SECURITY ANALYSIS OF AUTHENTICATION SCHEMES IN
M-COMMERCE BASED ON FUZZY COMPREHENSIVE
EVALUATION*
Rui Hua, Runtong Zhang and Dandan Li
Institute of Information Systems, Beijing Jiaotong University, Beijing 100044, China
Keywords: Identity authentication, M-commerce, Fuzzy comprehensive evaluation, Security analysis.
Abstract: The openness and transformation mode of the mobile network poses a serious problem of information
security in mobile commerce applications, and identity authentication is one of the main methods to solve
the problem. Along with the development of mobile commerce, considerable attentions on various identity
authentication schemes have been paid. However, there is still lack of a systematic and practical evaluation
system to evaluate these schemes. In this paper, an indicators analysis system of mobile commerce by using
the fuzzy inference approach is proposed to evaluate different mobile commerce identity authentication
schemes. Comprehensive simulation and comparsion show that the proposed indicator analysis systesm is
quantified and efficient.
*
This work is partially supported by the Fundamental Research Funds for the Central Universities (with contract number
2009YJS034) and the Special Funds of Beijing Jiaotong Universities (with contract number 2009JBZ011)
1 INTRODUCTION
M-commerce activities depend on wireless Internet
and mobile terminal. Based on characteristics of
wireless network, security of mobile commerce
platform has become the key problem (Minglei S,
2009). Features such as small storage, low
computational power supply and limited battery of
mobile terminal increase the difficulty to solve the
security problem of mobile platform (Kwok-Yan L,
2003). Limited to wireless networks and
characteristics of mobile terminal, developed
security system in traditional electronic commerce
cannot be directly applied in mobile business, which
brings mobile commerce a lot of dangers. Security is
a primary issue in the development of mobile
commerce, and identity authentication is the core of
entire security problem (Neuman B and Ts’T, 1994).
Identity
authentication is the use of one or more
mechanisms to prove that you are who you claim to be
(
Aloul F, 2009). In the absence of user identity
authentication, confidentiality and integrity of user's
information is of no significance. Currently, some
researchers have improved some traditional identity
authentication models for mobile environment (Oh
HG, 2009 ; Soleymani B, 2009 and Xuefei C, 2009).
These models use different key agreement method
and encryption algorithm, and each of them have
different emphases and defects. But there are only a
few of methods to evaluate these models security,
such as formal analysis method and simple
comparison. The lack of a complete and effective
security evaluation system makes the security
performance of existing models cannot be
comprehensive evaluated objectively. It becomes
more difficult for researchers and users to choose
authentication models.
According to characteristics of mobile terminal
and wireless network, we establish a scientific and
reasonable security analysis system based on mobile
commerce environment. The fuzzy comprehensive
evaluation method is used for modeling. To meet the
demand of comparing different models’ security, the
quantitative and qualitative analysis method will be
used to evaluate identity authentication models.
2 A BRIEF REVIEW
There are two mainly methods to evaluate the
245
Hua R., Zhang R. and Li D.
SECURITY ANALYSIS OF AUTHENTICATION SCHEMES IN M-COMMERCE BASED ON FUZZY COMPREHENSIVE EVALUATION.
DOI: 10.5220/0003268402450251
In Proceedings of the Twelfth International Conference on Informatics and Semiotics in Organisations (ICISO 2010), page
ISBN: 978-989-8425-26-3
Copyright
c
2010 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
security of identity authentication models. One is
formal security analysis; another one is building
security evaluation system.
Formal security analysis is a kind of methods
based on Dolev-Yao model (Hai-yan Q, 2008). This
model hypothesizes encryption is "perfect", and
attacker can intercept any text. Formal security
analysis includes strand spacer method (Javier T,
1999), BAN method (Burrows M, 1990) and
inductive approach method (Lawrence, 1998). These
methods can prove whether the identity
authentication scheme is security in ideal condition.
But actual application environments of the identity
authentication are not in ideal condition, so these
methods can prove the security of identity
authentication schemes only theoretically.
Identity authentication evaluation system can be
divided into evaluation systems for similar models
which have same authentication factor (such as
having the same authentication encryption
algorithm
or having the same authentication technology, etc.)
and evaluation systems for new models.
Identity authentication evaluation systems for
similar models comprehensively compare models’
security and performance in particular network
environment. Wei Liang (2005) and other
researchers divide analysis indictors into security
indictors and quality of service (QoS) indictors.
Based on the identity authentication data encryption,
data integrity, privacy protection and
non-repudiation, challenges / response models are
divided into four security index. By using the
method above, although boundaries clear to divide
and easy to operate, it cannot effectively compare
highly similarity identity authentication models.
Charlott E and other researchers propose identity
authentication evaluation system based on mobile
platform IP Multimedia Subsystem (IMS). The
system includes three basic parts: security,
user-friendly and simplicity and researchers point
out that three basic parts are interdependent. SWOT
analysis method, user ranking method and
simulation are used to compare different models
(Charlott E, 2009). Although this paper put forward
a complete and meticulous evaluation system, it does
not give a clear solution about how to use the system
to evaluation identity authentication models.
Patroklos G.A. (2004) and other three researchers
compare three security protocols (SSL, S/MINE and
IPSec) which are widely used in normal network
according to length of the key and time cost of
mutual authentication execution. The research closes
to reality application, but does not fit every model.
Evaluation systems for new models are
established to compare new models with similar
models, in order to indicate advantages of the new
one. Researcher Vipul Gupta (2002) proposes a new
SSL protocol based on ECC, and compares the
performance between the new protocol and SSL
protocol based on RSA. Yong-bin Zhou (2009) and
other researchers improve MAKAP protocol and
prove the security of the improved protocol as well
as compare the improved protocol with the old one
on calculate cost, communication cost and storage
cost. Zhi-qiang Xie (2009) and other researchers
propose new S/KEY authentication scheme and
compare the improved protocol with the old one on
mid-man-attack, decimal attack, secret key security
and so on. This kind of research focus on discussing
the model about it’s resistance to attack and unfit for
evaluation and comparison with other types of
authentication schemes. Most of this researches
limited in theoretical analysis and have no
quantitative evaluation results.
3 THE M-COMMERCE
SECURITY INDICATORS
EVALUATION SYSTEM
The evaluation system is established to meet mobile
commerce identity authentication schemes’
characteristics, which can reflect the characteristics
of mobile commerce properties and the effect of the
identity authentication.
Characteristics of mobile commerce mainly
embodies in three aspects: network environment,
terminal characteristics and service features. Small
bandwidth, high BER (bit error ratio) and the
opening of interface make mobile communication
network in numerous threats and these threats may
lead to different kinds of attacks. Mobile terminal
equipment has low storage capacity, battery power
specialty and its high secrecy request makes the
security of user information becoming more
important. At the same time, mobile commerce
customers generally require information immediately,
which means the immediacy of the service.
Evaluation system should be able to reflect the
characteristics of mobile commerce and make
mobile commerce identity authentication different
from identity authentication applications in common
net work environment.
Security is the most important part of identity
authentication effect. Researcher Mangipudi Kumar
V.K.N. (2005) proposes identity authentication
design framework based on the wireless network,
ICISO 2010 - International Conference on Informatics and Semiotics in Organisations
246
Table 1: M-commerce authentication security evaluation indicators system.
Target Level
First level
index
Second level index
Security
Analysis
Authentication
function
U
1
Mutual authentication U
11
Anonymity U
12
Non-repudiation U
13
Integrity U
14
Key agreement U
15
Multi-factor authentication U
16
Encryption
intensity
U
2
Theoretically decoding year U
21
Freshness of secret key U
22
Type of cryptograph U
23
Resistance of
attack
U
3
Type of cannot-resistance-attack U
31
Serious degree of cannot-resistance-attack U
32
and security is one of the most important factors,
which include security service and security
requirements. Security requirements indexes are
mainly about network attack, which must be satisfied
in identity authentication design. And security
services indexes, including mutual authentication,
non-repudiation etc, are optional functions in
identity authentication design.
To build a useable security evaluation system
should considerate a lot of factors. The evaluation
system can exert its effect only if the system is
established scientific and reasonable and each index
should reflect the evaluation objects. To establish a
set of perfect, rational and scientific identity
authentication model evaluation indicator system,
should follow the scientific, comprehensive,
feasibility and comparability principles. Indexes’
selection should cover all security requirements of
mobile commerce identity authentication. Besides,
indexes’ selection should avoid index correlate with
others at the same time, in order to keep eventually
evaluation result in a high quality. Otherwise,
identity authentication technology is an actual
operation process, so the evaluation index should be
convenient to get information and possible to
quantify information. This paper establishes mobile
commerce authentication security evaluation
indicators system shown in Table 1.
According to the traditional electronic commerce
security problems and mobile commerce’s
characteristics, mobile commerce security should
include mobile terminal security (Youqing G, 2005),
wireless network security, mutual authentication and
non-repudiation (Jiayuan L, 2006) four requirements.
According to the above requirements, evaluation
system is determined in Table 1. Authentication
function is an index set to assess schemes’ security
function, so indexes cannot be quantitative evaluated.
So a modeling method is introduced to quantization
the evaluation result.
4 MODELING BASED ON FUZZY
COMPREHENSIVE
EVALUATION
Fuzzy comprehensive evaluation method is an
important branch of the theory of fuzzy mathematics.
It has been widely used in economics, management,
environment, education and other fields. With the
social, economic development and complexity of
issues people consider, uncertainty and ambiguity of
the human mind has been strengthening, which
makes it very difficult for us to make an objective
evaluation. Fuzzy comprehensive evaluation method
as a solution of the problem is receiving more and
more attentions (Zeshui X, 2001).
Authentication technology is a very complex
issue. Variety of factors can impact the effect of
authentication and their impact on the way and size
are both different and vague. Fuzzy comprehensive
evaluation method is more objective on reflection
the ambiguity of the role of these effects and thus is
more scientific and feasible. Specific steps are as
follows:
The first step, determine the factor set
U
.
The factor set is divided into the target level
U
,
the first level evaluation factors
Ui
and the
second-level evaluation factors
Uim
.
0,,
) U... U, U,(UU
) U... U, U,(UU
321
321
>
=
=
nmi
imiiii
i
SECURITY ANALYSIS OF AUTHENTICATION SCHEMES IN M-COMMERCE BASED ON FUZZY
COMPREHENSIVE EVALUATION
247
Table 2: M-commerce authentication security evaluation criterion.
Target level
First level
index
Second level index
Possible
result
Optimal
result
Security
analysis
Authentication
function
U
1
Mutual authentication U
11
Yes/No
Yes
Anonymity U
12
Yes/No
Yes
Non-repudiation U
13
Yes/No
Yes
Integrity U
14
Yes/No
Yes
Key agreement U
15
Yes/No
Yes
Multi-factor authentication U
16
Yes/No
Yes
Encryption
intensity
U
2
Theoretically decoding year U
21
0-
Freshness of secret key U
22
Fresh/not
fresh
Fresh
Type of cryptograph U
23
0-
Resistance of
attack
U
3
Type of cannot-resistance-attack U
31
0-
0
Serious degree of cannot-resistance-attack U
32
Serious/not
serious
Not serious
Determined three-level fuzzy comprehensive
evaluation criteria are shown in Table 2.
The second step, determine the reviews set
V
. By
reference to fuzzy comprehensive evaluation method
application examples in other areas, this paper
reviews set
}V ...V ,V ,{VV 321 n=
as
V
= (Excellent,
Good, General, Poor, Terrible). The optimal result or
the compared optimal result will be the criterion of
mark. Indexes which can not be quantitative
evaluated will be marked according to the number of
possible values and five levels as shown in Table 3.
Table 3: Evaluation set description.
Evaluation
level
Reference Note
Excellent
V
1
Fully in line with index for evaluation criteria
Good
V
2
In line with the vast majority of index,
criteria, only the individual projects failed to
meet the index
General
V
3
Basically in line with targets, criteria, does
not meet the requirements of the less
Poor
V
4
Basically does not meet the targets, criteria,
does not meet the requirements of the more
Terrible
V
5
The vast majority of judges do not meet
targets, and only be able to meet the
requirements of individual projects
The third step, determine the evaluation factors
known to the weight vector
W
. Weight vector,
which relative with factors set, is used to display the
weight of each factor's value. Weight vector can be
divided into three levels according to factor set. All
levels of the weight vector can be expressed as
=
=
1w 0,w
) w... w, w,(wW
321
ii
i
For the establishment of mobile commerce
authentication evaluation system, expert scoring
method is used to get the initial data. We invite 10
experts for this research. Each questionnaire has
been filled independently. Table 4 shows 10 experts’
scoring statistics result of factors’ weight.
The fourth step, determine fuzzy relationship
matrix R. According to the scoring result of 3rd level
factors of 10 experts, fuzzy relationship matrix
imR
will be calculated as follows.
)(
)
10
,
10
,
10
,
10
,
10
(R
5,4,3,2,1
54321
imimimimim
im
rrrrr
kkkkk
=
=
Using the same method, a particular mobile
commerce authentication model’s final fuzzy
relationship matrix can be obtained.
The fifth step, determine fuzzy transformation
method. Obtained the results of comprehensive
evaluation through calculate
RW D
.
)b ,...,b ,b,(b
) w... w, w,(w
R
W
B
321
54321
2524232221
1514131211
321
m
mmmmm
m
rrrrr
rrrrr
rrrrr
=
=
=
#####
D
D
(1)
The sixth step, calculate the final evaluation
result.
=
==
5
1x
xx cbCBT D
(2)
ICISO 2010 - International Conference on Informatics and Semiotics in Organisations
248
Table 4: Evaluation index weight result.
Target
level
First level
index
Score/
Weight
Second level index Score/Weight
Security
analysis
Authentication
function
U
1
19
(0.32)
Mutual authentication U
11
53 (0.26)
Anonymity U
12
32 (0.15)
Non-repudiation U
13
35 (0.17)
Integrity U
14
45 (0.21)
Key agreement U
15
32 (0.15)
Multi-factor authentication
U
16
13 (0.06)
Encryption
intensity
U
2
19
(0.32)
Theoretically decoding year
U
21
23 (0.38)
Freshness of secret key U
22
23 (0.38)
Type of cryptograph U
23
14 (0.24)
Resistance of
attack
U
3
21
(0.36)
Type of
cannot-resistance-attack U
31
17 (0.57)
Serious degree of
cannot-resistance-attack U
32
13 (0.43)
5 AN ILLUSTRATIVE EXAMPLE
In this paper, we will use mobile commerce
authentication scheme based on OTP as an example
to begin illustrative example.
In this paper, we will use mobile commerce
authentication scheme based on OTP as an example
to begin illustrative example.
The main idea of OTP is to add indefinite factors
during the course of logging in and compress it to a
length definite digest, and then make the digest viz.
password used each time different to improve
security. The OTP based schemes have two common
advantages. One is the password is different each
time when logging in and therefore the adversary
capture cannot guess the password. Second is that all
of the schemes are using of hash function to produce
OTP. If the input information changes a little, the
output information would change a lot. Therefore,
this character can resist the probability that the
password would be juggled.
Two-dimension bar code is a kind of high density
bar code. It can record much information limited in a
small districts and itself is an integrated data file.
There are two common used two-dimension bar
codes. One is the stack bar code, such as PDF417...
Code49. The other is matrix bar code, such as Maxi
Code... Code1, etc.
For the problems of existed schemes and the
features of two-dimension bar code, a new OTP
scheme based on two dimension bar code is
proposed (Mu Yang, 2008). This new scheme
absorbed the advantages of the existed schemes; it
adopts the high security function-hash function. The
password is changed every time. The secret
password does not transmit directly in the network
and it also does not preserve directly in the server's
database.
In the new scheme, we take IMEI (International
Mobile Equipment Identity Number, IMEI) , and the
specific identity of mobile equipment, as an
important factor of authentication. This new scheme
takes the indefinite factor-Counter and the mobile
identity-lMEI or the service identity-Serl to generate
OTP. And then take advantages of two-dimension
bar code to cipher the authentication information
twice. Details about the new mobile commerce
authentication scheme will be got from the paper
written by Mu Yang in the reference.
The evaluation system factor set, evaluation set
and weight of each factor have been confirmed
above, so here start illustrative example from Step 4.
10 experts’ scoring statistics results as shown in
Table 5.
According to the formula (1), security evaluation
process and the results are as follows.
According to the formula (2), calculate the final
evaluation result as follows.
()
8.382208.3
1
2
3
4
5
048.00384.035072.016928.03936.0
CBT
=
=
=
D
D
The evaluation result is close to four,
so the security evaluation result of mobile
SECURITY ANALYSIS OF AUTHENTICATION SCHEMES IN M-COMMERCE BASED ON FUZZY
COMPREHENSIVE EVALUATION
249
Table 5: OTP authentication scheme security score.
Target
level
First level
index
Second level index
Membership grade
V
1
V
2
V
3
V
4
V
5
Security
analysis
Authentication
function
U
1
Mutual authentication U
11
1.00
Anonymity U
12
1.00
Non-repudiation U
13
1.00
Integrity U
14
1.00
Key agreement U
15
1.00
Multi-factor authentication U
16
1.00
Encryption
intensity
U
2
Theoretically decoding year U
21
0.80 0.20
Freshness of secret key U
22
1.00
Type of cryptograph U
23
0.50 0.50
Resistance of
attack
U
3
Type of cannot-resistance-attack U
31
0.20 0.80
Serious degree of cannot-resistance-attack U
32
0.20 0.80
()
()
()
()
()
()
048.00384.035072.016928.03936.0
0080.020.00
012.0196.0304.038.0
15.000085.0
36.032.032.0
RWB
0080.020.00
RWB
012.0196.0304.038.0
RWB
15.000085.0
0
00000.1
000000.1
000000.1
000000.1
00.10000
000000.1
06.015.021.017.015.026.0
R
B
11
333
222
111
=
=
=
=
=
=
=
=
=
=
D
D
D
D
D
D
commerce authentication scheme based on OTP is
close to "good" category.
6 CONCLUSIONS
According to the characteristics of mobile commerce,
this paper establishes mobile commerce
authentication security evaluation index system. This
system involves 11 specific indexes which basically
have no interrelated problems. On the basis of
evaluation index system, this paper use the method
of fuzzy comprehensive evaluation to build
evaluation model, and further to assess a mobile
commerce authentication scheme based on OTP as
an illustrative example.
Using the fuzzy comprehensive evaluation
method to study the authentication problem in
mobile commerce could theoretically consider all the
factors and make research more reasonable. Factors’
weights and membership adopts expert scoring.
Although each expert scoring also has certain
subjective, but from all the experts overall it can
reflect the objective and scientific.
As mobile commerce authentication security
involving many factors, some indexes can not be
quantitative evaluated yet and the selection of index
will be further discussed. Each index’s weight need
ICISO 2010 - International Conference on Informatics and Semiotics in Organisations
250
to be further optimized and researched, in order to
meet objective requirement of evaluation.
REFERENCES
Minglei S, Chengxiang T, Haihang W. Mobile
authentication scheme using SMS.
2009 IITA
International Conference on Services Science,
Management and Engineering
, 2009: 161–164.
Kwok-Yan L, Siu-Leung C, Ming G, Jia-Guang S.
Lightweight security for mobile commerce
transactions.
Computer Communications, 2003, 26:
2052-2060.
Neuman B, Ts’T. Kerberos: Authentication service for
computer networks.
IEEE Communication Magazine,
1994, 32(9): 33-38.
Aloul F, Zahidi S, El-Hajj W. Two factor authentication
using mobile phones,
IEEE/ACS International
Conference on Computer Systems and Applications
,
2009, 5: 641-644.
Oh HG, Kang SY, Seo JT, Lee IY, Moon J. Study on a
safe and efficient mOTP (mobile-OTP) authentication
mechanism for the mobile environment.
Journal of
Internet Technology, 2009, 10(5): 521-531.
Soleymani B, Maheswaran M. Social Authentication
Protocol for Mobile Phones.
CSE’ 09 International
Conference on Computational Science and
Engineering
, 2009, 8: 436–441.
Xuefei C, Xingwen Z, Weidong K, Liangbing H.
Identity-based anonymous remote authentication for
value-added services in mobile networks.
Vehicular
Technology, IEEE Transactions, 2009, 58(7):
3508–3517.
Hai-yan Q. Compairing the inductive method and strand
spaces for security protocol.
Verification Journal of
Computer Research and Development, 2008,
45(Suppl.): 137-142 (in Chinese).
Javier Thayer Fdbrega F, Jonathan Hermg C, Joshua
Guttman D. Strand spacers: Proving security protocols
correct [J].
Journal of Computer Security, 1999, 7(2-3):
191-230.
Burrows M, Abadi M, Needham R. A Logic of
Authentication [J].
ACM Trans on Computer Systems,
1990, 8(1): 18-36.
Lawrence Paulson C. The inductive approach to verifying
cryptographic protocols.
Journal of Computer Security,
1998, 6(2): 85-128.
Wei L, Wenye W. On performance analysis of
challenge/response based authentication in wireless
networks.
Computer Networks, 2005, 48: 267-288.
Charlott E, Markus F, Ivar J. A criteria-based evaluation
framework for authentication schemes in IMS.
ARES
'09. International Conference on Availability,
Reliability and Security
, 2009, 3: 865-869.
Patroklos Argyroudis G, Raja V, Hitesh T, Donal
O’Mahony. Performance analysis of cryptographic
protocols on handheld devices.
Third IEEE
International Symposium on Network Computing and
Applications, 2004: 169 – 174.
Vipul G, Sumit G, Sheueling C, Douglas S. Performance
analysis of elliptic curve cryptography for SSL.
ACM
Workshop on Wireless Security, 2002: 87-94.
Yong-bin Z, Zhen-feng Z, Guo-deng D. Analysis and
Improvement of a Security-Provable Mutually
Authenticated Key Agreement Protocol.
Journal of
Software, 2006,4(17): 868-875 (in Chinese).
Zhi-qiang X, Jun G, Jing Y. Analysis and design of new
S/KEY authorization solution.
Computer engineering,
2009 35(5): 175-176, 193 (in Chinese).
Kumar V.K.N. Mangipudi. New authentication and key
agreement protocols for wireless applications
[Graduate],
North Dakota State University, 2005.
You-qing G, Xiao-jun W, Xiao-yan D. Electronic
commerce security technology.
Beijing: Beijing
university of posts an telecommunications press, 2005
(in Chinese).
Jia-yuan L. Enaluation model based on status
authentication system.
Computer knowledge and
technology
, 2006, 7: 51-52, 57 (in Chinese).
Ze-shui X, Zhen-jun Y. Tow algorithms for fuzzy
synthetic judgment.
Journal of PLA university of
science and technology (Natural science edition)
, 2001,
2(4):5-8 (in Chinese).
Yang M, Runtong Z, Qin W, New Authentication Scheme
for M-Commerce Based on Two Dimension Bar Code.
IEEE International Conference on Service Operations
and Logistics, and Informatics, 2008: 1029-1034.
SECURITY ANALYSIS OF AUTHENTICATION SCHEMES IN M-COMMERCE BASED ON FUZZY
COMPREHENSIVE EVALUATION
251