Indonesi
an
Journa
l
of El
ect
ri
cal Engineer
ing
an
d
Comp
ut
er
Scie
nce
Vo
l.
1
3
,
No.
2
,
Febr
uar
y
201
9
, pp.
6
83
~
6
88
IS
S
N: 25
02
-
4752, DO
I: 10
.11
591/ijeecs
.v1
3
.i
2
.pp
6
83
-
6
88
683
Journ
al h
om
e
page
:
http:
//
ia
es
core.c
om/j
ourn
als/i
ndex.
ph
p/ij
eecs
Behavio
ra
l
and
perfo
rmance jell
yfish att
ac
k
Layth
A. Kh
al
il
A
l Dul
aimi
1
R.
Ba
dli
shah
Ah
m
ad
2
,
Nai
mah
Y
aakob
3
,
Syadi
ah
N
or
Wan Sh
ams
u
ddin
4
,
Moham
ed El
s
ha
ikh
5
1,3,5
Embedde
d
Networki
ng
and
Advanc
e
Com
pu
ti
ng
Clust
er,
Sch
ool
of
Com
pute
r
&
Com
m
unic
ati
on
Engi
n
ee
ring
,
Univer
siti
Malays
ia
Perl
is (
UniMA
P) T
ingka
t
1,
Kam
pus T
et
ap
Pauh,
Per
li
s,
Mal
a
y
sia
2,4
Facul
t
y
of
Inf
orm
at
ic
s a
nd
Co
m
puti
ng,
Univer
siti
Sult
an Za
in
a
l
Abidin
,
21300
Kuala
Te
r
engganu,
Mal
a
y
s
ia
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
Sep
3
, 2
018
Re
vised
N
ov
1
3
, 2
018
Accepte
d
Nov
2
6
, 201
8
W
e
provide
a
si
m
ula
ti
on
-
base
d
stud
y
of
the
eff
ec
ts
of
jell
y
fish
at
t
ac
ks
on
m
obil
e
ad
hoc
n
e
tworks
(MA
NE
Ts).
For
thi
s
pur
pose
we
suggest
a
sim
ula
ti
on
base
d
on
th
e
eff
ec
ts
of
je
l
l
y
fish
at
t
ac
ks
on
th
e
n
et
work
through
a
num
ber
o
f
diffe
ren
t
sc
ena
r
i
os.
In
p
articu
l
ar,
we
exa
m
ine
ho
w
the
num
b
er
o
f
atta
cke
rs
aff
ects
per
form
a
nce
m
ea
sures
su
ch
as
th
e
ratio
of
pac
ke
t
de
li
ver
y
,
throughput
,
and
end
-
to
-
end
del
a
y
s.
The
resu
lt
s
have
en
abl
ed
us
to
propose
m
ea
sures
to
red
uce t
h
e effe
c
t
s of
jell
y
fish
atta
cks
on
MA
NETs.
Ke
yw
or
d
s
:
AOD
V
Jel
ly
fish
at
ta
ck
Secu
rity
o
f A
d ho
c
n
et
wor
ks
Copyright
©
201
9
Instit
ut
e
o
f Ad
vanc
ed
Engi
n
ee
r
ing
and
S
cienc
e
.
Al
l
rights re
serv
ed
.
Corres
pond
in
g
Aut
h
or
:
Lay
th A.
Kh
al
i
l Al Dula
im
i,
School
of Com
pu
te
r
&
Com
m
un
ic
at
io
n
E
ngineeri
ng,
Un
i
ver
sit
i M
al
ay
sia
Per
li
s (U
niMAP
)
Ti
ngka
t 1,
Kam
pu
s Tetap
Pauh, Put
ra
02600 Ara
u, Pe
rlis, Ma
la
ys
ia
.
Em
a
il
: Lay
th.a.khali
l@gm
ai
l.
com
1.
INTROD
U
CTION
A
de
dicat
ed
m
ob
il
e
net
wor
k
is
a
netw
ork
of
m
ob
il
e
no
des
t
hat
com
m
un
ic
at
e
with
each
ot
her
with
the
help o
f wirel
es
s li
nk
s
w
it
ho
ut
any existi
ng
i
nfrastr
uctu
re
or
any cent
ral acc
ess point
or
ba
se stat
ion
[
1].
A
de
dicat
ed
ne
twork
is
a
wi
re
le
ss
netw
ork
in
wh
ic
h
the
co
nt
ract
colla
borat
es
to
form
a
se
t
of
no
des
that
do
not
re
quire
a
ny
c
ommun
ic
at
io
n
a
nd
work
in
fr
ast
ruct
ur
e.
It
is
a
sc
al
able,
sel
f
-
co
nfi
g
ur
in
g
netw
ork
,
a
nd
the
qual
it
y
of
the
m
ob
il
e
node
s
of
t
he
net
w
ork
nee
de
d
to
config
ur
e
it
qu
ic
kly
exp
a
nds
without
the
ne
ed
f
or
sp
eci
fic
or
s
pe
ci
al
infr
ast
r
uct
ur
e
.
MA
NETs
m
ay
lim
i
t
com
m
un
ic
at
ion
s
to
a
gro
up
s
o
that
each
node
m
us
t
perform
it
s
work
as
the
host
,
and
the
r
oute
r
al
so
wor
ks
a
nd
pac
kets
a
re
se
nt
to
nodes
tha
t
are
not
within
t
he
nex
t
direct
tra
ns
m
issi
on
area
in
a
m
ulti
-
ho
p
directi
on.
Eac
h
node
has
f
ull
pa
rtic
ipati
on
in
th
e
c
us
tom
ro
ut
i
ng
protoc
ol
with regar
d
t
o
ef
fecti
ve
sessi
ons
between n
odes
[2,
5
]
.
These
netw
ork
s
are
par
ti
c
ular
ly
su
it
able
in
c
ases
w
here
in
frast
ru
ct
ure
is
un
avail
able
or
im
po
s
sible
t
o
pu
t
into
place
because
it
is
to
o
e
xpen
sive,
w
eak
or
destr
uct
ive
[3
]
.T
he
qu
al
it
y
of
c
omm
un
ic
at
io
n
can
change
qu
ic
kly
ov
e
r
ti
m
e
or
eve
n
dis
app
ea
r
al
to
ge
th
er.
T
he
no
des
c
an
lo
ok,
disa
ppear
an
d
c
om
e
back
agai
n
ove
r
tim
e,
and
al
l
t
he
ti
m
e
net
w
ork
c
on
necti
ons
m
us
t
work
betwee
n
the
nodes
th
at
m
ake
it
up.
A
s
on
e
ca
n
no
doubt
i
m
agine,
the
se
ty
pes
of
dynam
ic
networ
ks
a
re
wea
k
and
un
im
ple
m
ented.
Be
ca
use
cust
om
net
works
dynam
ic
al
l
y
change
t
he
t
opol
og
y
of
a
netw
ork
a
nd
it
op
e
ns
up
i
n
the
m
idd
le
,
t
her
e
is
no
cl
ear
li
ne
of
de
fen
se
,
and
ther
e
is
a
l
ack
of
c
omm
u
nicat
ion
a
nd
ce
ntrali
zed
m
on
it
or
i
ng.
T
her
e
for
e,
MA
NET
al
s
o
in
her
it
s
t
he
s
ecur
it
y
threats
they
fa
ce
in
the
case
of
wire
d
an
d
wireless
netw
orks
,
an
d
offer
s
secur
it
y
at
ta
cks
that
are
uniqu
e
t
o
them
sel
ves
[4
]
.
T
he
sim
ulatio
n
an
d
stu
dy
of
j
el
ly
fish
at
ta
c
ks
bec
om
es
necessary
in
ord
e
r
to
pro
vide
de
fen
se
against t
hese t
ypes
of
att
ack
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
1
3
, N
o.
2
,
Fe
bru
ary
201
9
:
683
–
688
684
2.
JELL
YFISH
ATTA
CK
AOD
V
is
a
r
outi
ng
pr
oto
c
ol
that
is
aff
ect
e
d
by
j
el
ly
fish
at
t
acks.
[6
]
A
j
el
l
yfi
sh
at
ta
ck
is
a
kind
of
non
-
act
ive
at
ta
ck
that
is
dif
ficult
to
ide
ntify
in
the
li
ght
of
the
fact
that
the
a
ggress
or
does
not
ig
nore
a
ny
of
t
he
Pr
ot
oc
ol'
s
ru
le
s
.
It
is
a
kind
of
den
ia
l
of
ser
vic
e
at
ta
ck.
A
n
at
t
acker
cha
nges
t
he
orde
r
of
dat
a
pac
kets
or
re
du
ce
s
traff
ic
to
a
m
i
nim
u
m
or
zero
by
dro
ppin
g
da
ta
pack
et
s.
T
he
fir
st
ste
p
to
be
ta
ke
n
by
a
j
el
ly
fish
stri
ke
r
is
to
acce
ss
the
r
ou
t
ing
netw
ork
an
d
int
rod
uce
a
r
edirecti
on
pac
kag
e
.
It
is
li
ke
a
Bl
ack
hole
at
ta
ck
in
the
sens
e
that
in
an
at
ta
ck
on
the
Bl
ackhole
node
the
at
ta
ck
er
dro
ps
the d
a
ta
pack
et
s,
bu
t i
n
a
j
el
ly
fish
at
ta
ck
there
is
a d
el
ay
befor
e
t
he
pa
c
kets
are
se
nt
a
nd
t
hen
receiv
ed
f
ro
m
the
pa
ckets
in
the
gr
i
d.
T
his
only
m
eans
that
it
is
no
t
the
sam
e
as
a
Bl
ac
khole
at
ta
ck.
T
he
je
ll
yfi
sh
at
ta
ck
is
ai
m
ed
at
c
losed
l
oop
flo
ws
becau
se
the
se
flo
ws
inte
ra
ct
with
netw
ork
c
onditi
on
s
su
c
h
as
pa
cket
loss
an
d
pac
ket
delay
[7
]
.
A
j
el
ly
fish
at
ta
ck
is
sp
ec
ific
al
ly
da
m
aging
to
trans
fer
c
ontro
l
protoc
ol
traf
fic
i
n
that
th
ose
co
op
e
rati
ve
nodes
ca
n
ha
rd
ly
disti
ng
uish
these
at
ta
c
ks
from
netw
ork
c
onge
sti
on
.
Fig
ur
e
1
t
hr
ee
ways
of J
el
ly
fish
at
ta
ck [7].
Figure
1.
Th
re
e w
ay
s
of
j
el
ly
fish
at
ta
ck
2.1
.
Jellyfish
delay v
aria
nc
e att
ack
I
n
this
ty
pe
of
at
ta
ck,
the
re
quest
do
e
s
no
t
c
hange
the
pac
ke
ts.
Ra
t
her
,
th
e
pac
kets
a
re
de
la
ye
d
in
a
rand
om
way.
The
m
al
iciou
s
node
has
to
se
cur
e
t
he
r
outi
ng
pat
hs
a
nd
t
he
n
al
l
the
pac
ke
ts
are
recei
ve
d
by
a
pr
e
vious
r
e
dire
ct
ion
delay
. W
e h
a
ve
te
ste
d
t
hi
s ty
pe
of att
a
ck
i
n order
to
s
tud
y t
he
e
ff
ect
i
ven
e
ss a
nd im
pact
of
the
netw
ork
a
nd
t
he
perf
or
m
ance
of
the
prot
oco
l
un
der
di
fferent
ci
rcu
m
sta
nces,
a
nd
un
de
r
th
e
in
flue
nc
e
of
a
n
at
ta
cker
.
W
e
have test
ed
this
in the ca
se
of
m
or
e than
one
at
ta
cker
2.2
.
Jell
yf
ish
peri
od
ic
dr
oppi
ng
att
ack
This
is
m
os
t
co
m
m
on
ly
fo
un
d
in
the
relay
no
des.
Be
ca
us
e
of
co
ngest
io
n,
it
is
necessa
ry
f
or
t
he
node
to dr
op p
ac
kets
p
e
rio
dical
ly
an
d t
he
t
ran
sit
io
n of t
he
c
ontrol
pro
t
oco
l
will
dec
rease to
zer
o.
2.3
.
Jellyfish
reorder
att
a
ck
In su
c
h
a
n
a
ssa
ult, the m
al
ic
iou
s
node
forwar
d packets
in
t
he
r
a
ndom
o
rd
e
r
f
r
om
the queue, r
es
ulti
ng
In
ze
r
o
go
od
put,
in
it
s
place
of
i
niti
al
trans
m
ittal
(f
irst
in
f
irst
out).
Packe
ts
can
be
placed
in
t
he
a
rb
it
ra
r
y
buff
e
r
instea
d
of
the
f
irst
in
fi
rst
out
buff
e
r.
It
is
rearr
a
ng
e
d
by
the
Jel
ly
fish
node
and
pac
kets
ar
e
then
se
nt
fro
m
the
buff
e
r.
I
n
te
rm
s
of
t
he
destin
at
ion
,
after
the
pac
kets
do
not
arr
i
ve
in
acc
orda
nce
with
t
he
real
dem
and
,
t
he
ackno
wled
gem
ent freq
ue
ncy i
s sen
t t
o
the
s
ource
.
Ther
e
f
o
re,
i
f
you
rep
eat
the
receipt
of
th
r
ee
ackno
wled
gem
ents
on
th
e
source
si
de,
the
pac
ket
retransm
issi
on
s
are
the
be
ginnin
g
of
a
de
nial
of
wait
ing
a
s
far
as
the
retra
ns
m
issi
on
tim
e
ou
t
is
c
oncer
ne
d.
E
ve
n
if
the
pac
ket
re
aches
it
s
destin
at
ion
,
the
sou
rc
e
sti
ll
correct
ly
acce
pts
t
hat
the
pac
ket
has
bee
n
l
os
t
a
nd
m
ay
kee
p
retransm
i
tt
ing
the
pack
et
.
A Je
ll
yfi
sh
r
e
order
at
ta
ck
is sh
ow
n
in
Fig
ure
2.
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
Beh
avior
al an
d perfor
m
an
ce
je
ll
yf
ish a
tt
ack
(
La
yt
h
A
. K
ha
l
il
Al Du
l
aimi
)
685
Figure
2.
Jel
ly
fish
re
orde
r
at
ta
ck
3.
INV
E
STIG
A
TING T
HE J
EL
LYFISH
AT
TACK
The
e
xec
ution
of
a
Jel
ly
fish
at
ta
ck
is
trie
d
on
an
a
d
hoc
netw
ork.
To
be
s
ur
e
the
ex
ecuti
on
i
s
ef
fecti
vely
work
i
ng,
we
use
d
the
OM
NeT
++
in
that
it
is
one
of
t
he
m
os
t
com
m
on
ly
us
ed
sim
ulati
on
to
ols
in
t
he
netw
orkin
g
fiel
d.
Additi
onal
ly
,
the
NE
TA
fr
am
ewo
r
k
i
s
base
d
on
the
sam
e
idea
as
OM
Ne
T++
[
9].
T
o
te
st
the
i
m
ple
m
entat
io
n,
w
e
us
e
d
th
re
e
si
m
ulati
on
s.
In
th
e
first
sit
ua
ti
on
, w
e
di
d
not
us
e
a
ny
Jel
ly
fish
A
O
DV
nod
e
.
I
n
the
sec
ond
sit
uation,
we
a
dded
a
si
ng
le
Jel
ly
fish
at
ta
ck
i
n
t
he
A
ODV
node
to
the
si
m
ula
ti
on
.
I
n
t
he
t
hir
d
sit
uation,
a
c
oope
rati
ve
Jel
ly
fish
at
ta
ck
in
t
he
A
ODV
no
de
wa
s
a
dded
t
o
t
he
sim
ulatio
n.
At
that
point,
we
com
par
ed
t
he ou
tc
om
es o
f
th
e sim
ulatio
ns
.
3.1.
Sim
ulat
i
on
Par
ame
ter
s a
n
d
Me
as
ur
ed M
e
trics
We
util
ise
d
the
foll
ow
i
ng m
etr
ic
s in o
ur
re
view:
1)
Packet
delive
r
y
rati
o
(PDR):
cal
culat
ed
by
di
vid
in
g
t
he
nu
m
ber
of
pac
kets
rec
ei
ve
d
by
t
he
destinat
io
n
by
the num
ber
of
pack
et
s
origi
na
te
d
f
ro
m
the s
ource;
2)
Av
e
ra
ge
th
rou
ghput:
the a
verage
of su
cce
ssf
ul m
essage d
el
ivery
ov
e
r
a c
om
m
un
ic
at
ion
c
hannel;
an
d
3)
Av
e
ra
ge
e
nd
-
to
-
e
nd
delay
:
de
fine
d
a
s
t
he
ti
m
e
ta
ken
for
a
data
pac
ket
t
o
be
tran
sm
i
tt
ed
acro
s
s
M
ANE
Ts
from
so
ur
ce
to desti
natio
n [10
]
.
To
m
easur
e
t
he
outc
om
es
fro
m
the
sim
ulatio
ns,
t
he
c
ommon
pa
ram
et
ers
to
al
l
sit
uatio
ns
(sce
nar
i
os
)
wer
e
def
i
ned.
I
t
ge
ner
at
e
d
10
netw
orks
wit
h
20,
40,
60,
80
and
100
nodes
.
T
he
sim
ulati
o
n
area
was
lim
i
te
d
t
o
a
1000
m
x
10
00
m
fo
ur
-
si
de
d,
with
eac
h
node
us
i
ng
a
c
om
m
un
ic
at
ion
r
ang
e
of
250
m
.
The
sim
ulati
on
ti
m
e
was
set
to
600
s.
T
he
res
ults
wer
e
de
rive
d
f
ro
m
50
sim
ulatio
n
r
un
s
.
AOD
V
a
nd
802.1
1g
we
re
sel
ect
ed
as
the
routin
g
a
nd
m
edium
acce
ss
con
t
ro
l
(MAC
)
la
ye
r
protoc
ol
s
res
pecti
vely
,
an
d
the
RTS/
CTS
m
echan
is
m
wa
s
us
e
d
to
sen
d
pa
ckets.
T
he
num
ber
of
at
ta
cke
rs
va
ried
from
one
to
th
ree.
T
he
at
ta
cks
we
re
pe
rfor
m
ed
duri
ng
the
ent
ire
sim
ulatio
n
ti
m
e
and
the
co
rr
e
spo
nd
i
ng
at
ta
ck
rate
was
set
to
100%,
where
the
at
ta
ck
rate
w
as
the
pro
bab
il
it
y
of
an
at
ta
cker
no
de
trig
ge
rin
g
it
s
at
ta
ck.
T
her
e
was
a
c
onsta
nt
bitrat
e
(CBR
)
co
nnect
ion
of
f
our
pack
et
s
,
w
her
e
the
pack
et
pa
yl
oad
siz
e
was
512
byte
s.
T
he
m
ov
em
ents
ra
ndom
ly
sta
rted
bet
wee
n
0.5
a
nd
1.5
s
an
d
they
e
nd
ed
betwee
n
290
a
nd
295
s.
T
he
m
ini
m
u
m
s
peed
was
set
to
1
m
/s
and
th
e
m
axi
m
u
m
w
as
set
to
betwee
n 5 to 2
0
m
/s, w
it
h
a
pa
us
e ti
m
e o
f 15 s.
[8
]
.
4.
APP
RA
I
SA
L
OF
RESU
LT
S
Thr
ee
div
er
se
si
m
ul
at
ion
s
we
re
te
ste
d
j
el
ly
fi
sh
delay
va
rian
ce
at
ta
ck.
I
n
th
e
first,
eac
h
node
wor
ke
d
in
c
ollaborat
io
n
with
each
oth
er
to
gua
rd
t
he
netw
ork
in
com
m
un
ic
at
ion
.
The
sec
ond
sim
ulati
on
ha
d
one
m
al
ic
iou
s
node
w
hile
the
thi
r
d
had
fi
ve
m
alici
ou
s
nodes
th
at
transm
it
te
d
t
he
Jel
ly
fish
at
ta
ck.In
this
a
na
ly
sis,
the r
es
ults
of
t
hese t
hr
ee
sim
ulati
on
s
are c
om
par
ed
to
unde
rstan
d
the
n
et
work a
nd no
de
b
e
hav
i
or
s
.
To
sta
rt
with,
pa
cket
delive
ry
rati
o
(PDR)
.
A
s
Table
1
sho
w
s,
the
P
DR
is
al
m
os
t
1
befor
e
t
he
Jel
ly
fish
at
ta
ck.
T
his
m
eans
that
al
m
os
t
al
l
the
pac
ke
ts
sent
by
the
s
end
e
r
no
de
a
re
recei
ved
by
th
e
recei
ver
node
,
bu
t
for
a
netw
ork
with
a
Jel
ly
fis
h
node,
t
he
P
D
R
decr
ease
s
du
e
to
inc
rease
d
retransm
issi
on
tim
eou
t
le
d
by
delay
In
tr
oduce
d
by
at
ta
cker
.
T
his
m
eans
that
al
m
os
t
al
l
of
t
he
pac
kets
sent
by
the
se
nder
node
are
delay
by
the
j
el
ly
fish nodes
.
Table
1.
Packe
t Deli
ver
y R
at
i
o
C
om
par
ison
Nu
m
b
e
r
o
f
no
d
es
No
Jelly
f
ish
attack
On
e Jelly
f
ish
attac
k
Fiv
e
Jell
y
f
ish
attack
20
9
3
.67
7
6
2
9
0
.03
2
7
5
8
1
.63
0
8
7
40
9
2
.71
2
8
2
9
1
.34
9
4
5
8
9
.62
6
6
8
60
9
7
.36
5
7
5
9
8
.04
6
6
5
9
8
.00
6
4
9
80
8
5
.58
2
5
4
8
7
.79
6
9
8
7
6
.02
2
1
3
100
6
9
.67
8
0
9
7
4
.54
4
9
6
8
1
.28
2
8
4
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
1
3
, N
o.
2
,
Fe
bru
ary
201
9
:
683
–
688
686
Figure
3
is
dra
wn
from
the
da
ta
con
ta
ine
d
i
n
Ta
ble
2.
Thi
s
com
par
es
the
PD
R
befor
e
t
he
Jel
ly
fish
at
ta
ck
an
d
afte
r
one
a
nd
five
J
el
ly
fish
at
ta
cks,
we
no
ti
ce
nods
be
hav
i
or
wh
e
n
the
nu
m
ber
of
no
des
bec
om
es
60
because
s
par
s
e
netw
ork
(is
a
netw
ork
with
l
ess
num
ber
of
li
nk
s
t
han
the
m
axi
m
u
m
po
ss
ible
num
ber
of
li
nk
s
within t
he
sam
e n
et
w
ork
)
:
Fi
gure
3
PD
R
.
Figure
3.
P
DR
Table
2
s
hows
the
E
nd
-
to
-
E
nd
Del
ay
:
En
d
-
to
-
E
nd
delay
is
the
a
ver
a
ge
ti
m
e
of
t
he
data
pack
et
to
be
su
ccess
fu
ll
y
tr
ansm
itted
acr
oss
a
MA
NET
f
ro
m
source
to
destinat
io
n
[
11]
.
We
can
see
an
e
nd
-
to
-
e
nd
dela
y
increase
in
sit
ua
ti
on
s
in
volvin
g
a
Jel
ly
fish
at
ta
ck.
T
his
is
the
resu
lt
of
delay
pack
et
s
w
hen
receivin
g
packet
s
by
Jel
ly
fish
no
des
,
the
reas
on for t
he dela
y i
ncr
e
ase is d
ue
t
o
th
e att
ack
node b
ehav
i
or
.
Table
2.
E
nd
-
to
-
E
nd
Delay
Com
par
ison
Nu
m
b
e
r
o
f
no
d
es
No
Jelly
f
ish
attack
On
e Jelly
f
ish
attac
k
Fiv
e Jell
y
f
ish
attack
20
0
.00
3
6
0
1
0
.02
2
1
2
5
0
.00
4
5
5
8
40
0
.00
0
6
2
7
0
.00
2
4
4
6
0
.00
1
0
6
7
60
0
.00
5
7
2
3
0
.00
9
2
9
0
.00
6
8
9
4
80
0
.01
0
5
4
8
0
.01
4
6
8
2
0
.01
2
0
3
8
100
0
.01
4
7
6
8
0
.01
7
2
4
3
0
.01
6
6
1
5
Figure
4
is
dr
a
wn
from
Table
2’
s
in
f
or
m
at
ion
.
It
com
par
e
s
the
en
d
-
to
-
e
nd
delay
be
fore
a
Jel
ly
fish
at
ta
ck
an
d
a
fter
on
e
an
d five
Jel
ly
fish
at
ta
ck
s:
Figure
4
.
Dela
y
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
Beh
avior
al an
d perfor
m
an
ce
je
ll
yf
ish a
tt
ack
(
La
yt
h
A
. K
ha
l
il
Al Du
l
aimi
)
687
Table
3
s
hows
the
th
rou
ghput
for
div
e
rse
sit
ua
ti
on
s
(
T
he
r
atio
of
the
total
amo
unt
of
data
tha
t
reac
he
s
a
recei
ver
fr
om
a
se
nder
t
o
the
ti
me
it
t
akes
for
the
r
ecei
ver
to
get
the
la
st
pa
ck
et
is
ref
erre
d
to
as
throu
ghput
)[1
1]
. N
ext,
on
e
nt
erin
g
Jel
ly
fish nodes int
o
the
netw
ork
r
outi
ng r
e
qu
e
st, the t
hro
ughput d
e
cr
eases.
This
im
plies
that
after
t
he
Jel
ly
fish
at
ta
ck,
th
e
num
ber
of
da
ta
pack
et
s
that
are
c
omm
un
icated
is
le
ss
t
ha
n
th
e
nu
m
ber
of co
nt
ro
l
pac
kets g
e
n
erated.
Table
3.
T
hro
ughp
ut Com
par
ison
Nu
m
b
e
r
o
f
no
d
es
No
Jelly
f
ish
attack
On
e Jelly
f
ish
attac
k
Fiv
e Jell
y
f
ish
attack
20
1
3
4
3
5
7
1
2
9
1
3
1
.2
1
1
7
0
8
0
.1
40
3
0
8
7
7
8
.3
3
0
4
8
7
1
.4
2
9
9
9
3
8
.5
60
4
1
8
7
8
1
.9
4
2
1
7
1
0
.5
4
2
1
5
3
7
.1
80
4
9
0
6
4
4
.1
4
0
9
6
6
4
.9
4
5
1
5
8
5
.4
100
4
9
9
4
3
8
.3
3
3
6
0
3
9
.3
4
9
0
7
7
5
.3
Figure
5
is
dra
wn
from
the
da
ta
in
the
a
bove
ta
ble.
T
his
al
lows
us
to
com
par
e
th
r
oughpu
t
be
fore
the
Jel
ly
fish
at
ta
ck
and a
fter
on
e
and fi
ve
Jel
ly
fish
at
ta
cks:
Figure
5.
Th
r
ough
pu
t
5.
CONCL
US
I
O
N
In
m
ob
il
e
ad
hoc
net
wor
ks
(
MANET
),
node
s
de
pend
on
e
ach
ot
her
t
o
ke
ep
the
netw
ork
associat
ed.
I
n
this
m
ann
er,
unli
ke
c
us
t
om
ar
y
wireless
ar
r
ang
em
ents,
s
uc
h
netw
orks
don'
t
re
qu
i
re
a
ny
pr
e
vious
(s
e
tt
le
d)
fou
nd
at
io
n.
T
his
li
m
it
s
their
cost
a
nd
se
nd
i
ng
tim
e.
H
enc
e,
sec
ur
it
y
in
a
m
ob
il
e
ad
-
ho
c
netw
ork
is
the
m
os
t
i
m
per
at
ive wor
ry f
or the fu
nda
m
ental
u
sefu
l
ness of
the
net
work. MA
NET
s r
eg
ularly
exp
erience the il
l effects
of
sec
ur
it
y
ass
aults
in
view
of
s
uc
h
el
em
e
nts
as
it
s
op
e
n
m
edium
,
chan
gi
ng
it
s
to
po
l
og
y
dynam
i
cally
,
the
abse
nce
of
fo
c
al
ob
se
r
ving
a
nd
adm
inist
rati
on
,
hel
pful
al
gorithm
s,
and
no
reas
onable
def
e
ns
e
m
echan
ism
.
A jell
yfi
sh
att
a
ck
is a
v
e
ry im
portant ty
pe
of
at
ta
ck
that
has becom
e o
f
inte
rest to m
any re
searche
rs.
Hen
ce
,
i
n
t
his
work,
we
ha
ve
analy
zed
the
e
ff
ect
of
Jel
ly
fis
h
at
ta
ck
s
i
n
a
n
A
O
DV
ne
tw
ork
.
F
or
t
his
reason,
we
ha
ve
ap
plied
a
n
A
ODV
proto
col
that
be
ha
ves
as
a
Jel
ly
fish
at
ta
ck
us
in
g
a
n
O
MNET++
sim
ulato
r
(4.2.2
).
W
e
sim
ula
te
d
3
sce
nar
i
os
on
20,
40,
60,
80
an
d
100
no
des.
T
he
sim
ulati
on
was
done
us
i
ng
UDP
pack
et
s
.
I
n
ea
ch
sit
uatio
n,
w
e
app
li
e
d
net
w
ork
perform
ance
befor
e
an
d
after
one
Jel
ly
fish
at
ta
ck
a
nd
five
Jel
ly
fish
at
ta
cks.
W
e
in
vestig
at
ed
the
ef
fects
of
t
he
Jel
ly
fish
at
ta
cks
on
ne
twork
perf
orm
ance
j
el
ly
fish
at
ta
c
k
aff
ect
s
a
dvers
el
y
throu
ghpu
t
of
net
wor
k
be
cause
of
co
ngest
io
n
ca
us
ed
due
to
retra
nsm
issi
on
s
a
nd
PDR
decr
ease
s due to
increa
sed ret
ran
sm
issi
on
ti
m
eou
t l
ed
by del
ay
intro
duce
d
by att
acker
. It
incr
eases en
d
-
to
-
e
nd
delay
of
the
ne
twork
.
The
e
ffec
t
is
m
or
e
dev
ast
at
ing
as
num
ber
of
j
el
ly
f
ish
no
de
increa
ses.
The
perf
orm
anc
e
degra
dation de
pends
on t
he d
el
ay
introdu
ce
d by the
j
el
ly
fish nodes
.
REFERE
NCE
S
[1]
N.Hoang
La
n
,
a
nd
U.
Tra
ng
Ngu
y
en
.
A
stud
y
of
d
iffe
ren
t
t
y
pes
of
at
t
ac
ks
on
m
ult
icast
in
m
obil
e
ad
hoc
net
works
.
A
d
Hoc
Networks
6
(1):
32
-
46
(2008
).
.
[2]
H.
Lu
.
W
ireless
Ad
-
hoc
Network
s.
W
ire
l
ess pe
rsonal
Com
m
unic
ation
journ
al
4
(20
04).
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
1
3
, N
o.
2
,
Fe
bru
ary
201
9
:
683
–
688
688
[3]
M.
Sergio
,
T
.
J
.
Giuli
,
K.
Lai,
an
d
M.
B
ake
r
.
Mit
i
gat
ing
routi
ng
m
i
sbeha
vior
in
m
ob
il
e
ad
hoc
net
wor
ks.
In
Proc
ee
din
gs
of
the 6t
h
annu
al i
nt
ern
ational c
o
nfe
re
nc
e
on
Mobile c
om
puti
ng
a
nd
net
working
,
p
p.
255
-
265
(200
0).
[4]
J.
Apurva
,
and
A.
Shrotri
y
a
.
In
vesti
gating
the
e
ffe
ct
s
of
bl
ac
k
h
ole
atta
ck
in
MA
NET
under
sh
adowing
m
odel
with
diffe
ren
t
tr
aff
i
c
condi
ti
ons
.
In
C
om
pute
r,
Com
m
unic
a
ti
on
and
C
ontrol
(IC4)
,
20
15
Int
ern
at
ion
al
Confer
ence
on
,
pp.
1
-
6(
2015).
[5]
P.
Nidhi,
R.
Sin
ha,
and
K.
Maur
y
a
.
Sim
ulation
stud
y
of
B
la
ck
h
ole
and
Je
lly
fish
attac
k
on
MA
NET
using
NS
3.
I
n
Engi
ne
eri
ng
(N
UiCON
E),
2011
Nirm
a
Univer
sit
y
Inte
rn
at
ion
al C
onfe
ren
c
e
on
IE
EE
,
pp.
1
-
5
(201
1).
[6]
G.
Sakshi
,
and
S
.
Ch
and.
Enh
anced
AO
DV
proto
col
fo
r
d
efe
nc
e
a
gai
nst
Je
lly
Fis
h
Atta
ck
on
MA
NETs.
In
Advan
ces
in
Com
puti
ng,
C
om
m
unic
at
ions
and
Inform
at
i
cs
(ICACCI,
2014
I
nte
rna
ti
ona
l
Con
fer
ence ,
pp
.
227
9
-
2284
(2014).
[7]
Sachde
va
,
Saks
hi,
and
Parne
e
t
Kaur.
"D
etec
t
i
on
an
d
an
aly
sis
of
Je
lly
fish
a
t
ta
ck
in
MA
NETs.
"
In
Inv
ent
i
v
e
Computati
on
Te
chnol
ogi
es
(
ICICT)
,
Inte
rnation
al
Conf
ere
nce o
n
,
vol
.
2
,
pp
.
1
-
5
.
IE
EE,
2016
.
[8]
S.
Casado,
Le
ov
i
gil
do,
R
.
A.
Rodrígue
z
-
Góm
ez,
R.
Magá
n
-
C
arr
ió
n,
and
G
.
Mac
iá
-
Ferná
ndez.
"N
E
TA:
evalua
ti
ng
t
h
e
eff
ects
of
NETw
ork
at
t
ac
ks.
MA
NETs
as
a
c
ase
stud
y
.
"
In
Advan
ce
s
in
Secu
rity
o
f
Inform
at
ion
an
d
Com
m
unic
at
io
n
Networks,
Sprin
ger
Ber
li
n
Heidelbe
rg, 2013
pp
.
1
-
10.
[9]
T.
Nava
m
an
i,
a
nd
P.
Yogesh.
“
Secur
e
Eff
i
ci
e
nt
Routi
ng
ag
ains
t
Packe
t
Dro
pping
Attacks
i
n
W
i
rel
ess
Mesh
Networks”.
In
P
roc
ee
d
ings
of
th
e
3rd
Int
ern
a
ti
o
nal
Confer
en
ce
on
Frontiers
of
Inte
lligen
t
Sprin
ger
In
te
rn
at
ion
al
Publishing,
2015.
pp.
673
-
686
.
[10]
K.
Sim
ran
pre
et
,
R.
Kaur
,
and
A.
K.
Verm
a
.
J
el
l
y
fish
a
tt
a
ck
i
n
MA
NETs:
A
rev
ie
w.
In
El
e
ctrical,
Com
pu
te
r
and
Com
m
unic
at
ion Te
chno
logi
es
(I
CECCT
), 2015
I
EE
E
Inte
rn
ation
al
Conf
ere
n
ce o
n,
pp
.
1
-
5
(2015)
.
[11]
B,
Ta
runpr
eet,
a
nd
A.
K.
Verm
a.
"P
erf
orm
anc
e
E
val
ua
ti
on
of
AO
DV
under
Bla
ck
hole
Att
ac
k
.
"
Int
ern
ational
Journ
al
of
Com
pute
r
Ne
t
work a
nd
Inform
at
ion
Secur
i
t
y
5
,
no.
:
12,
pp35
(2
013).
Evaluation Warning : The document was created with Spire.PDF for Python.