AuthenLink: A User-Centred Authentication System
for a Secure Mobile Commerce
Christina Braz
1
, Esma Aïmeur
2
1
Department of Mathematics and Industrial Engineering, École Polytechnique de Montréal,
C.P. 6079, succ. Centre-Ville, Montreal (Quebec) Canada H3C 3A7
2
Department of Computer Science and Operations Research, University of Montreal,
C.P. 6128, succ. Centre-Ville, Montreal (Quebec) Canada H3C 3J7
Abstract. We envision an environment where humans communicate directly
with computers without additional authentication inputs like passwords,
passphrases, PINs (Personal Identification Numbers), biometrics, or other
existent authentication systems; and where humans network (intercommuni-
cate) continually with wireless (mobile) devices. In this paper, we propose a
new mobile authentication system, not yet implemented, called “AuthenLink”,
coupled with a new approach to distinguishing characteristics to authenticate
people (authentication factor): something you CONVEY. The utmost purpose
of this paper is to provide an ease, user-centred and acceptable security
authentication system against fraud, counterfeit, and theft for the mobile
commerce (m-Commerce) domain, more specifically for mobile devices. Our
authentication system achieves its goal through a microprocessor chip
(ChipTag) computer implanted under human skin. This ChipTag is able to
authenticate user’s access to systems, connect them wirelessly, through the
Radio Frequency Identification (RFID) technology, and enable mobile devices
perform mobile transactions, access files, or shop online.
1 Introduction
Mobile devices like mobile phones, Personal Digital Assistants (PDAs), hand-held
devices, Portable Communications Services (PCS), and 3G devices
1
are now capable
of retrieving email, managing our calendar, browsing the Web, using instant messag-
ing, viewing media, playing games. Moreover, they afford us occasionally supply
d
ocument processing, printing, and scanning abilities. In fact, the possibilities pro-
vided by mobile devices have been seen as a new paradigm: Mobile Computing.
Mobile Computing implies the “availability” concept, which refers to the omnipres-
ence of anytime, anywhere [1]. Furthermore, companies are, in an increasing degree,
using a variety of mobile devices to store crucial business information, boost produc-
tivity, and improve customer relationships to achieve the competitive advantage. The
1
3G is a short term for third-generation wireless, and refers to a new wireless standard promising increased
capacity and high-speed data applications up to two megabits, especially for mobile communications.
Braz C. and Aïmeur E. (2004).
AuthenLink: A User-Centred Authentication System for a Secure Mobile Commerce.
In Proceedings of the 3rd International Workshop on Wireless Information Systems, pages 1-11
DOI: 10.5220/0002660000010011
Copyright
c
SciTePress
statistics on the mobile commerce (m-Commerce) industry are loud and clear. Ac-
cording to Forrester Research, the m-Commerce revenues for the global market is
growing each year and will reach a US$ 22 billion in 2005 against US$7.5 billion
achieved in 2003.
The authentication process is one of the basic frameworks of computing security.
Thus, to enable a mobile device to distinguish between legitimate and non-legitimate
users, most authentication systems provide passwords to authorize mobile users.
The primary issue is the lack of usability and acceptable security authentication
system against fraud, counterfeit, and theft for mobile electronic transactions. The
ability to securely trade business or shopping online and wirelessly is dependent on
securely authenticating participants and digitally signing transactions.b
This paper presents a new mobile user authentication system called AuthenLink,
designed for mobile devices, which integrates a microprocessor chip [2], a ChipTag,
implanted under human skin and a mobile device antenna-embedded. AuthenLink
gives the user automatic access to different resources in an acceptable secure authen-
tication process, especially against fraud, counterfeit, and theft. In this way, a legiti-
mate user will be able to conveniently prove her/his identity through the Radio Fre-
quency Identification2 (RFID) technology, and gain access to the wireless network
without threatening the safety of the organization. Furthermore, this article introduces
a new approach to distinguishing characteristics to authenticate people, which we
consider a fourth authentication factor: something you CONVEY.
This paper is structured as follows: We begin with a quick overview of the Perva-
sive and Mobile Computing issues in Section 1. Then, we define an authentication
system, the usefulness role of a strong authentication and the authentication factors in
Section 2. We describe the state-of-the-art of the authentication systems industry in
section 3. Afterwards, we illustrate the technology overview of our system in section
4, while in section 5 we present the Architecture Usage Scenarios for the Au-
thenLink. Then, we describe the security aspects of our system in section 6. We vali-
date our assumptions in Section 7 with an empirical evaluation of our system in com-
parison of other authentication systems. In Section 8, we state our reasoning for pro-
posing a new mobile authentication system and a new authentication factor. Finally,
the last Section presents the conclusions, and outlines opportunities for future work.
2 Authentication
Authentication is the process of establishing whether someone is who he or she de-
clares himself or herself to be. In private and public computer networks (encompass-
ing the Internet), authentication is popularly done through the use of logon pass-
words. The logon is the process used to get access to an operating system or applica-
tion, generally in a remote computer. Usually a logon requires that the user have a
user ID (username) and a password.
2
Radio Frequency Identification (RFID) is a method of identifying unique items using radio waves.
Typically, a reader communicates with a tag, which holds digital information in a microchip.
2
Authentication is one of the critical elements of a set of services that constitute a
security sub-system in a communications infrastructure and encompasses the follow-
ing security services: Authentication, Confidentiality, Integrity, Non-repudiation,
Access Control, and finally Availability.
2.1 Strong Authentication Issues
Strong authentication refers to systems that require rigorous user identity verification,
which is realized through multiple factors for authentication and employs advanced
technology. The goal of strong authentication is to reinforce the security by replacing
the classic authentication method of password for a software-only authentication
solution with dynamic password generators, or software-hardware authentication
solutions like smart cards, tokens, biometrics, etc. The greatest challenge of strong
authentication is to make fraud more difficult for an attacker while respecting the
constraints associated with an applications technical, economical, and organizational
environment.
Until very recently, the suitable method for strong authentication was a smart card.
For good reason, smart cards use Public Cryptography Infrastructure (PKI) digital
certificates - the standard for digital authentication and signatures [3], which steadily
protect the user’s private key with hardware. However, certain characteristics of
smart cards do not support the needs of today’s business environment. These include
the following: Lack of omnipresence (end-users are severely limited by the need to
have access to card readers), Difficult to deploy (costly to administrate, and support),
Expensive not cost-effective for large, distributed user communities.
Finally, an organization’s authentication service should be suitable to the risks, and
should consider the impact on users, as well as the cost of integration with its existing
technology architecture, and total cost of ownership.
2.2 Authentication Factors
An authentication factor is Authentication Information (AI), information used to set
up the validity of a claimed identity, utilized to check an identity demanded by or for
a user. Consider the following scenario: before a Reliable Security System (RSS)
gives Bob (a legitimate user) access to a computer system, network, or secure re-
source, the RSS must determine who he is, if he belongs to this system, if he has the
right to access the system, and if he is the person he says he is. Actually, the RSS has
demanded three distinct elements – identification, authentication, and authorization –
that all together comprise the so-called access control. However, how does the RSS
confirms that Bob is who he says he is? For example, entering his password does not
prove it is him. Hence, the RSS needs the AI to authorize access for Bob. The AI may
be gathered from one of the following authentication factors, as shown in Table 1.
We can notice that associating two or more factors presents greater security (i.e. A
PIN and a smart card). In this way, an authentication system using a single authenti-
cation factor may be vulnerable, but it depends on the employed technology. Our
system will demonstrate that a single authentication factor is also possible.
3
Table 1. Authentication Factors.
Classification (NCSC-TG-017
3
)
Factor Examples
Type 1: Authentication by
Knowledge
Something only
the user KNOWS
. Password or passphrase.
. Personal Identification Number (PIN)
. Information about the user or family
members.
Type 2: Authentication by
Ownership
Something only
the user
POSSESSES
. Physical key
. Magnetic-stripe card
. A token that generates a One-Time
Password (OTP)
Type 3: Authentication by
Characteristic
Something only
the user IS (or
does)
A Biometric trait:
. Fingerprint
. Iris pattern
. Hand geometry
. Voice
Or combination of the above
3 Related Work
As of this writing, there is no related work developed that performs exactly as our
system does, especially client’s side. However it is important to present an overview
relative to existing authentication methodologies on the market.
3.1. Passwords and PINs
For user authentication to an information system, the use of a password is by far the
most common knowledge-based, Type 1 (See Table 1), authentication method. A
long password, especially one with inserted spaces, is called a passphrase.
3.2. Authentication Tokens
Authentication Tokens (ATs) supply a means of authenticating and identifying an
end-user. End-users protect their identity by using a physical object that is unique to
them, for example, using a driver’s license to prove a person’s identity. To verify the
identity of the token’s owner, the host system performs its authentication protocol
using data encoded on the token.
ATs come in a variety of physical forms. The size, shape, and materials from
which a token is manufactured are referred to conjointly as the token’s form factor.
There are three main types of token form factors: Non-Contact Tokens (demand no
electrical or physical contact with a token reader device such as proximity cards, One-
Time Password generators, and handheld challenge-response calculators), Contact
Tokens (make physical contact with the reader device like magnetic stripe tokens
3
NCSC-TG-017 is a “Guide to Understanding Identification and Authentication in Trusted Systems”,
published by the U.S. National Computer Security Center (http://security.isu.edu/pdf/idenauth.pdf).
4
used in Automated Teller Machines (ATM) and Smart Card and Public Key Authen-
tication.
3.3. Biometrics
Biometrics is a form of authentication that uses the user’s physical or behavioural
characteristics to verify his or her claimed identity. Physical characteristics like fin-
gerprints, retinas and irises, palm prints, facial structure, and voice are some of the
several existing biometric authentication methods.
3.4. Kerberos
An interesting variant of the authentication methods shown herein is Kerberos. It was
created by MIT as a solution to network security problems [U1]. Kerberos is a net-
work authentication protocol that supplies strong authentication and shares temporary
base secrets for client/server applications by using secret-key cryptography. Authen-
ticating mobile computing users might demand a considerable amount of processing
and communications resources. Hence, research efforts have been directed towards
developing some adaptations in this protocol in order to provide a better performance
of public key-enabled Kerberos authentication in mobile computing applications [4].
Yet another sub-variant of the MIT-Kerberos authentication scheme is the security
protocol [5] in which assigns authentication keys to the mobile nodes, dynamically
thus, overcoming the problems related to static passwords in traditional schemes.
4 Technology Overview - AuthenLink
Those entire authentication methods described above have security problems: they
lack usability, security (especially against fraud, counterfeit, and theft), and evolutiv-
ity. An authentication method must be flexible, interoperable, and anticipate the
user’s needs leaving an open door for future developments.
Our system is focused strongly on the user-side not on hardware itself. It is the re-
sult of the integration of a wireless semiconductor integrated circuit (IC) that stores
an ID number in its memory (Chip_User_ID), implanted under human skin, and the
mobile device antenna-embedded. This latter device will authenticate the user by
making a connection between him or her, and the authentication server.
The main AuthenLink’s components are the ChipTag, Radio Frequency Identifica-
tion (RFID) technology, Mobile Reader, Authentication Server, and Database.
5
5 Architecture Usage Scenarios
In order to implement AuthenLink, we can make use of three distinct scenarios [U3]:
UMTS
4
Architecture Mode (Maximum Mobility), WLAN
5
Architecture Mode (Me-
dium Mobility), and Ad Hoc
6
Architecture Mode (Minimum Mobility). All of these
scenarios can be implemented by an enterprise or an organization. In fact, the choice
depends on the cost the enterprise or organization is willing to incur in terms of
equipment, administration system, and human resources. In this paper, however, we
focus on the UMTS Architecture Mode (Maximum Mobility) according to Figure 1.
Let us see how it works:
Step 1: When the Mobile Reader (MR) antenna-embedded, is activated, say, when
the ChipUser turns on the MR, it radiates a small amount of radio frequency energy
through its antenna onto the ChipTag. Note: In this case, the Mobile Reader is a cellu-
lar phone.
Step 2: Radio frequency energy passes through the skin energizing the inactive Chip-
Tag, which then emits a radio frequency signal conveying the ChipUser’s unique ID
to the MR for the purposes of user authentication. Using the energy it receives from
the signal when it enters the radio field, the ChipTag will briefly converse with the
MR for verification and data exchange. The ChipTag has no power supply, and a tiny
transmitter on the ChipTag sends out the data (unique ID).
(*) ChipTag implanted under skin of the user's upper-arm with a unique
ID embedded. Example: aliceg milkshell8suitcase.
Chip_User_Name Chip_User_ID Chip_Tag_Number Time_Stamp
aliceg milkshell8suitcase AB54545YT 03:51 P.M.
ID Database
2
1
MOBILE READE
(Antenna-embedded
cell phone or PDA).
R
CHIPUSER
Base Station
Network
ChipTag (*)
4
3
AuthenLink
INTERNET
{ aliceg milkshell8suitcase }
SSH Secure Shell Session
AUTHENTICATION
SERVER
UMTS MOBILE
NETWORK
Fig. 1. UMTS Architecture Mode (Maximum Mobility).
Step 3: Once that data is received by the MR, it automatically authenticates the
ChipUser’s ID with the Authentication Server (AS) by means of a Base Station (Cell
4
It is a third-generation (3G) broadband, packet-based transmission of text, digitized voice, video, and
multimedia at data rates up to 2 megabits per second (Mbps) that offers a consistent set of services to
mobile computer and phone users no matter where they are located in the world.
5
WLAN is a local area network (LAN) without wires.
6
ADHOC architecture is a networking framework in which devices or stations communicate directly with
each other without the use of an Access Point (AP).
6
Phone Tower), and an UMTS Mobile Network, through the Internet. An SSH (Secure
Shell) session automatically logs the ChipUser onto a remote AS [6]. The ChipUser
gives his or her public key to the AS and then, when it connects, the AS knows access
is permitted and automatically enables the connection. In fact, SSH uses a pub-
lic/private encryption system to authenticate the ChipUser to the AS without the in-
tervention of the ChipUser. We merely need to create a public/private key pair for the
ChipUser, and then store the public key on the AS. Then, our SSH session client can
use that key pair to automatically authenticate the ChipUser to the AS.
Step 4: Once the data is received by the AS, it can be sent to the database for proc-
essing and management. Linking each ID from the database to the Ch
ipUser is per-
6 Security
e describe only the security issues
7
to be implemented between the
Mobile Reader (RFID Technology) and the ChipUser.
ing data exchanges. The main
dif
7 Empirical Evaluation
henLink system but we did carry out an empiri-
cal evaluation and comparison analysis of the authentication methods related to our
formed.
In this Section, w
In the RFI industry we have seen a huge effort to protect consumer privacy by se-
curing information from “eavesdropping” or intercept
ference between RFID and, for example, a magnetic stripe technology (bank cards)
is operability Over-the-Air (OTA8). The risk of eavesdropping, or intercepting, con-
veyed data is well acknowledged as is, for example, someone using a hide and mali-
cious mobile reader. These risks are greatly reduced through the design of appropriate
Over-the-Air protocols and data encryption methods. This protocol requires the Chip-
Tag to be within range of both the mobile reader and the eavesdropper. Moreover, the
mobile reader changes frequency quickly and the eavesdropping reader has to follow
precisely the main mobile reader. In fact, this task is very difficult due to the random-
ness of the hopping sequence. Then, there is the data encryption algorithm code,
which must be cracked to use the data. A well-designed system will protect consum-
ers by implementing the proper protocol to achieve a level of security comparable and
even beyond more evolved technologies.
There is no implementation of the Aut
system. In this way, we present in this section, an empirical evaluation of the authen-
tication methods in comparison of our system (See: Table 2) with respect to different
features encountered in other authentication methods.
7
A complete description of the security mechanisms between the mobile reader and the authentication
server can be seen at http://www27.brinkster.com/chbraz/index.html
8
Over-the-air (OTA): is a standard for the transmission and reception of application-related information in
a wireless communications system. OTA messages can be encrypted to ensure user privacy and data secu-
rity.
7
Table 2. Comparative Analysis of the Authentication Methods
AUTHENTICATION FACTORS
FEATURES
Autthen
Link
User-
names &
Passwords
OTPs/
Ch
Tokens Smart
Cards &
PKI
allenge
Response
Biomet-
rics
Kerberos
Accessibility
Durability
Mobility
Reliability
Performance
Security
Flexibility
Tamper-proof
Ergonomics
Privacy
Data Integrity
Ease of deployment
Interoperability
Compatibility
Extendability
Architecture Model
N.A.
In each authenticatio ethod, we rated the “features” on a Very-good (), Fa ),
sis.
Fo desc on of t advan es and s of th u-
8 Discussion
arises from this analysis: Why do we need another user authentica-
her, a new authentication factor?
n m ir (
A
and Poor ( ) ba
r a ripti he tag disadvantage e
thenLink go to: http://step.polymtl.ca/~chbraa/advantagesDisadvantages.htm
A critical question
tion system, or furt
In today’s mobile computing environment, solutions that provide an authentication
system supported by combining several authentication factors are limited because
they may be viewed as extremely cumbersome by the mobile user: something you
KNOW, something you HAVE, and something you ARE. Passwords, passphrases,
PINs, smart cards, and authentication tokens may be stolen, counterfeited, damaged,
misused, and intercepted directly from the authentication system. Furthermore, we
cannot trust biometric authentication on an unreliable wireless network unless we
distribute base secrets (installed in the biometric reader) to authenticate the biometric
readers [7]. Indeed, that’s a cumbersome two-authentication factor. Moreover, each
mobile user leaves a trace of his or her fingerprints, voice, and appearance wherever
he or she goes.
A multi-factor authentication is another technical hurdle for a mobile user. In fact,
it hides the weaknesses of distinct techniques (passwords, tokens, biometrics, etc.) by
compounding two or more authentication factors in one mechanism [8].
As we can see, it is crucial to introduce a new mobile authentication system that
improves security especially against fraud, counterfeit, and theft, and gives people on
8
the
ation keystone process di-
rec
arily be implemented to perform mobile access control with
curity, defence, financial, homeland security, and high-level secure-
ac
embedded beneath a person’s skin may sound
is. However, when the Social Security Number
(S
plications, implement an RFID tracking system to locate
ev
tication system. “The problem is a
move fast and easy transactions. Therefore, we will introduce AuthenLink, a mo-
bile one-factor authentication system; a new approach to authenticate people by dis-
tinguishing characteristics, which we consider a fourth authentication factor: Type 4 -
Authentication by Emanation: something you CONVEY.
The user authentication main task in AuthenLink is made on the client side
(ChipUser), and our effort is to internalize the authentic
tly to an individual (chip is implanted under the skin of the user) instead of inter-
nalizing it to hardware (mobile device). Hence, we are confident that the base secret
originates from a reliable source – the ChipUser.
Target group
AuthenLink could prim
a variety of se
cess applications such as government, research centres, business, and organizations.
It could also be appropriated to m-Commerce to allow end-users to perform mobile
electronic transactions. AuthenLink has to target the most tech-savvy mobile users
such as Innovators, Technology Enthusiasts, and Early Adopters. They are the main
target group for AuthenLink. However, our system is not suitable to Skeptics con-
sumers.
Implications on the User Experience
The use of an authentication chip
pedantic, or a little techy. In fact, it
SN) was implemented in North America in 1935 as an all-purpose identifier (ID),
people worried that the government would use it for other purposes. Some suspected
that we could be tracked and linked to one another with sensitive data. Today, the
vast majority of North Americans support some form of national identification like
the Health Insurance Card, Driving Permit, or Social Security Card. If we accept a
National ID system as we have accepted SSNs, five years from now the idea of an
authentication chip may not appear as threatening as it does today. As with SSNs,
people will get used to it.
Other people may be concerned about privacy because of the possibility that gov-
ernments may, in future ap
ery citizen at any time. Let us consider the colossal infrastructure costs for a gov-
ernment institution to track all citizens, not to mention the massive database that
would have to be generated. In fact, the viability of an application like this is beyond
any government’s capability. At present, in the U.S., one must obtain a court mandate
to use private information like cell phone records and credit card purchases. Hence,
the data generated from the use of RFID would be considered private and include the
same privacy protections that are in place today [U2]. “The notion of embedding an
authentication device in one's body is an interesting one. The U.S. government has
recently passed regulations approving the implantation of such devices in humans”
(Smith, R.E., personal communication, 2003
9
).
Finally, there is no way to control what could be realized, for example, with a bio-
metric authentication or, with any other authen
9
R. E. Smith, Information Systems Security Consulting. Web Site: http://www.smat.us/crypto/index.html,
Minnesota, USA, personal communication, July, 2003.
9
sim
9 Conclusions and Future Work
ndated with more electronic information
than ever before. The open networks, which are easily accessible and inexpensive,
ssues are moving from the corporate
rea
hentication could also be
ext
tion represents our major challenge. In this work, we have
taken the first step toward meeting this challenge by examining the need for a user-
ple one: computing equipment is completely amoral and cannot tell whether it is
being used for “good” or “bad” purposes. If a system can find an identity based on a
biometric signature, then there is no purely technological way of controlling WHY a
given search is performed. Given enough collusion among system operators and pro-
prietors, there is always a way to fool the system into performing its function for
unintended purposes” [9].
The m-Commerce “channels” are being inu
surpass the more expensive and functionally limited communications channels. Since
open networks are intrinsically less secure than private networks, secure m-
Commerce depends largely upon information security itself (ChipTag implanted
under skin) rather than channel security. Hence, the m-Commerce aims a mobile
authentication method that is user-friendly, flexible and adequate security in order to
boost the m-Commerce industry. AuthenLink is a single authentication factor, which
provides an acceptable degree of security against fraud, counterfeit, and theft. An-
other major point is that AuthenLink is strongly focused on usability (easy user au-
thentication) in order to take away much of the burdensome job of memorization and
typing in usernames and passwords from users.
A fundamental sign that things are changing for new techy-savvy-based systems
like AuthenLink is that worries about security i
lm - where security traditionally has been a major issue - to the individual user
level. Indeed, the end-users are becoming increasingly corporate-wise due to the fact
that they urge to reap the benefits of the technological advancements in a more secure
way, and it is going to be tough for them to keep pace of those advancements due to
the technological complexity. They aim an authentication system that is not complex
to understand and use [10], and that is available on an ongoing basis. Definitely, this
pronounced shift reinforces our belief that this is a significant factor in the growing
trend of the acceptance of a user-centred system. The AuthenLink provides usability
and always-on authentication system, speed and performance, flexibility, and contrib-
utes to the consolidation of the m-Commerce industry.
Following the investigations described in this work, a number of projects could be
taken up. The concept of using AuthenLink for mobile aut
ended to include a range of consumer products such as PCs, cars, and even homes
and apartments. Furthermore, the authentication ChipTag could not only be used as
an implant in a human being but also be attached to the mobile device, desktop, lap-
top, or any computer system as a hardware component. Another considerable im-
provement would be the possibility to allow the ChipTag be not “read only” (infor-
mation can only be read, never changed) but read/write providing thus numerous
promising applications.
Developing a secure and ease of use authentication system that can handle diverse,
mobile device authentica
10
fri
es
stèmes informatiques mobiles – Fondements, architectures et appli-
es Internationales Polytechnique, École Polytechnique de Montréal (Quebec)
Canada (2003).
4.
7.
U
e Network Authentication Protocol (2004).
eb.mit.edu/kerberos/www/
U2 Hibbert, C., Frequently Asked Questions on SSNs and Privacy, Computer Professionals for
endly and secure mobile authentication system especially against fraud, counterfeit,
and theft.
Referenc
1. Pierre, S., Réseaux et sy
cations, Press
2. Bassiouni, M. & Mukherjee, A., A VLSI Chip for Efficient Transmission and Retrieval of
Information, ACM Journal of the ACM, Department of Computer Science University of
Central Florida Orlando, Florida 32816 - USA, (1987).
3. Burnett, S. & Paine, S., RSA Security’s Official Guide To Cryptography, RSA Press,
McGraw-Hill Companies, Berkeley, California - U.S.A. (2002).
Harbitter, A. & Menascé, A., The Performance of Public Key-Enabled Kerberos Authenti-
cation in Mobile Computing Applications, Proceedings of the 8th ACM (Journal of the
ACM), Conference on Computer and Communications Security, Philadelphia, PA - U.S.A.,
Session: Mobile Code and Distributed Systems, pp. 78 – 85 (2001).
5. Raman, B. & Ramanathan, A., Artificial Intelligence Based Authentication Scheme for
Mobile Adhoc Networks, White-Paper Dept. of. Computer Science/Dept. of Electrical
Engg. - Texas A&M University, College Station, Texas 77843 U.S.A. (2001).
6. Barrett, D. & Silverman R., SSH, The Secure Shell - The Definitive Guide, 1st Edition,
O’Reilly & Associates, Inc., 101 Morris West, Sebastopol, CA – U.S. (2001).
Smith. E. R., Authentication: From Passwords to Public Keys, Addison-Wesley, Addison-
Wesley, 1st edition U.S.A. (October 1, 2001a).
8. Smith. E. R., Authentication: From Passwords to Public Keys, Addison-Wesley, Addison-
Wesley, 1st edition U.S.A. (October 1, 2001b).
9. Smith. E. R., Authentication: From Passwords to Public Keys, Addison-Wesley, Addison-
Wesley, 1st edition U.S.A. (October 1, 2001c).
10. Jøsang, A. & Patton, M. A., User Interface Requirements for Authentication of Communi-
cation, in the Proceedings of the Australasian User Interface Conference, Adelaide, Austra-
lia (February, 2003).
RLs:
U1 The MIT Kerberos Team, Kerberos: Th
http://w
Social Responsibility - Palo Alto, CA – U.S. (2004).
http://www.cpsr.org/cpsr/privacy/ssn/ssn.faq.html
U3 Braz, C., AuthenLink's Architectures Usage Scenarios (2003).
http://step.polymtl.ca/~chbraa/
11