Biometric Authentication: System Security & Privacy

This research project explores the usability, technology, and applications of biometric authentication systems. It discusses the most used biometric technologies, their pros and cons, and highlights the vulnerabilities related to present biometric techniques. The report also covers the advantages, limitations, and challenges of biometric authentication.

Biometric Authentication: System Security & Privacy
Enter Your Name
ITC595 MIT, School of Computing & Mathematics, Charles Sturt University
[e-mail]
Abstract – Technology and the underling
innovations are expending day by day, and with this
burst in technological advancements, modification in
biometric security is also taking place. Biometric
Authentication System is based on physical and
behavioural characteristics of a person. Biometric
technologies are introduced to make use of
individual’s unique biological characteristics which
are saved in the database for future references
includes. These characteristics include fingerprints,
iris pattern, voice pattern and facial details etc. In
this research project, there will be an explanation of
the usability of system, technology and applications
related to this system. Here we will describe the
most used biometric technologies in brief and will
also discuss the pros and cons of the biometric
technology systems.
Keywords— Authentication methods, biometrics,
face recognition, Privacy
Aim
The objective of this report is to highlight the
important biometric authentication methods in use
with their pros and cons. We will also highlight the
vulnerabilities related to present biometric
techniques.
Introduction
In the 21st century, 'Privacy' is the most
resonating word. The sudden surge of the digital
revolution has made it possible that in order to
protect our data and thereby our privacy we need
some updated 21st-century methods. Biometric
system of authentication is one such way to
ensure privacy and data security.
Biometric authentication
Biometric authentication is a modern method to
authenticate the identity of a person. Biometric
authentication takes into account one's anatomical
traits (e.g. fingerprint, face, palmprint,
iris or voice) and behavioural traits (e.g. signature,
gait (way of walking), keystroke dynamics, lips
movements). Because of the uniqueness of these
traits, these are the most reliable means to
authenticate the identity of a person. (Alsaadi, 2015)
Given to its reliability, some countries have even
issued biometrics based unique IDs. The database
created in such a way is further utilized to
government's social schemes and by various
institutions to identify the legitimate person.
Biometric authentication has been in use for a time
in identifying the activities involving criminal
functionalities.
Traditional and Modern authentication system
Task Examples Properties
What do you
know
User ID
Password
PIN
Can be Shared
Easy to guess
Often forgotten
What do you
have
Cards
Badges
Keys
Can be Shared
Can be duplicated
May be Stolen/lost
What do you
know and
have
ATM + PIN
Can be shared or
tracked
What is
something
unique you
have
Fingerprint
Face
Iris/retina
Voice
Ear
These things cannot
be shared,
duplicated,
guessed, forgotten
or stolen
Classification of Biometric Authentication
biometric authentication is broadly classified into
two categories. (Eng & Wahsheh 2013) The first one is
the physiological biometric system which involves
fingerprint, geometry of hand, face, iris and retina
related recognition and the second one is the
behavioural biometric system which involves
signature, gait etc.

Figure1. General classification of Biometric
schemes
Biometric System Process -
The process used is very simple, the first step is
'enrollment' in which the user is enrolled for the
record and used as the reference for future
authentication processes.( Kataria, Adhyaru,
Sharma, Zaveri 2013) For example, a face is
recorded using a camera and all the salient and
distinct features are then extracted and stored as the
templates in the database along with the other
information like name, sex, date of birth etc.
If a user wants to authenticate his/her identity, a
sample of his/her face, for example, is captured
using a camera then using the software the salient
features of the face are extracted and compared with
the files which are stored already in the database. If
the match is fulfilled then the message is shown on
the screen that proves the legitimacy of the user.
The basic process used in a biometric system is
described in the block diagram below. (Weizhi, Furnell,
Jianying 2015)
Figure 2. Block diagram of biometric process
Physiological biometric system
Physiological traits are the static traits which are not
subject to change with age. These traits include face
recognition, hand and palm print/geometry,
fingerprint, iris recognition, DNA, retina and
recognition of unique blood veins pattern.
Fingerprint Recognition - One of the popular and
oldest methods of the biometric system is the
fingerprint recognition. There may be cases when
people have similar faces or in the case of twins, this
method stands true.(Pahuja, Nagabhushan, 2015)
The basic principle utilized in this technique is that
an individual has a distinct fingerprint pattern.
Fingerprint recognition method was first devised by
Francis Galton in 1892.
Figure 3. A Sample image of human Fingerprint
Fingerprints are comprised of three main patterns.
These pattern include arch, loop and whorl. The
fingerprints are identified using a unique position
and direction of valleys and ridges of the loop, arch
and whorl. The scanning process is done by using a
device called 'fingerprint sensor'.
Face Recognition - Today the facial feature (such as
nose, eyes, lips & chin) have a great contribution in
biometric technology. Each person has a unique
shape of their face and this shape is easily identified
by using Face Recognition System (FRS). FRS can
be availed via software packages.
Figure 4. Automatic face recognition system

There are two methods of facial recognition, the first
one is 'static recognition' in which an image is used
to identify the face and the other one is 'real-time
recognition' where a real-time image using a video
camera is captured to identify the face.
Retina Recognition - Retina is the part of an eye
where the image of an object is formed. Retin is
made of special kind of cells to transmit signals to
the brain. These cells are connected to the blood
vessels. The position of blood vessels is distinct for
every retina and hence for every person. This
characteristic is used in biometric recognition and is
one of the reliable means for identification.
Iris Recognition - Iris is responsible for controlling
the amount of light entering into our eyes by
changing the size of the pupil. A retina is also
responsible for the color of an eye. Every iris has is
made of a specific network of tissues because of its
biological composition. This trait is used in
identifying an individual using a scanner.
Figure 5. Iris recognition
Palm Geometry Recognition - In this method of
biometric recognition, various aspects of geometry
are computed and compared with already stored
templates for the purpose of identification. Some
features like the length of the fingers, the distance
between these fingers, the relative position of the
fingers etc.
Figure 6. hand/palm geometry scanning
Ear Recognition - You might wonder but your ear
has a distinct shape and this particular feature can be
utilized in order to identify the legitimacy of a
person. The shape of the human ear does not change
with the increase in age.
Figure 7 Biometric authentication system based of
shape of ear
Voice recognition - This method is not that reliable
compared to the other in this category because this
method is easy to crack. In voice recognition, a
person's voice is recorded and it is subjected to
frequency analyzer where the comparison is done
and identity is revealed. Any person can make the
system fool id he can mimic the other person.
Behavioral biometric system
Signature – Signature analysis is a foolproof way to
identify the authenticity of a person. The way the
signature has been done cannot be copied exactly. In
the signature analysis, the parameters taken into
account are
 Distinct shape of the signature
 Amount of time taken to do the signature
 The stroke order of the signature
 The pressure applied on pen during the
signature
By using the above-mentioned parameters a
foolproof signature analysis can be realized. (Ye
Wang, Rane, Draper, Ishwar, 2012) This technique
is not much of use at present compared to the other
biometric technique but this is one of the best
technique to be used in the future biometric journey.
Gait - 'Gait' means the specific manner a person
walks. A camera records the walk of the user and
this is then subjected to analyze by computing the
mathematical functions performed on the inclination
of legs, by calculating the frequency of balancing of
the body. This method can be utilized in a telemetric
manner. This method also not is much use but will
be used in the future for advanced purposes.
Keystrokes - In this method, a user is made to type
something on the keyboard and the manner of typing
is analyzed. The parameters used in this technique
are -

 User’s typing manner on a keyboard
 Time taken by user in holding a key
 The time taken by the user to switch from
one key to another
This method is reliable but the reliability can vary
depending upon the stress level or tiredness of the
subject. However, it is a good comfortable
authentication method. This is easy to set up and the
system requires software and a keyboard.
Applications of the biometric system
There are numerous applications of biometric
systems. The most used are discussed below -
1. Logical Access Control – Use of the
biometric system is performed in Logical
Access Control in an extensive manner. To
gain access to the computer system either
from the current location or from the remote
location, biometric authentication is used to
fulfill the purpose. (Deutschmann,
Nordstrom, & Nilsson, 2013). By using this
we can get rid of remembering passwords
and all.
2. Physical Access Control - To give someone
permission to enter in the office building or a
room biometric authentication is used. There
is no need for keys and lock to secure the
location.
3. Time and Attendance - To note the entering
time and exit time and for the purpose of
attendance biometric authentication is used.
This makes the process more accurate and
fast.
4. Law enforcement - Government bodies
widely use biometric authentication to
identify a criminal as law enforcement.
5. Surveillance - CCTV cameras are installed
for the recording purposes and if a suspected
criminal is found to be roaming around in the
crowd, biometric authentication will easily
identify the subjected person. (Narhar &
Joshi 2015) This saves both time and efforts.
Advantages of the biometric authentication
Biometric authentication is way better than the
traditional authentication methods. Following are the
advantages of the biometric authentication system -
1. The biometric characteristics that are used to
authenticate the identity of a person are
unique and cannot be copied or duplicated in
most of the cases.
2. In traditional methods, the IP address of the
computer is taken into account which is not
the case here.
3. The major advantage of the biometric
authentication system is that you always
carry your identity with yourself so you
cannot forget it or lose it.
4. The biometric authentication system is more
practical and easy to use compared to
traditional passwords in the era where you
know only a few contact numbers in your
head.
5. The biometric authentication system is fully
attack-proof and if you are safe your is also
safe. Moreover, if you don't access it no one
ever will be able to do so.
Limitations of the biometric authentication
Every technology has some limitations. Limitations
are what inspire us to innovate.
1. As mentioned in the advantages that
biometric authentication systems use unique
biometric characteristics and if a person lacks
a particular characteristic then his/her
identity cannot be established.
2. If a person wears a new pair of shoe and he is
not comfortable to walk properly then his
gait test will declare him illegitimate.
3. If someone's fingers or palm get injured then
the authentication will not take place and the
person will be able to prove his legitimacy.
4. Biometric authentication systems are still in
the developing phase and their extensive use
has not been realized yet.
5. In most cases, the setup cost of a biometric
authentication system is very high which is
not appropriate for everyone and for common
people.
6. The major drawback is privacy which always
makes a user hesitate to use the biometric
authentication system. (Zaeri, 2011) We
don't even know where the data we are
providing is going to be stored and who is
using is for fulfilling their greed. You must
be aware of the Facebook–Cambridge
Analytica data scandal.

7. If we talk about retina authentication, iris
authentication etc., these are not the user-
friendly authentication ways.
8. There is no law governing the issues of data
security and privacy.
Vulnerabilities of biometric authentication
Biometric authentication can suffer from two kinds
of failures. One is 'denial of service' which occurs
when the system denies the identity of a legitimate
person and other is 'intrusion' which is opposite of
the first case in which the system proves the
legitimacy of an illegitimate user. The 'denial of
services' are classified under "intrinsic limitations"
and all the 'intrusions' are classified under "adversary
attacks".
Challenges in biometric authentication
The main problem related to biometric
authentication is the ‘preservation of privacy’. This
is the loophole that makes it a ‘not so’ user friendly
technology. The other problem in implementation is
regarding the setup cost which is very high which
makes the technology unreliable and is a great
setback for the technology.
Conclusion
Above discussion suggests that biometric
authentication is way better than the traditional
methods where the passwords or PINs have to be
remembered. Biometric authentications are fully
foolproof and are the reliable means to protect our
important data even and also the remote access is
available there.( Rathgeb, Breitinger, Busch & Baier 2014)
However, the concerns related to privacy have not
been answered yet. Hence we can say biometric
authentication technology is still in developing phase
and more research is needed to realize its true
potential.
Acknowledgement
I am the author report and would like to thank all the
anonymous and open reviewers for their invisible
contribution. A big thank all the teachers who have
supported and resolved every single query in order
to realize this paper.
References
Alsaadi M. I. (2015). International ournal of science and
technology research vol.4, issue 12. ISSN 2277-
8616 285
Deutschmann, I., Nordstrom, P., & Nilsson, L. (2013).
Continuous authentication using behavioral
biometrics. IT Professional, 15(4), 12-15
Eng, A.; Wahsheh, L.A. (2013). "Look into My Eyes: A
Survey of Biometric Security," in Information
Technology: New Generations (ITNG), Tenth
International Conference on , vol., no., pp.422-427,
15-17
Kataria, A.N.; Adhyaru, D.M.; Sharma, A.K.; Zaveri, T.H.
(2013). "A survey of automated biometric
authentication techniques," in Engineering
(NUiCONE). Nirma University International
Conference on , vol., no., pp.1-6, 28-30
Narhar, U.K.; Joshi, R.B. (2015). "Highly Secure
Authentication Scheme," in Computing
Communication Control and Automation
(ICCUBEA), 2015 International Conference on , vol.,
no., pp.270- 274, 26-27
Pahuja, G.; Nagabhushan, T.N. (2015). "Biometric
authentication & identification through behavioral
biometrics: A survey," in Cognitive Computing and
Information Processing (CCIP), 2015 International
Conference on , vol., no., pp.1-7, 3-4
Rathgeb, C. ,Breitinger, F. Busch, C. and Baier, H. (2014).
On application of bloom filters to iris biometrics. IET
Biometrics, 3(4):207–218
Weizhi Meng; Wong, D.S.; Furnell, S.; Jianying, Z.
(2015)"Surveying the Development of Biometric
User Authentication on Mobile Phones," in

Communications Surveys & Tutorials, IEEE ,
vol.17, no.3, pp.1268-1293
Ye Wang; Rane, S.; Draper, S.C.; Ishwar, P. (2012). "A
Theoretical Analysis of Authentication, Privacy,
andReusability Across Secure Biometric Systems,"
in Information Forensics and Security, IEEE
Transactions on , vol.7, no.6, pp.1825-1840
Zaeri, N. (2011) Minutiae-based fingerprint extraction and
recognition. In Computer and Information Science
- Artificial Intelligence - ”Biometrics”