009.
can let different privilege entities access their permitted data [19] L. Sumter, “Cloud computing: security risk,” in Proceedings of the 48th
and files. The scheme not only accomplishes the hierarchical Annual Southeast Regional Conference. ACM, 2010, p. 112.
access control of mobile sensing data in the mobile cloud [20] B. R. Moyers, J. P. Dunning, R. C. Marchany, and J. G. Tront, “Effects
computing model, but protects the data from being obtained of wi-fi and bluetooth battery exhaustion attacks on mobile devices,” in
System Sciences (HICSS), 2010 43rd Hawaii International Conference
by an untrusted third party. on. IEEE, 2010, pp. 1–9.
[21] J. Oberheide and F. Jahanian, “When mobile is harder than fixed (and
vice versa): demystifying security challenges in mobile environments,”
ACKNOWLEDGMENT in Proceedings of the Eleventh Workshop on Mobile Computing Systems
& Applications. ACM, 2010, pp. 43–48.
The work is sponsored by EPRI, CSG research found- [22] W. Zhang, Y. Wen, and H.-H. Chen, “Toward transcoding as a service:
ing, NSFC (No. 61271172, 61572114), 863 High Tech- energy-efficient offloading policy for green mobile cloud,” Network,
nology Plan (Grant No. 2015AA01A707), NSFC A3 Pro- IEEE, vol. 28, no. 6, pp. 67–73, 2014.
[23] J. Horwitz and B. Lynn, “Toward hierarchical identity-based encryption,”
gram (No.61140320) and RFDP (Grant No. 20120185110030, in Advances in CryptologyEUROCRYPT 2002. Springer, 2002, pp. 466–
20130185130002). 481.

2168-7161 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TCC.2015.2513388, IEEE
Transactions on Cloud Computing
JOURNAL OF LATEX CLASS FILES, VOL. 13, NO. 9, SEPTEMBER 2014 9

[24] V. Goyal, O. Pandey, A. Sahai, and B. Waters, “Attribute-based encryp- Jiaxiao Meng received the B.S. degree in Traffic
tion for fine-grained access control of encrypted data,” in Proceedings and Transportation from Central South University
of the 13th ACM conference on Computer and communications security. of Forestry and Technology, Changsha, Chian, in
Acm, 2006, pp. 89–98. 2004. He is currently pursuing the M.D. degree in
[25] M. Armbrust, A. Fox, R. Griffith, A. D. Joseph, R. Katz, A. Konwinski, Telecommunications engineering at Beijing Univer-
G. Lee, D. Patterson, A. Rabkin, I. Stoica et al., “A view of cloud sity of Posts and Telecommunications. From 2012
computing,” Communications of the ACM, vol. 53, no. 4, pp. 50–58, to 2015, he was a Research Assistant with Electric
2010. Power Research Institute,CSG, Guangzhou, China.
[26] T.-D. Nguyen and E.-N. Huh, “An efficient key management for His research interest includes the Network Security
secure multicast in sensor-cloud,” in Computers, Networks, Systems and Cloud Computing.
and Industrial Engineering (CNSI), 2011 First ACIS/JNU International
Conference on. IEEE, 2011, pp. 3–9.

Yuanpeng Xie received the bachelor degree in

communication engineering from the University of
Electronic Science and Technology of China in
July 2013. He is currently studying for his master
degree in communication and information system
in National Key Laboratory in Communication at
University of Electronic Science and Technology of
China. His research interests include the security
issues in cloud computing, the integration of WSNs
and cloud computing, and security issues in smart
grid.

Hong Wen received the M.Sc. degree in Elec-

trical Engineering from Sichuan University, P. R.
China, in 1997. She pursued her Ph.D. degree in
Communication and Computer Engineering Dept. at
the Southwest Jiaotong University (Chengdu, P. R.
China). Then she worked as an associate professor
in the National Key Laboratory of Science and
Technology on Communications at UESTC, P. R.
China. From Jan. 2008 to Sept. 2009, she was a
visiting scholar and postdoctoral fellow in the ECE
Dept. at University of Waterloo. Now she holds the
professor position at UESTC, P. R. China. Her major interests focus on
wireless communication system and security.

Bin Wu received his Ph.D. degree in Electrical and

Electronic Engineering from The University of Hong
Kong (Pokfulam, Hong Kong) in 2007. He worked
as a postdoctoral research fellow from 2007-2012 in
the ECE Dept. at University of Waterloo (Waterloo,
Canada). He is now a professor in the School of
Computer Science and Technology at Tianjin Uni-
versity (Tianjin, P. R. China). His research interests
include computer systems and networking as well as
cloud computing.

Yixin Jiang received the Ph.D. degree in Computer

Science from Tsinghua University in 2006. From
2011, he was a senior researcher of EPRI, China
Southern Grid. His research interests include smart
grid, security and performance evaluation in wire-
less communication and mobile computing. He has
published more than 80 papers in research journals
and IEEE conference proceedings in these areas.

2168-7161 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.