International Journal of Applied Engineering

Research and Development (IJAERD)

ISSN 22501584
Vol.2, Issue 2 June 2012 19-31
TJPRC Pvt. Ltd.,

FIXED LENGTH OF INFECTIVE PERIOD FOR
ATTACKING WORMS IN COMPUTER NETWORK
a1
BIMAL KUMAR MISHRA,
b
UPENDRA KUMAR &
c
G. SAHOO
a
Department of Applied Mathematics, Birla Institute of Technology, Mesra,
Ranchi, India.
b,c
Department of Information Technology, Birla Institute of Technology, Mesra,
Ranchi, India
ABSTRACT
Susceptible(S)-Infectious (I) model for transmission of worms in Computer
network is formulated. In the SIS model for the infective periods of fixed length
due to the attack of computer worms three different epidemic models have been
formulated. Threshold, equilibria and their stability are found with cyber mass
actions incidence. Thresholds, R
0
determine the outcomes of the disease. If R
0
1, the infective fraction of model disappear so the disease die out, while R
0
>1,
the infected fraction persist and the feasible region is an asymptotic stable region
for the endemic equilibrium state. Numerical methods are employed to solve and
simulate the system of equation developed. The effect of time delay on infected
nodes has been analyzed which also includes worms transmission in vertical
ways on the nodes of computer network.
KEYWORDS : SIS, Computer network,Attacking Worms.
INTRODUCTION
Computer worms arose in early eighties which can self replicate, damage
computer system, erase data, steal information as well as modify normal
operation. The major source of computer worms is electronic mail and


Bimal Kumar Mishra, Upendra Kumar & G. Sahoo

20

S I S
S(t)I(t) S(t-)I(t-)+I(t-)
I(t
secondary storage devices. Their action throughout the network can be studied
by using classical epidemiological model for diseases propagation[4-13].Based
on the Kermack and McKendrick SIR classical epidemic model[14-
16],dynamical models for malicious objects propagation were proposed,
providing estimations for temporal evaluations of infected nodes depending on
the network parameters considering topological aspects of the network[4-6,17-
20]. The kind of approach is applied to email propagation schemes [21] and
modification of SIR model generates guide for infection prevention by using the
concept of epidemiological threshold[4-6,22].Richard and Mark[23] proposed an
improved SEI model to simulate virus propagation. The concept of time delay on
the transmission of malicious code has been implemented by Mishra and Saini
[4] who has presented a SEIRS model with latent and temporary immune
periods, which can reveal worm propagation. Mishra and Jha[6] implement the
idea of fixed period of temporary immunity in computer network. Y.Michel,
H.Smith, L.Wang [24] considered global dynamics of SIER epidemic model
with vertical transmission. Fred Brauer, Carlos Castillo Chavez [1] gives
concept of infective periods of fixed length. SIS model has been designed to
know how worms are vertically transmitted and how much time it has taken to
self replicate.
In this paper three different models for the infective periods of fixed length
in computer network due to the attack of the worms has been formulated.






21 Fixed Length of Infective Period for Attacking Worms in Computer
Network
Figure 1 : Schematic diagram for the flow of Worms
MATHEMATICAL MODEL FORMULATION
(A) The population size K is constant
The constant population (that is, the total number of nodes in computer
network) K is partitioned into two sub classes susceptible and infectious denoted
by S(t) and I(t) respectively as depicted in figure 1. Our assumption is to
implement the concept of time delay including vertical transmission of worms in
a computer network. The SIS model with mass action incidence is

[ ] ) ( ) ( ) ( ) ( ) ( ) ( + + = t I t I t S t I t I t S
dt
dI
(1)
Where and are real positive constant, is time delay for transmission of
worms. Since every constant value of I is a solution of (1), to accomplish this we
are incorporating the initial data for - 0 t into model by writing it in
integrated form

{ } [ ]dx
t
t
x I x I x I k I

+ =

) ( ) ( ) ( (2)
describing I(t) as the integral of member infected at time x between t- to t and
therefore still infective. The equilibrium condition for (2) is I=0 or
I K I + = This gives a diseases-free equilibrium I=0 and endemic
equilibrium


1

|
|

\
|
+ = K I (3)
Bimal Kumar Mishra, Upendra Kumar & G. Sahoo

22
To linearize about an equilibrium I , we substitute I=I +u(t). We
neglect the quadratic term to give the linearization
I

+u(t)
= ( ) ( ) [ ]dx
t
x u I dx x u I k

+ +

) ( ) ( 2
= ( ) [ ]


+ +
t
t
dx I dx x u I K
t

) ( 2
The characteristic equation is condition on that the linearization has a
solution that is a constant multiple of e
t
, namely

( ) { }
1
1 )( ) 2 ( (
) ( 2
) (t
=
)
`

+
)
`

+ =

e I k
or
t
e I
e
I k
t
e
t
e
(4)

At the disease free equilibrium state I

=0

)
`

+ =

e
k I
1
(5)
Lemma1-: If 0 > and 0 R then

e 1

It follows that if 1 < + k then for 0 R the absolute value of left
side of equation(5)is no greater than one and thus there can be no root of the
23 Fixed Length of Infective Period for Attacking Worms in Computer
Network
equation(5) with 0 and the disease free equilibrium is asymptotically
stable if
1 0 + = K R , at the endemic equilibrium

+
|

\
|

+
|

\
|

)
`

= I
e e k
I
1 1 2 2

if R0 the absolute value of the left side of characteristic equation is
K 2 thus if R
0
>1 then there is no root with R0 and equilibrium
is asymptotically stable.
(B) The total population size is not necessarily constant
We begin with assumption that there is no crashing of the nodes due to
attack of worms, this imply that if the birth rate (new attached nodes) is k and
death rate is N, the differential equation for N is
N K N =
In order to derive differential equation for I we have to use following two
facts:
(I) The rate of infective at time t is
( ) ( ) ( ) = + t I t I t S ( ) ( ) ( ) [ ] ( ) t I t I t N t I +
(II) The rate of recoveries after running of antivirus at time t is the rate
of new infection S(t-)I(t-) at time (t-) multiplied by the fraction
e
-
of these new infection who survive until time and the rate of
recoveries due to vertical infection transmission.
The model is represented in following form,
( ) ( ) ( ) [ ] ( ) ( ) ( ) ( ) [ ] ( )

+ = t I t I t N t I
t
e t I t I t N t I I (6)
Bimal Kumar Mishra, Upendra Kumar & G. Sahoo

24
N K N =
There is a worm free equilibrium I=0, N=0 and endemic equilibrium
becomes

) 1 (

=
e
K I , N=K which exist if
1
) 1 (
0 f

=
e K
R
The characteristic equation at an equilibrium = I I , K N = is

[ ]
1
) (
1 ) 2 (
=
+
+



e I K
(7)
Lemma-2: If 0 , 0 R then
( )

+
+

e 1

e 1

It follows that if
1
) 1 (
0 <

e K
R
then for R0 the absolute value of left side of equation (7) with I

=0 is no
greater than one and there can be no root of the equation (7) with R0 and thus
the worm free equilibrium is asymptotically stable if R
0
<1. At the endemic
equilibrium
I K 2 ( )= K
e

1
2

and absolute value of the left side of equation is no greater than 2-R
0
which is
no greater than one. Also ,since R
0
>1 again there can be no root of the
25 Fixed Length of Infective Period for Attacking Worms in Computer
Network
characteristic equation with R0 and thus endemic equilibrium is
asymptotically stable if R
0
>1
(C) Crashing of the nodes due to attack of worms model (B)
We assume that at the end of the fixed infective period a fraction p of
infective nodes crashes die while remainder return to the susceptible class then
death rate due to worm is
( ) ( ) ( ) [ ] ( )

t I t I t N t I
t
pe
The model is represented in following form
( ) t I I = ( ) ( ) [ ] ( ) t I t I t N + -
( ) ( ) ( ) [ ] ( )

t I t I t N t I
t
e (8)
( ) = t N K N ( ) ( ) ( ) [ ] t I t N t I
t
pe

The analysis of the model leads to the same value for R
0

) 1 (
0

=
e K
R

To linearize about an equilibrium I ,we substitute


( ) t u I I + =

,gives

( ) [ ]
)
`

+ + = +

e
x u I P K x u I t u I
1
) ( )) ( (


The characteristic equation is the condition on that the linearization have a
solution that a constant multiple of e

=
( )
( )
{ }

+
+
|

\
|
e I P N 1 2

Bimal Kumar Mishra, Upendra Kumar & G. Sahoo

26

+
+ =
I p
1 (9)
If R
0
<1, I

=0 the right side of equation(9) is one, while left side has

absolute value at most one. If R
0
>1at the endemic equilibrium, for which,
( )

=
)
`

e
I N
1
, we may denote left side of equation (9)
has absolute value at most one using ( ) I P N 2 I N ,it may be
shown that if R0 The real part of
+
PBI
is non negative and it follows the
right side of equation(9)has absolute value greater than one. From this we might
deduce that the disease free equilibrium is asymptotically stable if and only if R
0
<1, and endemic equilibrium which exist if and only if R
0
>1 is asymptotic
stable. This is true for every p, 0p1.
Lemma(3):- Lemma(1) has been extended which follows from the estimate
,valid for 0 & 0 > R

( )
dx e
xdx
e

+

+ |

\
|

0
0

=

0 0
xdx
e
xdx
e
x
e
=

e 1

and the integral expression is,
27 Fixed Length of Infective Period for Attacking Worms in Computer
Network

( )

+
+

=
+
|

\
|

e
x
xd
e
1
0

The case p=1 of (9) in which all infective nodes crashes from the worm at
the end of infective period.
NUMERICAL METHODS AND DISCUSSION
In the SIS model for the infective periods of fixed length due to the attack of
computer worms, three different epidemic models have been formulated. Runge-
Kutta Method of order four is employed to solve and simulate the system of
equation developed. After simulation from the figure 3, we observe that number
of infectivity nodes in a computer network oscillates after constant period of
time (approximately 180 minutes). From the figure 2, we again observe that
initially number of susceptible nodes is 10000 but after 90 minutes (approx)
number of susceptible nodes is 0 which after approximately same period,
number of susceptible nodes comes back to 10000. This shows the oscillatory
behavior of the susceptible nodes in the computer network.

Bimal Kumar Mishra, Upendra Kumar & G. Sahoo

28
Figure 2 : Susceptibility verses time (in minutes), S(0)= 10000, I(0)=1633,
= 0.2 and =0.01

Figure 3 : Infectivity verses time (in minutes) S(0)= 10000, I(0)=1633, =
0.2 and =0.01
29 Fixed Length of Infective Period for Attacking Worms in Computer
Network
CONCLUSIONS
The computer epidemic SIS model for the infective periods of fixed length
due to the attack of computer worms has been developed. Threshold equilibria
and their stability are found with cyber mass action incidence. Threshold, R
0

determines the outcome of the disease. If R
0
1, the infective fraction of model
disappear, so the disease die out, while R
0
>1 ,the infected fraction persist and
the feasible region is an asymptotic stable region for the endemic equilibrium
state. R-K method of order 4 was employed to solve and simulate the system of
equation developed. It has been observed that the effect of time delay on
infected nodes play an important role on worm transmission in computer
network
REFERENCES
[1] Fred Brauer, Carlos Castillo_Chavez, Mathematical models in
Population Biology and Epidemiology, Springer-Verlag New York
,2001.
[2] R.M. Anderson, R.M. May, Infectious Diseases of Humans, Dynamics
and Control, Oxford University Press, Oxford, 1992.
[3] R.M. Anderson, R.M. May, Population biology of infectious disease I,
Nature 180, pp.361367,1999
[4] Bimal Kumar Mishra, D.K. Saini, SEIRS epidemic model with delay for
transmission of malicious objects in computer network, Appl. Math.
Comput. 188 (2) pp.14761482.,2007
[5] Bimal Kumar Mishra, Dinesh Saini, Mathematical models on computer
virus, Appl. Math. Comput. 187 (2) 929936,2007.
[6] Bimal Kumar Mishra, Navnit Jha, Fixed period of temporary immunity
after run of anti-malicious software on computer nodes, Appl. Math.
Comput.190 (2) 12071212., 2007
Bimal Kumar Mishra, Upendra Kumar & G. Sahoo

30
[7] E. Gelenbe, Dealing with software viruses: a biological paradigm,
Inform. Security Technical Rep. 12 (4) 242250. 2007
[8] Erol Gelenbe, Keeping viruses under control, in: Computer and
Information Sciences ISCIS 2005, 20th International Symposium, vol.
3733, Lecture Notes in Computer Science, Springer, October 2005.
[9] Erol Gelenbe, Varol Kaptan, Yu Wang, Biological metaphors for agent
behaviour, in: Computer and Information Sciences ISCIS, 19th
International Symposium, vol. 3280, Lecture Notes in Computer Science,
Springer-Verlag, October 2004, pp. 667675. 2004
[10] J.R.C. Piqueira, F.B. Cesar, Dynamical models for computer virus
propagation, Math. Prob. Eng., doi:10.1155//940526, 2008
[11] J.R.C. Piqueira, B.F. Navarro, L.H.A. Monteiro, Epidemiological models
applied to virus in computer networks, J. Comput. Sci. 1 (1) pp.3134.,
2005
[12] S. Forest, S. Hofmeyr, A. Somayaji, T. Longstaff, Selfnonself
discrimination in a computer, in: Proceedings of IEEE Symposium on
Computer Security and Privacy, , pp. 202212.1994.
[13] Y. Wang, C.X. Wang, Modeling the effects of timing parameters on virus
propagation, in: 2003 ACM Workshop on Rapid Malcode, ACM,
October,pp. 6166. 2003
[14] W.O. Kermack, A.G. McKendrick, Contributions of mathematical theory
to epidemics, Proc. Royal Soc. London Series A 115 pp.700721. 1927.
[15] W.O. Kermack, A.G. McKendrick, Contributions of mathematical theory
to epidemics, Proc. Roy. Soc. London Series A 138 pp.5583,1932.
[16] W.O. Kermack, A.G. McKendrick, Contributions of mathematical theory
to epidemics, Proc. Roy. Soc. London Series A 141 pp.94122.,1933
31 Fixed Length of Infective Period for Attacking Worms in Computer
Network
[17] C.C. Zou, W.B. Gong, D. Towsley, L.X. Gao, The monitoring and early
detection of internet worms, IEEE/ACM Trans. Network. 13 (5) 961
974,2005.
[18] J.O. Kephart, S.R. White, D.M. Chess, Computers and epidemiology,
IEEE Spectrum (2026, 1993
[19] M.J. Keeling, K.T.D. Eames, Networks and epidemic models, J. Roy.
Soc. Interf. 2 (4) 295307, 2005.
[20] Ma. M. Williamson, J. Leill, An Epidemiological Model of Virus Spread
and Cleanup, , <http://www.hpl.hp.com/techreports/>.2003,
[21] M.E.J. Newman, S. Forrest, J. Balthrop, Email networks and the spread of
computer virus, Phys. Rev. E 66 () 035101-1035101-4. 2002,
[22] M. Draief, A. Ganesh, L. Massouili, Thresholds for virus spread on
networks, Ann. Appl. Prob. 18 (2) pp.359378. 2008.
[23] W.T. Richard, J.C. Mark, Modeling virus propagation in peer-to-peer
networks, in: IEEE International Conference on Information,
Communications and Signal Processing (ICICS pp. 981985., 2005.
[24] Y. Michael, H. Smith, L. Wang, Global dynamics of SIER epidemic
model with vertical transmission, SIAM Journal of Applied Mathematics
62 (1) pp.5869. 2001
[25] J.K. Hale, Ordinary Differential Equations, second ed., Krieger, Basel,
1980.
[26] M.E.J. Newman, Stephanie Forrest, Justin Balthrop, Email networks and
the spread of computer viruses, Phys. Rev. E 66 pp.035101035104. 715,
2002.