net banking

Abstract—In this article we propose to model Net-banking
system by game theory. We adopt extensive game to model our web
application. We present the model in term of players and strategy.
We present UML diagram related the protocol game.
Keywords—Game theory, model, state, web application.
I. INTRODUCTION
AME theory is a branch of mathematics that studies the
interactions of multiple independent decision makers that
try to fulfill their own objectives. Today, it is applied to
telecommunications as the users try to ensure the best possible
quality of service. In recent years, game theoretic research on
ad hoc networking has emerged. Game theory is a set of
analytical developed to study situation in which self-interested
parties interact with each other according to rules. Thus, it
makes possible to model a wider range of real life situations.
The distance visualization with different interfaces knows
an important evolution. The interfaces permit to access to data
of type text and images. Several domains see an expansion of
using the web like e-commerce, e-business, e-learning… Net
banking application is one of more important web application.
These transfer systems require a security system which
protects these data during their transfer, because for raisons of
confidentiality.
Exchanged information is very important and should be
organized. We have to define who sends, who receives. Also,
we have to define constraints for each participant. Therefore
we have to offer to users (bank and client) fair, reliable and
rational exchange information on the different types of
networks. Game theory is adequate to model Net-Banking
system. Article contains an introduction as section 1. Section
2 gives background in theory game. Section 3 presents
extensive game. Section 4 presents Net baking system
modeled as game. Section 5 presents UML diagrams related
to game model. A conclusion finishes the article.
II. THEORY GAME BACKGROUND
Game theory is a set of analytical developed to study
situation in which self-interested parties interact with each
other according to rules. Since these kinds of situations occur
in, exchange protocols game theory is very appropriate.
N. Ghoualmi-Zine is with the Computer Sciences Department, Badji
Mokhtar University Annaba, 23000, Algeria (e-mail: ghoualmi@yahoo.fr).
A. Araar is with the Faculty of Computer Sciences, University of Ajman,
UAE (e-mail: araar@ajman.ac.ae).
Parties of a given exchange protocol find themselves as a
game. Game is called the protocol game. The protocol parties
are modeled as players. The protocol itself is represented as a
set of strategies (one strategy for each protocol party).
Games can be classified into different categories according
to their properties. We present below a brief classification [1].
• Non cooperative and cooperative games
• Strategic and extensive games
• Zero-sum games
• Games with perfect and imperfect information
• Games with complete and incomplete information
III. EXTENSIVE GAMES
Extensive games eliminate the limitation of the
simultaneous decisions, thus they make possible to model a
wider range of real life situations. Thus, we present below a
formal presentation and protocol game [2].
A Formal Presentation
Next, we formulate an extensive game based on [2]. It
should be noted that for simplicity the following formulation
does not allow simultaneous actions of the players, i.e. the
game has perfect information. An extensive game with
imperfect information can be formulated similarly. In the
strategic and extensive games, the solution of a game is a set
of actions or strategies that will result in Nash equilibrium.
• A set N (the set of players).
• A set H of sequences (finite or infinite) of actions
that satisfies the following three properties.
o The empty sequence θ; is a member of H.
o If ( )k K
ak =1... K ∈ H (where K may be
infinite) and L < K then ( ) k L
ak =1... ∈H.
o If an infinite sequence ( ) ∞
k =1
ak satisfies H for
every positive integer L then ( ) ∞
k =1
ak ∈H.
(Each member of H is a history; each component of a history
is an action taken by a player.) A history ( ) k K
ak =1... ∈ H is
terminal if it is infinite or if there is no (ak +1) such that
( ) k =1...K+1
ak ∈H. The set of terminal histories are denoted Z.
• A function P that assigns to each non terminal history
(each member of H\Z) a member of N. (P is the
Net-Banking System as a Game
N. Ghoualmi-Zine, and A. Araar
G
PROCEEDINGS OF WORLD ACADEMY OF SCIENCE, ENGINEERING AND TECHNOLOGY VOLUME 6 JUNE 2005 ISSN 1307-6884
PWASET VOLUME 6 JUNE 2005 ISSN 1307-6884 26 © 2005 WASET.ORG
player function, P(h) being the player who takes an
action after the history h.)
• For each player i ∈ N a utility function i U on Z.
B. Protocol games
The protocol game of an exchange protocol is intended to
model all the possible interactions of the (potentially
misbehaving) protocol parties. The correct behavior of each
party is represented by a particular strategy within the
protocol game.
We should note that we consider only two-party exchange
protocols (i.e., protocols that involve only two main parties
and possibly a trusted third party) because most of the
exchange protocols proposed in the literature are two-party
exchange protocols.
IV. NET-BANKING AS GAME
We assume that the network that is used by the protocol
participants to communicate with each other is reliable, which
means that it delivers messages to their intended destinations
within a constant time interval. Such a network allows the
protocol participants to interact in a synchronous fashion. We
will model this by assuming that the protocol participants
interact with each other in rounds, where each round consists
of the following two phases [3]:
1. each participant generates some messages based on her
current state, and sends them to some other participants;
2. each participant receives the messages that were sent to
her in the current round, and performs a state transition
based on her current state and the received messages.
A. Players in Net-Banking System
We model each protocol participant (i.e., the two main
parties) as player. In addition, we model the communication
network as a player too. Therefore, the player set P of the
protocol game is defined as P = (p , p , p , net) 1 2 3 where p1
and p2 represent the two main parties of the protocol, p3
stands for the trusted third party, and net denotes the network.
If the protocol does not use a trusted third party, then p3 is
omitted. We denote the set P \ {net} by P’.
Therefore, main players in Net-banking are costumer which
represents client and account that represents bank’s server
denoted by P’. We assume that network is reliable by
cryptographic systems. We shall present in next section a
cryptographic method.
B. Information sets in Net-banking
We define two types of events: send and receive events. The
send event snd (m; j) is generated for player i ∈ P' when she
submits a message π m∈M with intended destination
j ∈ P' to the network, and the receive event rcv (m) is
generated for player i ∈ P' when the network delivers a
message π m∈M to i. We denote the set of all events by E.
The local state Σ ( ) i q of player i ∈ P' after action
sequence q is defined as a tuple (q) H (q) r (q) α i , i , i
where:
o (q) i
α ∈{true; false} is a Boolean, which is true
iff player i is still active after action sequence q (i.e.,
she did not quit the protocol);
o H (q) i ⊆ E × N is player i’s local history after
action sequence q, which contains the events that
were generated for i together with the round number
of their generation;
o r (q) i ∈N is a non-negative integer that represents
the round numbers for player i after action sequence
q.
Initially, we have: (q) i
α = true,
H (q) i =φ , and r (q) i = 1 for every player i ∈ P' .
C. Protocol Game in Net-Banking system
In previous work we presented Syverson’s protocol as
protocol game that guaranties integrity, authentication, norepudiation
and confidentiality [4]. To secure exchange in
Net-Banking system we had proposed to mix cipher methods:
cipher at character level and at bit level. At Character level we
use θ- Vigenere that we developed in previous work [3]. At bit
level we use standards cipher like AES, DES, etc. We
introduce θ- Vigenere in Syverson’s protocol because
encryption key is random and encryption system is symmetric.
Syverson’s protocol has to transmit the triple, which
represents key in θ-Vigenere. So the receipt is allowed to
decrypt message at character level. The application consists
to encrypt at character level by θ- Vigenere cipher and with
random K at bit level. Where K’ is the basic key used by
Vigenere cipher and on which message is divided in blocs [5].
Then we obtain: 1- A and B denote the two participants; 2-
−1 , −1
A B K K denote their private keys, 3- Item A, item B denote
the items that they exchange, 4- dsc A denotes the description
of item A, 5- K denotes a randomly chosen secret key, 6- enc
is a symmetric-key encryption at bit level, 7- K’, θ denotes a
randomly chosen key, 8- enc’ is symmetric-key encryption at
character level where enc’= θ- Vigenere
PROCEEDINGS OF WORLD ACADEMY OF SCIENCE, ENGINEERING AND TECHNOLOGY VOLUME 6 JUNE 2005 ISSN 1307-6884
PWASET VOLUME 6 JUNE 2005 ISSN 1307-6884 27 © 2005 WASET.ORG
V. UML DIAGRAMS RELATED TO GAME MODEL
A. The ObjectDiagram
We present below the object diagram[7] that represents player parties (see Fig.1). The object customer represents client party
and account represents bank’s server party. Transaction is an object that represents history of client. For each transaction
between customer and his account we save in transaction class information, which save historic of last ten transactions.
Fig. 1 Object diagram
B. Connection Process Sequence Diagram
Each object in the Net-banking passes by different states through transitions; we study each object apart showing its different
states, events provoking transitions of these states.
PROCEEDINGS OF WORLD ACADEMY OF SCIENCE, ENGINEERING AND TECHNOLOGY VOLUME 6 JUNE 2005 ISSN 1307-6884
PWASET VOLUME 6 JUNE 2005 ISSN 1307-6884 28 © 2005 WASET.ORG
C. Payment Credit with Activated Account Sequence Diagram
VI. CONCLUSION
We have developed a Net-banking application because
this kind of application is in expansion on the web. We
modeled the application as game. Game theory is very
adequate for such application. We presented UML Diagram
that represents interactions. In Future work we’ll develop
protocol game as secured exchange protocol.
REFERENCES
[1] M. J. Osborne and A. Rubinstein. A Course in Game Theory. MIT
Press, Cambridge, 1994.
[2] T. Fent, G. Feichtinger, and G. Tragler. A dynamic game of
offending and law enforcement. International Game Theory Review,
4(1):71–89, 2002.
[3] N. Ghoualmi-Zine & A.Araar, “Secured Net-Banking by θ-vigenere
in Syverson’s protocol”, IEEE Catalog Number:05EX949, ISBN 0-
7803-8735-X, Library congress:2004110879, AICCSA 2005.
[4] Levente Buttyan, Jean Pierre Hubaux, Srdjan Capkun, ‘A formal
model of rational exchange and its application to the analysis of
Syverson’s protocol’, 2003 IOS Press, journal on computer security,
15th IEEE security foundation workshop, 2003.
[5] P. Syverson, ‘Weakly secret bit commitment: Applications to
lotteries and fair exchange’, In Proceedings of the IEEE computer
security foundations workshop, pp. 2-13, 1998.
[6] M. Jakobsson, J.P. Hubaux, and L. Huttyan. A micropayment
scheme encouraging collaboration in multi-hop cellular networks In
Proceedings of Financial Crypto 2003, January 2003.
[7] Conallen J., Concevoir des applications Web avec UML,Editions
Eyrolles, 2000.
PROCEEDINGS OF WORLD ACADEMY OF SCIENCE, ENGINEERING AND TECHNOLOGy