Content uploaded by Gaurav Agrawal
Author content
All content in this area was uploaded by Gaurav Agrawal on Jan 15, 2017
Content may be subject to copyright.
Abstract—Debit card or ATM card frauds had been a major
sector of concern due to which Reserve Bank of India (RBI) has
set new guidelines since December 1, 2013. Now you will have to
enter your personal identification number (PIN) every time you
swipe your card at any merchant outlet. Taking the existing state
of affairs into consideration, in this paper the conventional
security issues of the use of ATM and debit cards are discussed
along with the feasibility of other alternatives. Then, the research
paper will propose a model for secure use of the debit cards and
ATM cards via Chaos function and QR code (DACQ model) that
bolsters both speed and security without confounding the process
or making it undesirable to users.
Index Terms—ATM card, Chaos function, Debit card, QR
Code.
I. INTRODUCTION
uring the last three decades, the use of debit cards have
rapidly grown into extensively used payment instrument
at the point-of–sale (POS) in India. The rolling increase in the
use of debit cards has made it more alluring for various frauds.
This in turn is going to affect customers’ confidence on the use
of debit/ATM cards and eventually they will shift away to
other means of payment.
Security requirement of such cards include account
verification, user identity verification, information access
restriction and prevention of card tampering. Card security
should be included at all level of card’s life cycle which is
during its issuance, personalization, distribution and debit and
credit transaction [1]. In this paper we propose a DACQ model
to completely eradicate such frauds occurring at the POS.
Various approaches for debit transaction can be categorized as
follows:
Gaurav Agrawal is Asst. Professor in Department of Computer Science
and Engineering at Inderprastha Engineering College, Uttar Pradesh
Technical University, Sahibabad, Ghaziabad, U.P.-201010. (phone:
+919899987137; e-mail: gaurav.agrawal@ipec.org.in).
Akash Singh, student in Computer Science and Engineering, 4th year (B-
Tech), Inderprastha Engineering College, Uttar Pradesh Technical University,
Sahibabad, Ghaziabad, U.P.-201010 (phone: 9555582581 e-mail:
akashsingh.skyhorn@gmail.com).
A. Conventional approach of debit card usage:
1. Contemporary approach: As per Reserve Bank of
India (RBI) guidelines from 1st December, 2013 all
the debit card transactions at retail outlets will need to
be validated using your existing ATM PIN [2].
However, entering your PIN at POS is vulnerable to
threats like shoulder snooping, hidden pinhole
camera, etc. Now, all that the culprit needs is to get a
glimpse or an image of your debit card that contains
your card number, expiry date and CCV number
which is not a big deal considering the current
technical advancements. It should be noted here, now
the customer will be exposed to on-line frauds which
requires the above mentioned obstacles [3], [4].
2. Signature debit: This approach uses signature as a
means of user authentication at the POS [2]. Since,
the signature can easily be forged. So, such a
technique suffers from debit card theft.
3. Aadhaar-based biometric authentication: This
provides a secure payment infrastructure but incurs
with itself cost of maintaining huge amount of
database and is also less time efficient as compared to
the contemporary approaches [5].
4. NFC technology for contactless payments: A NFC-
enabled phone is provisioned with a payment
application and payment account information issued
by customers’ financial institution. The phone uses
NFC technology to merchants’ contactless payments-
capable POS system and the transaction occurs [6].
However, the major concerns for this approach is
that, in a developing country like India, it is not
possible for everyone to get their mobile replaced by
NFC embedded one. This is also prone to theft and
other frauds.
5. Based on the short message service (SMS): As SMS is
a universal source of communication and is available
in all mobile phones, pagers or other handsets.
However, service providers can try to deliver it but
cannot ensure its delivery. Moreover, SMS incurs
surplus charges. Thus, it is an infeasible approach and
is not necessary a low cost solution [7].
Debit /ATM card security based on chaos
function and QR code
Gaurav Agrawal, Akash Singh
D
International Conference on Communication and Signal Processing, April 3-5, 2014, India
978-1-4799-3357-0
Adhiparasakthi Engineering College, Melmaruvathur
372
6. QR code activated transaction: In this a QR code is
embedded in our mobile devices and all the purchases
are made scanning this QR code. If the merchant has
a QR code scanner, it can scan the QR code on the
mobile device to complete the transaction. If in case
the merchant does not have a scanner, a four-digit
code pops up on the shopper’s phone to be entered on
the PIN pad. QR code transactions ensure speedy
transactions but it is only applicable to smartphone
toting customers [8], [9]. In India where a major
section of population don’t possess a smartphone, you
cannot guarantee such a transaction procedure for
common people. This scheme is also liable to theft
frauds.
The above mentioned hindrances have evidently restricted
the use of debit cards and has ameliorated the fraud cost and
trends. Therefore, it is very relevant to devise a solution to the
above mentioned problems.
The rest of this paper is organized as fallows. Section II
deals with the introduction to DACQ model, section III
exemplifies DACQ model using henon map, section IV
provides us the speed analysis of this model, section V
describes changes required in the conventional approach,
section VI describes with its advantages and disadvantages,
section VII deals with the possible attacks. Further section
VIII takes into consideration its future scope. In section IX we
demonstrate our results. Finally, in section X we draw the
conclusions.
II. DEBIT CARD AND ATM CARD SECURITY USING CHAOS
FUNCTION AND QR CODE (DACQ) MODEL:
This paper proposes a DACQ model that will address both
concerns of security and transaction speed.
This will be another way to eliminate fraud transactions
caused by debit card theft or by shoulder snooping others PIN
at the POS. The traditional payments via debit card requires at
least two tokens which are expected from the customers to
prove an authenticated transaction. These two involves the
existence of card itself, and the signature or the PIN which is
getting more popular at POS than signatures. Both the tokens
exist with the customer in different places. The debit card is
placed in customers’ wallet, while the signature or the ATM
PIN usually exist in his brain. Following the same concept, this
proposal requires two tokens to verify customers’ identity. One
is your enhanced debit card which has a QR code on it, while
the other is the token supplied to you specifically for any
public transaction or to be used at any POS, which is to say
that this token will not be same as your ATM PIN.
For the issue of speed of setting a transaction, the
transaction initiator must be a merchant [3]. This is because
the merchants generally possess a more reliable and
continuous connection with the third party involved in the
transaction, who could be either a bank or any other service
provider. Secondly, since we are using QR code in our debit
card, it ensures that we get a large storage capacity, small
printout size and a very high speed scan [10], [11]. Also
because you are already verifying it via QR code at the
merchant store, you can further compensate this time by
removing the signature token delivered by the customers at the
POS, as signatures can easily be forged and merchants are no
experts at judging your identity through a signature and by the
time authorities recognizes the criminal’s false identity, fraud
transactions have already occurred. Further, since there is no
handshaking involved between the customer and the third
party for transaction confirmation, the processing time is
expected to be faster than the traditional approach.
Having discussed both the concerns, we can describe the
following scenario. While the customer waits in line for the
cashier, on his turn he scans his card to the QR code reader
and enters the seed value provided to him by the bank. This
step acknowledges the identity of the card holder. On scanning
QR code, the card reader takes the stored pattern as input and
further asks for a private key. With the help of this private key
the card reader machine generates the pattern based on the
algorithm provided by different cards. This pattern is further
matched with the one scanned by the QR code reader. If they
are found to be similar then the user’s identity is proved. This
will further enable the gateway to the third party or the bank
itself for further transaction processing which requires swiping
of the card on the card reader. Depending on the card used,
customers’ bank will transfer the money, or the service
provider will carry on the money transfer from the customers’
to the merchants’ bank [12].
III. WORKING AND CONFIGURATIONS
DACQ model uses a QR code and a chaos function to
generate a random pattern which is stored to form a string of
characters in QR code. In an attempt to enhance the security of
such cards this paper uses a chaotic map to generate cypher
based crypto system. Chaotic map possess properties like
pseudorandom behavior, topological transitivity and sensitive
dependence on initial conditions. Due to such properties they
have been used to develop various block and stream cipher
based cryptosystem [13]. The proposed cryptosystem involves
the use of henon map. It is a 2D map which takes a point (xn,
yn) and maps it to a new point (xn+1, yn+1) [14]. This is
expressed as:
xn+1 = yn+1 – axn2
yn+1 = bxn
The map depends on two parameters ‘a’ and ‘b’ and also on
the initial values of ‘x’ and ‘y’. At different values of ‘a’ and
‘b’ the map may be chaotic, intermittent or converge to a
periodic orbit. The randomized pattern generated is stored in a
QR code.
QR code is a matrix type two dimensional barcode which
indicates information with black and white cross stripes. QR
code is used in this scheme due to its following properties
[10], [11]:
1. Sufficient data capacity:
Numeric : 7089
Alphanumeric: 4296
373
Binary : 2953
Kanji : 1817
2. High speed scan
3. Small printout size
4. Omni-direction readability
5. Error correction capability
Level L: about 7% or less error can be corrected.
Level M: about 15% or less error can be corrected.
Level Q: about 25% or less error can be corrected.
Level H: about 30% or less errors can be corrected.
- Algorithm to generate pattern:
Step1: Perform several iterations with some initial conditions
of x and y. Here we have taken number of iterations as 16.
Step2: Generate a chaos matrix by assembling the x and y as
alternate elements of the chaos matrix as shown in fig. 1.
X1
Y1
X2
Y2
0.1000
0
0.9860
0.0300
-0.3311
0.2958
1.1423
-0.0993
-0.9263
0.3427
0.1416
-0.2779
0.6941
0.0425
0.3681
0.2082
Fig. 1. Chaos matrix taking initial values as x=0.1 and y=0. Values of
constants a=1.40 and b=0.30
Step3: Generate a pattern by appending the matrix elements
after ignoring the initial values of x and y. A positive value is
denoted by a ‘0’ whereas a negative value is represented by a
‘1’. The pattern for the chaos matrix shown in fig. 1 is given
below,
00986000030010331100295801142310099310926300342700
1416102779006941000425003681002082
The pattern generated will be very different for different
values of ‘a’ and ‘b’. To exemplify this let’s take a=1.35 and
b=0.30. In this case the value of pattern generated is,
00986500030010283800296001187210085110988000356200
0385102964007016000115003470002105
Step4: This pattern is further appended to the end a numeric
code which will be used as a token to distinguish between
varieties of cards. This numeric code should be of fixed
number of digits. In this research paper we have taken two
digits to represent the card type. For example, Master card can
have code: 32. This 32 will be appended at the beginning of
the generated pattern.
32009860000300103311002958011423100993109263003427
001416102779006941000425003681002082
Step5: This pattern is further stored in QR code. The generated
QR code of the above pattern is shown in fig. 2
Fig. 2. QR code with embedded chaos pattern when a=1.40 and b=0.30.
Step5: Further, this QR code is embedded on your debit cards.
Step6: Key can be generated by directly appending the decimal
part of seed value of the chaos function used. For example, if
we are using a=1.30 and b=.40 as seed values then we append
the decimal values as 3040. This acts as the key for the
customer. This can be further ameliorated by using various
encryption algorithms but for the simplicity of our proposal we
have used a direct approach in this research paper.
IV. SPEED TEST FOR DACQ BASED VERIFICATION:
In order to test the efficiency of our model, the simulator
analysis is done on the computer of 4.0 Giga byte of memory,
64 bit operating system, Intel® Core™ i7-2670QM CPU of
2.20Ghz (8 CPUs), ~2.2GHz processor. The experiment is
implemented using MATLAB R2012a software. The analysis
is shown in table1:
S.
N
o.
Pattern
generation
time(sec)->t1
Pattern
matching
time(sec)->t2
Total
time(sec)
t=t1+t2
Cumulativ
-e t(sec)
1
0.035896
0.000196
0.036092
0.036092
2
0.036972
0.000165
0.037137
0.073229
3
0.035247
0.000268
0.035515
0.108744
4
0.035393
0.000164
0.035557
0.144301
5
0.037549
0.000166
0.037715
0.182016
Table1: Statistical analysis of time consumption
Fig. 3. Total Pattern generation time with increasing card swipes.
QR code takes around 1 sec to get scanned whereas PIN
based approach requires a handshaking between bank or any
other third party and the merchant and also a signature to
verify customers’ authenticity. Also, the time spent for
processing for five customers is nearly equivalent to 0.2
seconds by which we can infer that for a thousand transactions
the net time spent will be 200 seconds, which can be easily
neglected for enhancing the security at POS, no matter whether
the bank continues its signature policies or not.
374
V. PRELIMINARIES
A. The Card manufacturers:
The card manufacturers ensure that the following things are
present on the debit cards distributed by them to the
customers. The following steps are required to make enhance
the existing debit cards:
1. Generate a chaos pattern using a suitable chaos
function. To exemplify this, Hanon map is used in
this research paper.
2. A unique pattern gets generated at different initial
values or different constant values. These initial
values are converted in the form a private key before
distributing it to the customers as a token for their
authenticity.
3. The pattern values are stored to form a pattern of
characters in QR code.
4. Place this QR code on the conventional Debit cards.
B. At the Point-of –sale (POS):
The following steps have to be followed in the defined order:
1. Card holder hands over his debit card to the merchant.
2. The merchant uses a QR code reader to take the
random pattern as input from the customers debit
card.
3. Then the customer is asked to enter his private key in
the swiping machine. Not to forget that this key is not
your ATM PIN.
4. Based on the key a cipher is produced which is
converted into the pattern and is further matched with
input pattern of customers’ debit card.
5. If the pattern matches with the input pattern then
normal procedure for the transaction is carried out
which involves swapping the magnetic strip on the
swapping machine and further making the transaction.
6. If the pattern does not match which implies that the
customer is not an authenticated user of the debit card
and he is given 2 more attempts before the card gets
blocked.
C. The debit card machine manufacturers:
1. Embed the algorithm provided by different card
manufacturers on the card reader.
2. Enhance the existing card readers by affixing a QR
code reader which acts as a source of input to the
above mentioned algorithm.
3. Don’t open the gateway to bank’s processor until the
authenticity of user is verified via above mentioned
algorithm.
Note: Access to card agent should be disabled unless
the client’s authenticity is verified through QR code
VI. ADVANTAGES AND DISADVANTAGES
Advantages of this method:
1. The actual decryption key is neither stored in the card
nor in the card reader. This protects the system from
direct key recovery attacks.
2. Each card has its own private key.
3. Authentication procedure requires mutual collaboration
of both card and the card reader. So, to compromise
with the system, the culprit requires access to not only
the card of the victim but also the algorithm stored in
the card reader. However this alone will not suffice
the culprit’s requirements, he will further need the
private key which again is only available in victim’s
brain. Hence the procedure becomes almost
impossible to crack.
4. It ensures that the users don’t have to use their ATM
PIN publically which in turn secures them from fraud
internet purchase or any other transactions, as all
culprit needs for any e-purchase is the victim’s card
number, CCV number, his name and his ATM PIN.
Card number and CCV number can be easily acquired
as it is directly displayed on the victim’s card itself
and his name is his public identity which makes it
easily available. The only thing that avoid fraud e-
transactions by the culprit is the victim’s ATM PIN.
This procedure completely avoids this risk.
5. This approach requires no change in the business logic
at the bank’s processor and no extra database
handling is required.
Disadvantages of this method:
1. The card holder has to get his card enhanced.
2. This approach requires new card readers.
No (3 times)
Yes
Fig. 3. Flowchart to demonstrate the working of DACQ model
Debit Card
QR code reader to read the
pattern stored in QR code (Pc)
Enter the private key
Pattern gets generated (P) based on the
pattern key
Pc= =P
Swipe the card in the swiping
machine
Bank’s processor for transaction processing
375
VII. ATTACKS
1. Card theft: Debit/ATM card theft is no more a matter
of concern as it is not the only token required for
transactions at the POS, you will also need the private
key of the victim. By the time you plan to get his
private key, the user will get sufficient time to report
it to authorities and disable his bank account until a
new card is issued to him [15].
2. Card tampering: Once the QR code is embedded on
the card, write access to it gets disabled which further
protects it against accidental or intentional accidents.
3. Card forgery: Each card is personalized with a
specific PIN for ATM machines or for any internet
based transactions and a special private key used for
QR code based authentication at the POS [16].
4. Private Key guessing approach: This technique will
not work because the private key used here is quite
different from the seed value of the chaos function
and also you need other tokens to get yourself
authenticated which obviously becomes a impossible
combination [16].
VIII. FUTURE SCOPE
You can see QR code everywhere on marketing materials.
Newspapers, on receipts, posters, etc. It is an easy way to gain
information at a quick snap of your camera phone.
Conventionally when we want to know our balance, either we
have to visit some ATM machine or we have to make a call to
the customer care of our bank and then after taking us to
several different options we finally have to enter our 16 digit
card number which itself makes it a time consuming process.
This procedure can further be enhanced to resolve the above
mentioned issue. The idea is very simple and all you have to
do is to scan the QR code and you can easily retrieve the
information that you want. To ameliorate it further, it can also
be utilized to find the closest ATM.
All that needs to be done is, just make an application freely
available to the customers which is capable of scanning QR
code and which transfers your scanned data to the respective
bank for information regarding your account. The customer on
the other side will download the application and can easily use
it like any other QR code scanner.
This will not only enhance customers’ satisfaction but also
reduce the number of phone calls that the bank receives on
daily basis [17]. Further, this reduced time spent no receiving
calls will payoff for the money spent on implementing this
procedure. The bank capable of implementing this much can
be an innovator which might fulfill their dream of placing their
cards on top of everyone’s wallet.
IX. RESULT
The result demonstrates that DACQ approach fulfills both
security and speed issues and also makes it more reliable and
easy for the customers to get used to it.
X. CONCLUSION
DACQ approach is a very feasible method to fight against
Debit/ATM card based frauds. This research paper is the first
to propose the chaos base QR code verification at the POS
which not only eliminates the usage of the PIN verification
table, but also is a very cost effective way to eliminate such
frauds.
REFERENCES
[1] Attoh-Okine, N. O. and Shen, L. D., "Security Issues of Emerging Smart
Cards Fare Collection Application in Mass Transit," Pacific Rim
TransTech Conference, pp. 523-526, 1995.
[2] Master Circular on Credit Card, Debit Card and Rupee Denominated
Co-branded Prepaid Card operations of banks,
www.rbi.org.in/scripts/BS_ViewMasCirculardetails.aspx?id=8087
[online], R.B.I., pp. 17-18, 2013.
[3] Martin Emms, Budi Arief, Nicholas Little, and Aad van Moorsel,” Risk
of Offline Verify PIN on Contactless Cards,” Financial Cryptography
and Data Security Lecture Notes in Computer Science, vol. 7859, pp.
313-321, 2013
[4] Malcolm Reay, Siemens Insight Consulting,” Chip and PIN, a new
challenge for data security,” Computer Fraud and Security, vol. 2006,
pp. 4-5, 2006.
[5] S. Dharanya, and A. Umamakeswari,” Embedded Based Conveyance
Authentication and Notification System,” International Journal of
Engineering and Technology( IJET), vol. 5 no. 1, pp. 410, 2013.
[6] Harvey Glickenstein (Senior Editor),”Contactless Payments Debuts on
London’s Buses New and Improved Railways Around the World,”
Vehicular Technology Magazine, vol. 8, pp. 19-25, 2013.
[7] Kuan-Chieh Liao, Min-Hsuan Sung, Wei-Hsun Lee, Ting-Ching Lin,”A
one-time password scheme with QR code on mobile phone,” INC, IMS
and IDC.Fifth International Joint Conference, pp. 2069-2071, 2009.
[8] Sana Nseir, Nael Hizrallah and Musbah Aqel, “ A Secure Mobile
Payment System using QR code,” 5th International Conference on
Computer Science and Information Technology (CSIT),pp 111-114,
2013.
[9] Young-Gon Kim and Moon-Seog Jun,” Design of user authentication
System using QR code Identification Method,” 6th International
Conference on Computer Sciences and Convergence Information
Technology (ICCIT), pp. 31-35, 2011.
[10] Phaisarn Sutheebanjard and Wichian Premchaiswadi,”QR-Code
Generator,” eighth International conference on ICT and knowledge
engineering, pp. 89-92, 2010.
[11] Satrid Vongpradhip,” Use Multiplexing to increase information in QR
code,” The 8th International Conference on Computer Science and
Education (ICCSE), pp. 361-364, 2013.
[12] Fumiko Hayashi, Richard Sullivan, and Stuart E. Weiner,”A guide to
the ATM and Debit Card Industry”, Federal Reserve Bank of Kansas
City, pp. 17-68, 2003.
[13] Ljupco Kocarev, “Chaos based Cryptography: a brief overview,”
Circuits and Systems, IEEE, vol. 1, pp. 6-21, 2001.
[14] Nidhi Taneja, Balasubramanian Raman and Indra Gupta, “Chaos based
cryptosystem for still visual data,” Multimedia Tools and Applications,
vol. 61, no. 2, pp. 281-298, 2012.
[15] Jose M. Pavia, Ernesto J. Veres-Ferrer, Gabriel Foix-Escura, “Credit
card incidents and control systems,” International Journal of
Information Management, vol. 32, pp. 501-503, 2012.
[16] C. Brzuska, N.P. Smart, B. Warinschi, and G.J. Watson,”An analysis of
the EMV Channel Establishment Protocol,” ACM SIGSAC conference
on Computer & communications security, pp. 373-386, 2013.
[17] Reghunathan Sukumara Pillai, and Santhy Sreedhar,“Infosys Finacle-
Banking in India: evolution in technology,”2012.
376
