ON THE IMPROVEMENT OF REMOTE AUTHENTICATION
SCHEME WITH SMART CARDS
Lih-Yang Wang and Chao-Chih Chen
Department of Electronic Engineering, Southern Taiwan University of Technology
Yung-Kang City, Tainan, Taiwan Roc
Keywords: Remote authentication, Privileged Insider attack, smart card.
Abstract: In 2005, Sun et al’s proposed a user-friendly remote au
thentication scheme. In order to improve the
efficiency of the authentication process, their method is based on one-way hash function. Unlike previous
methods, Sun’s method allows the user to choose and change the password locally without connecting to the
server. It can resist replay attack, impersonation attack, guessing password attack, denial of service attack.
However, in this paper we will point out that their scheme is vulnerable to privileged insider attack, and an
enhanced scheme is proposed to eliminate the weakness.
1 INTRODUCTION
In 2000, Sun proposed a remote user authentication
scheme using smart card base on one-way hash
function (sun, 2000). His method does not need to
store user's password and can resist a replay attack
of an attacker. However, the user has no power to
choose his password. This inconvenience was
removed by Wu and Chieu’s improved method few
years later (Wu and Chieu, 2003), with an additional
merit that the selection of password can be
accomplished locally without the help of the server.
Nevertheless, it is found that their method have
some weakness (Yang and Wang, 2004; Wu and
Chieu, 2004; Wang et al, 2005; Hsu, 2005; Hwang
and Liao, 2005; Hwang et al, 2005; Lee and Chiu,
2005). If the attacker can intercept and forge
messages between the user and the remote server, he
can login to the server without having the correct
password. To cope with this problem, a lot of
enhanced schemes are proposed later, including Wu
himself (Wu and Chieu, 2004; Wang et al, 2005;
Hsu, 2005; Lee and Chiu, 2005). Unfortunately, it is
found that Wu's own improved version is still risky
under stolen smart card attack, forgery attack,
privileged insider's attack and off-line password
guessing attack (Yoon and Yoo, 2007 and Ku et al,
2005).
While most of the above methods, some
ope
ration of are used the logarithm functions and
modulus exponentiation, Sun et al’s improvement
method (Sun et al, 2005) adopts only simple but
effective hash functions. Their method can resist
replay attack, impersonation attack, guessing
password attack, denial of service attack, and only
need to use the smart card to finish the demand of
changing the password. But, it is still vulnerable to
some attack.
In this paper, we will analyze Sun et al’s method,
an
d show that it is not secure under Privileged
Insider's Attack (PIA). We then propose a new
enhanced scheme to overcome this problem.
This paper is organized as follows: In Section 2,
we
review Sun et al’s user-friendly remote
authentication scheme. In Section 3, we prove that it
is vulnerable under PIA. An improved scheme to
eliminate the weakness is proposed in Section 4. The
security analysis of the proposed method is given in
Section 5. Finally, a brief conclusion is given in
Section 6.
2 REVIEW OF SUN ET AL’S
SCHEME
In this section, we briefly review Sun et al's scheme.
Table 1 lists the definitions for all required symbols,
there are needed for the successive description:
93
Wang L. and Chen C. (2007).
ON THE IMPROVEMENT OF REMOTE AUTHENTICATION SCHEME WITH SMART CARDS.
In Proceedings of the Second International Conference on Security and Cryptography, pages 93-97
DOI: 10.5220/0002124900930097
Copyright
c
 SciTePress

Table 1: Meaning of representatives of symbols.
symbols meaning
U
i
the remote user
ID
i
U
i
’s identifier
PW
i
the password chosen by U
i
PW
i
*
the password inputted by U
i
PW
i
’
the new password chosen by U
i
S
the server
x
secret key of the server
R
a randomly selected value by the user
R
*
a randomly inputted value by the user
R’
a randomly selected new value by the user
T
the current timestamp of the U
i
input message.
T’
the current timestamp of the S receive message.
T
#
the current timestamp of the attacker vary message.
h(.)
a cryptographic hash function
⊕
a bitwise XOR operation
║
a string concatenation operation
2.1 Registration Phase
In this phase, the legal user applies for his
authorization of login.
Step 1: Transfer personal identity and password
{ID
i
, PW
i
} from U
i
to S via a secure channel.
Step 2: S calculates A
i
and B
i
, where
A
i
=h(ID
i
, x) (1)
B
B
i
= A
i
⊕ PW
i
(2)
Step 3: Store the smart card with U
i
's information {
ID
i
, BB
i
, h(⋅)}. Deliver the smart card to U
i
via a secure channel.
2.2 Login Phase
In this phase, U
i
will login to the server.
Step 1: U
i
puts the smart card into the card reader,
and inputs ID
i
*
and PW
i
*
Step 2: The smart card computes A
i
*
and C
1
, where
A
i
*
= B
i
⊕ PW
i
*
(3)
C
*
1
= h( T , A
i
*
) (4)
Step 3: The login message: m={ ID
i
*
, C
1
*
, T} are
forward to S through a public channel.
2.3 Authentication Phase
In this phase, S verifies U
i
's login demand.
Step 1: S checks the U
i
's ID. If it is incorrect or
already login, refuse the login request.
Step 2: S confirms the conveyance of login and
receives time (
T and T’), if unreasonable,
refuse login request.
Step 3: S computes A
i
, C
1
and check C
1
*
. If
inconsistent, refuse the login request.
A
i
= h ( ID
i
, x )
C
1
= h ( T ⊕ A
I
) (5)
If C
1
= C
1
*
, an inconsistency occurred, refuse the
login request.
2.4 Password Change Phase
In this phase, U
i
changes the password.
Step 1: U
i
puts the smart card into the card reader,
and input ID
i
*
, PW
i
*
and PW
i
’.
Step 2: It computes a new value B
B
i
’ for the smart
card to replace B
i
B . The symbol PW
i
’
denotes
the new password.
B
B
i
’ = B
i
B ⊕ PW
i
⊕ PW
i
’
(6)
3 ANALYSE THE SUN ET AL'S
METHOD
In this Section, we will analyze Sun et al's method
from two aspects, that is, security and authentication
efficiency.
3.1 Privileged Insider's Attack
Though the user conveys identity and password to
the server and applies for the smart card through the
secure channel. However, if the user's password is
not protected, one single privileged insider can get
the user's password, imitate user's login to system
and gain improper benefit (Ku et al, 2005).
3.2 Efficiency Analysis
According to Sun et al's scheme, whenever a user
inputs a wrong password, he has to wait for the
server to decline his request. That is, it needs two
transmissions between the user end and the server
end. This mechanism seems troublesome and
inefficient. If the smart card itself has the ability to
verify the password, it can considerably simplify the
whole authentication process. In addition, if the
password is not confirmed, then the smart card will
face the risk revised illegally.
SECRYPT 2007 - International Conference on Security and Cryptography
94

4 THE PROPOSED METHOD
In this section, we propose an improved method can
prevent the privileged insider's attack, and add the
mechanism of password authentication in the smart
card. The scheme includes four phases: registration
phase, login phase, authentication phase and
password change phase.
4.1 Registration Phase
In this phase, the legal U
i
applies for the
authorization of login. For presentation neat, we
follow the same convention as Sun et al's.
Step 1: U
i
randomly chooses a value R, and
computes the register password G.
G = h( PW
i
║ R ) (7)
Step 2: U
i
submits {ID
i
,G} to S via a secure channel.
Step 3: S calculates , and V, where
i
A
i
B
A
i
= h( ID
i
⊕ x )
B
B
i
= A
i
⊕ G (8)
V = h( ID
i
⊕ G ) (9)
Step 4: S send U
i
's information { ID
i
, BB
i
, h(.), V} to
U
i
, it is then stored in the smart card.
4.2 Login Phase
In this phase, U
i
will login to S.
Step 1: U
i
puts the smart card into the card reader,
and input ID
i
*
, PW
i
*
and the number R
*
.
Step 2: Smart card computes G
*
and V
*
, and then
compare it with V. If equivalent, go to the
next step; otherwise, stop.
G
*
= h(PW
i
*
║ R
*
) (10)
V
*
= h(ID
i
*
⊕ G
*
) (11)
Step 3: The smart card computes A
i
*
and C
1
*
.
A
i
*
= B
i
⊕ G
*
(12)
C
1
*
=h(T ⊕ A
i
*
) (13)
Step 4: The login message m= { ID
i
, C
1
*
, T} is
forward to S through a public channel.
4.3 Authentication Phase
In this phase, S verifies U
i
's login demand.
Step 1: S confirms the U
i
's ID
i
, if the identity is
incorrect or already logged in, refuse login
request.
Step 2: S compares the time stamp T in the login
message with the time
T’ that S
acknowledges the login request. If
unreasonable, refuses the login request.
Step 3: S computes A
i
and C
1
, where
A
i
= h( ID
i
⊕ x )
C
1
= h( T ⊕ A
i
)
If C
1
= C
1
*
, an inconsistency occurred, refuse the
login request.
4.4 Password Change Phase
In this phase, U
i
changes his PW
i
.
Step 1: U
i
puts the smart card into the card reader,
and input personal account number ID
i
*
,
current password PW
i
*
, the number R
*
, a new
password PW
i
’
, and another randomly chosen
number R’.
Step 2: Smart card computes G
*
and V
*
, and then
compares it with V. If equivalent, goes to the
next step; otherwise, stop.
G
*
= h( PW
i
*
║ R
*
)
V
*
= h( ID
i
*
⊕ G
*
)
Step 3: Smart card computes a new value G
’
, V
’
and
B
B
i
’, where
G
’
= h( PW
i
’
║ R
’
) (14)
V
’
= h( ID
i
*
⊕ G
’
) (15)
B
B
i
’ = B
i
B ⊕ G ⊕ G
’
(16)
Step 4: The smart card uses V
’
and BB
i
’ to replace old
V and B
i
B , and store them.
5 SECURITY ANALYSIS OF OUR
METHOD
In this section, we analyze the proposed method.
Prove that our method can resist more kinds of
attacks than that of Sun et al's method.
5.1 Privileged Insider's Attack
Recently, a lot of financial crimes against banks or
enterprises are committed by insiders. This has
become a serious problem. In the proposed method,
as can be seen in Eq. (6), the message G is used to
register to the server. It is the result of a hash
function, where the input parameters are the
personal password PW
i
and random number R.
Because of the strength of one-way hash function,
the proposed algorithm can resist a PIA, as long as
the user does not reveal his registration password.
5.2 Replay Attack
If the attacker sends the interceptive information {
ID
i
*
, C
1
*
, T} to the server, his login request will be
ON THE IMPROVEMENT OF REMOTE AUTHENTICATION SCHEME WITH SMART CARDS
95

refused, because of the timestamp mechanism. Refer
to Eq. (13), since C
1
*
consists of A
i
*
and T, the
attacker must face the complexity of one-way hash
function. That is to say, assume that the attacker
replaces the conveyance time T with a false one, said
T
#
, in the interceptive information, and pass it to the
server. His conspiracy will be discovered, because
he can not produce the value C
1
#
corresponding to
T
#
. Therefore, the proposed method also can resist
the replay attack.
5.3 Impersonation Attack
The attacker has two major approaches to conduct
the Impersonation attack. Firstly, he can steal the
legal user’s smart card, and input the password of
guessing. According to Eq. (12), because we
increase the mechanism of password authentication,
whenever the attacker inputs wrong ID
i
*
, PW
i
*
and
R
*
values, he can't obtain the service of the smart
card. Secondly, the attacker makes use of old
messages to perform an intercept attack. Again,
according to Eq. (1) and Eq. (13), he will have to
resolve the difficult problem of one-way hash
function. Moreover, he has to steal the secret
information(x and G
*
) from the user and the server.
So, their attack will not succeed.
5.4 Guessing Password Attack
Similar to the above analysis, if the attacker wants to
utilize some interceptive information { ID
i
*
, C
1
*
, T}
to attack, then he must face the challenge of
breaking a double-hash function. In addition, based
on Eq. (10) and Eq. (12), if the attacker wants to
obtain the secret of G
*
, then he must face the
problem of solving a one-way hash function. Even if
the attacker gets the smart card, it is difficult to
guess the correct values of PW
i
and R by using the
dictionary attack, because value R is chosen at
random. So, the proposed method can resist the
guessing password attack.
5.5 Denial of Service Attack
We divide the denial of service attack into three
phases, and discuss them one by one. Consider the
login phase. Refer to Eq. (11), users can confirm the
legitimacy of the input information (ID
i
*
, PW
i
*
and
R
*
) and transfer the message of login by himself.
Then, consider the authentication phase. The server
can distinguish the legitimacy of login information
alone. Finally, consider the process of password
changes phase. Refer to Eq. (11), users also utilize
the smart card to verify the correctness of the input
information (ID
i
*
, PW
i
*
and the number R
*
), and
offer change service of register password. As can be
seen, all these phases require only local operations,
no message exchanges between the smart card and
the server are required, and thus it can resist the
denial of service attack.
6 CONCLUSION
In this paper, a remote user authentication scheme
using smart card base on one-way hash function was
proposed. This method provides an enhancement
protocol to Sun et al’s scheme, which has been
shown insecure. In their method, a privileged insider
can impersonate the lawful user to login remote
server. We not only resolve this security problem,
but also increase the verification efficiency of the
password.
ACKNOWLEDGEMENTS
This research was partially supported by the
National Science Council, Taiwan, R.O.C., under
grant NSC95-2220-E-218-002.
REFERENCES
Eun-Jun Yoon, Kee-Young Yoo, 2007. Comments on
Modified User Friendly Remote Authentication
Scheme with Smart Cards. IEICE TRANS. COMMUN.
Da-Zhi Su , Ji-Dong Zhong, Yu Sun, 2005. Weakness and
improvement on Wang–Li–Tie’s user-friendly remote
authentication scheme. Applied Mathematics and
Computation 170.
Narn-Yih Lee, Yu-Chung Chiu, 2005. Improved remote
authentication scheme with smart card. Computer
Standards & Interfaces.
Wei-Chi KU, Hsiu-Mei CHUANG, Maw-Jinn TSAUR,
2005. Vulnerabilities of Wu-Chieu’s Improved
Password Authentication Scheme Using Smart Cards.
IEICE TRANS. FUNDAMENTALS.
Min-Shiang Hwang, Jung-Wen Lo, Chi-Yu Liu, Shu-Chen
Lin, 2005. Cryptanalysis of a User Friendly Remote
Authentication Scheme with Smart Card, Journal of
Applied Sciences.
Kuo-Feng Hwang, I-En Liao, 2005. Two attacks on a user
friendly remote authentication scheme with smart
cards, ACM SIGOPS Operating Systems Review.
Chien-Lung Hsu, 2005. A user friendly remote
authentication scheme with smart cards against
SECRYPT 2007 - International Conference on Security and Cryptography
96

impersonation attacks, Applied Mathematics and
Computation.
Yingjie Wang, J.H. Li, L. Tie, 2005. Security analysis and
improvement of a user-friendly remote authentication
protocol, Applied Mathematics and Computation.
Shyi-Tsong Wu, Bin-Chang Chieu, 2004. A note on a user
friendly remote authentication scheme with smart
cards. IEICE Trans. Fundamentals.
Chou-Chen Yang, Ren-Chiun Wang, 2004. Cryptanalysis
of a user friendly remote authentication scheme with
smart, Computers & Security.
Shyi-Tsong Wu, Bin-Chang Chieu, 2003. A User Friendly
Remote Authentication Scheme with Smart Cards.
Computer and Security.
Hung-Min Sun. 2000. An efficient remote use
authentication scheme using smart cards. IEEE
Transactions on Consumer Electronics.
ON THE IMPROVEMENT OF REMOTE AUTHENTICATION SCHEME WITH SMART CARDS
97