Internati
o
nal
Journal of Ele
c
trical
and Computer
Engineering
(IJE
CE)
V
o
l.
6, N
o
. 5
,
O
c
tob
e
r
201
6, p
p
. 2
322
~233
0
I
S
SN
: 208
8-8
7
0
8
,
D
O
I
:
10.115
91
/ij
ece.v6
i
5.1
162
4
2
322
Jo
urn
a
l
h
o
me
pa
ge
: h
ttp
://iaesjo
u
r
na
l.com/
o
n
lin
e/ind
e
x.ph
p
/
IJECE
Node Cooperation to Avoid Earl
y Congestion Detection Based
on Cross
-
Layer f
o
r Wi
rel
ess Ad Hoc Networks
Abdalraz
ak
T
a
req Rahem
,
Mahamod
Is
mail,
Nor
Fad
z
ilah Abdulla
h, Mohamme
d
Balfaqih
Departm
e
nt o
f
E
l
ec
tric
al
, E
l
e
c
tro
n
ics
and
S
y
s
t
em
s
Engine
ering
,
F
acul
t
y
of
Engin
e
ering and
Bu
ilt Environment,
Universiti
Keba
ngsaan Mal
a
y
s
ia
(UKM), Mala
ys
ia
Article Info
A
B
STRAC
T
Article histo
r
y:
Received
May 16, 2016
Rev
i
sed
Ju
l 12
,
20
16
Accepte
d
J
u
l 29, 2016
The resen
t
application of wireless
ad hoc netw
orks (WANET)
demands a
high and
reliab
le data load
.
The simultaneous
tr
ansfer of
larg
e
amounts of
data d
i
ffer
e
nt n
earb
y
s
ources
t
o
near
b
y
destin
ations in
a massive network
under these cir
c
um
stances result
s in th
e possibilit
y
of network congestion.
Congestion is an extrem
el
y
un
wanted
condi
tio
n because it cr
eat
es extr
a
overhead
to
the
alre
ad
y de
epl
y
l
o
aded
environm
ent,
which u
ltim
ate
l
y l
eads to
res
ource exh
a
us
tion, and
can l
ead to
packet d
r
ops and retran
smission at
eith
er
th
e
MAC or
upper lay
e
rs. We
present a lig
htweight conges
tion
con
t
rol
and early
avo
i
dance congestion control
scheme, which can eff
ective contr
o
l
congestion while keeping overh
ead to a
minimum. This sc
heme
is based on
the Cross-la
ye
r
between
the MA
C and ne
tw
ork lay
e
rs lead to
ear
l
y
detection
of congestion
.
With the h
e
lp of
node c
oop
eratio
n the send
er nod
e is tr
iggered
to find an alter
n
ativ
e route b
a
sed on
TMT. Th
is m
echanism
controls th
e
network resourc
e
s rather
than
th
e dat
a
tr
affic
.
D
e
ta
iled p
e
rform
a
n
ce resul
t
s
show enhancement in
the
throu
ghput and
packet deliver
y
r
a
tio
,
as well as
a
reduction in
packet drop
. Gen
e
rall
y
,
network
perf
ormance in
creases.
Keyword:
Ad
h
o
c
wireles
s
Altern
ativ
e route
Co
ng
estion
Net
w
or
k t
o
p
o
l
ogy
R
out
i
n
g pr
ot
oc
ol
Copyright ©
201
6 Institut
e
o
f
Ad
vanced
Engin
eer
ing and S
c
i
e
nce.
All rights re
se
rve
d
.
Co
rresp
ond
i
ng
Autho
r
:
Ab
dal
r
aza
k T
a
req
R
a
hem
,
Depa
rt
m
e
nt
of
El
ect
ri
cal, Electronics and
Sy
s
t
em
s Engi
n
eeri
n
g
,
Facu
lty of En
gin
eering
and
B
u
ilt Env
i
ron
m
en
t,
Natio
n
a
l
Un
iv
ersity o
f
Malaysia (UKM)
Em
a
il: ab
d
t
areq
1@siswa.u
k
m
.
ed
u
.
my
1.
INTRODUCTION
A wi
rel
e
ss ad
hoc
net
w
or
k (
W
A
N
ET
) refe
rs t
o
a pa
rt
i
c
ul
ar wi
rel
e
ss net
w
o
r
k
.
Part
i
c
ul
ar net
w
or
ks
(ad
hoc
) aim to obtain speci
fic charact
eri
s
t
i
c
s, suc
h
as dy
n
a
m
i
c, i
ndepe
nd
ent, self-c
o
n
fi
g
u
ri
ng
, dece
ntra
lized,
and i
n
f
r
ast
r
uct
u
re
-l
ess. T
h
e r
out
i
n
g
pr
ot
oc
o
l
pl
ay
s an esse
nt
i
a
l
rol
e
i
n
i
m
pr
o
v
i
n
g t
h
e pe
rf
orm
a
nce of
wi
rel
e
s
s
net
w
or
ks
[1]
.
The m
a
i
n
go
al
of a
n
y
ro
ut
i
n
g
pr
ot
oc
ol
i
s
t
o
d
e
t
e
rm
i
n
e dy
na
m
i
cal
ly
t
h
e cor
r
ect
r
out
e
bet
w
een
a
sou
r
ce n
o
d
e an
d a dest
i
n
at
i
o
n
no
de [2]
.
I
n
t
h
e case of co
nt
r
o
l
m
e
ssages, f
o
r
w
ar
di
n
g
o
f
l
a
rge am
ount
s
of dat
a
from
one node to anot
her
w
ithout the cooperation of each node leads
to conge
stion. Congestion occ
u
rs in any
m
i
dway
fr
om
a so
urce
n
o
d
e
t
o
a
dest
i
n
at
i
o
n
no
de i
f
m
a
ssi
v
e dat
a
pac
k
et
s t
r
avel
.
C
o
nse
que
nt
l
y
, hi
gh
packet
lo
ss and
long d
e
lay ar
e enco
un
ter
e
d
.
Th
i
s
situ
atio
n
lead
s to
th
e
d
e
gr
ad
ation
o
f
n
e
tw
or
k
p
e
rf
or
man
ce.
N
e
two
r
k
congestio
n
can
b
e
ad
dr
essed
th
ro
ugh
eith
er
traf
f
i
c con
t
ro
l o
r
r
e
sou
r
ce co
n
t
r
o
l.
H
o
w
e
ver
,
th
e
situ
atio
n
wo
rsen
s if resou
r
ces are
i
n
creased
with
ou
t con
s
id
ering
th
e
co
ng
estion
type, traffic p
a
ttern, and
net
w
or
k
t
o
p
o
l
ogy
. As
s
h
ow
n
i
n
Fi
g
u
re 1, con
g
est
i
o
n occ
u
rs
i
n
R
o
ut
e 1
:
20
→
21
→
22
→
23
→
24
→
25
→
26
→
27
→
28
→
29. Becaus
e
m
o
st routing pr
otocols choos
e that path without consi
d
eri
ng the
network
t
o
p
o
l
o
gy
[
3
]
.
AO
D
V
,
DS
D
V
,
OLSR
, a
n
d
DSR
ha
ve
no
co
nge
st
i
o
n
co
nt
r
o
l
al
g
o
ri
t
h
m
s
. Ty
pi
cal
l
y
, r
e
duci
n
g
packet
l
o
ss i
n
vol
ves c
o
n
g
es
t
i
on c
ont
r
o
l
.
We
have
use
d
det
ect
i
o
n
m
e
t
h
o
d
s t
o
p
r
ee
m
p
t
congest
i
o
n.
O
u
r
pr
o
pose
d
r
o
ut
i
ng
pr
ot
oc
ol
co
nsi
d
e
r
s an
d sel
ect
s t
h
e opt
i
m
al
and m
o
st
effi
ci
ent
ro
ut
e b
y
readi
n
g t
h
e net
w
or
k
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
N
ode
C
o
ope
rat
i
on t
o
Av
oi
d
E
a
rl
y C
o
n
g
est
i
o
n
Det
ect
i
on
Ba
sed
o
n
C
r
oss .
.
.. (
A
b
d
a
l
r
az
ak
Tare
q R
a
hem)
2
323
to
po
log
y
as well as av
o
i
d
s
co
ng
estion
b
e
fo
re it o
c
c
u
rs
. The propose
d
mechan
ism
w
a
s analyzed using
a
math
e
m
atica
l
m
o
d
e
l an
d
ev
al
u
a
ted
u
s
i
n
g a
NS-3
sim
u
lato
r.
100
m
10
0
m
10
no
de
s
(L
e
n
g
t
h
)
4
node
s
(w
i
d
t
h
)
So
u
r
c
e
No
d
e
s
Sin
k
No
d
e
Fi
gu
re 1.
C
o
ng
est
i
on ro
ut
e
2.
RELATED STUDIES
High
data loa
d
s can ea
sily lead
to congestion because
of
the lim
i
ted node
s res
o
urces a
v
a
ilable in the
wireless ad
hoc network. Conge
stion
is a highly undesira
ble situation b
ecause it creates additional overhea
d
to the alrea
d
y
heavily loa
d
ed enviro
n
m
en
t, wh
ich
ev
en
t
u
ally lead
s to
re
source de
pletion.
To reduce
t
h
e delay
and
b
u
f
f
er
ov
erfl
o
w
pr
od
uc
ed by
n
e
t
w
or
k co
n
g
est
i
o
n and t
o
en
ha
nc
e per
f
o
r
m
a
nce, co
ngest
i
o
n
m
u
st
be
avoi
ded
bef
o
re
i
t
occurs. C
o
nge
st
i
on
det
ect
i
on i
n
wi
re
l
e
ss ad h
o
c net
w
or
ks ha
s bee
n
st
udi
ed
[4]
-
[
8
]
. An
earl
i
e
r st
udy
[
9
]
i
n
t
r
o
duce
d
a no
vel
cross
-
l
a
y
e
r hy
bri
d
m
e
t
r
i
c
based o
n
AO
D
V
pr
ot
oc
ol
fo
r ad h
o
c n
e
t
w
o
r
ks
base
d o
n
i
n
f
o
rm
ati
on o
b
t
a
i
n
ed f
r
o
m
wi
rel
e
ss
ch
an
n
e
l co
nd
itio
ns in
t
h
e
p
h
y
sical layer, lin
k qu
ality and
congestion
i
n
M
A
C
l
a
y
e
r,
a
n
d m
i
nim
u
m
ho
ps
i
n
net
w
o
r
k l
a
y
e
r.
Aft
e
r
che
c
ki
n
g
t
h
e
r
o
ut
e f
o
r
t
h
e e
x
i
s
t
e
n
ce
o
f
co
ng
estion
in
an
y in
term
ed
iate n
o
d
e
, th
e
new ro
u
t
e is in
i
tialized
. Go
lnaz Karb
asch
i [10
]
ad
dressed
a n
e
w
lin
k
-
qu
ality an
d
con
g
e
stio
n-aware m
e
tric fo
r m
u
lt
i-h
o
p
wi
reless rou
ting
.
Th
e au
t
h
or fou
n
d
th
at cro
ss
layer
bet
w
ee
n
ro
ut
i
n
g
pr
ot
oc
ol
a
n
d
M
A
C
l
a
y
e
r
i
s
hel
p
f
u
l
i
n
en
h
a
nci
n
g
ro
ut
i
n
g
i
n
t
e
rm
s of e
n
d
-
t
o
-en
d
del
a
y
an
d
th
ro
ugh
pu
t in
t
h
e jud
g
m
en
t of th
e m
i
n
i
m
u
m-hop
coun
t m
e
tric. Mo
reov
er, cro
s
s-layer
rou
tin
g
is i
n
tended
to
pl
ay
an essent
i
a
l
rol
e
i
n
im
pro
v
i
n
g t
h
e
per
f
o
r
m
a
nce of w
i
rel
e
ss net
w
or
ks. C
o
n
g
est
i
o
n
det
ect
i
on i
n
s
e
ns
or
net
w
or
ks
has
b
een st
udi
e
d
[
6
]
,
[
7
]
,
[
11]
,
[
1
2
]
.
A pri
o
r
st
udy
[
13]
prese
n
t
e
d a
t
o
p
o
l
o
gy
-a
w
a
re
re
sou
r
ce ad
ap
tation
(TARA), wh
ich is
an
ad
ap
tation
strateg
y
for allev
i
atin
g
co
ng
estio
n
.
Th
e m
a
i
n
id
ea of
T
A
RA is the capac
ity analys
i
s
model
,
w
h
i
c
h ca
n b
e
use
d
t
o
p
r
edi
c
t
t
h
e ca
paci
t
y
o
f
di
ffe
rent
t
o
p
o
l
ogi
es.
Thi
s
m
odel
i
s
f
o
rm
ul
at
ed u
s
i
n
g a
gra
p
h
-
c
o
l
o
ri
ng
p
r
obl
em
.
TAR
A
i
s
a
d
va
nt
age
o
u
s
beca
use i
t
i
s
di
st
ri
but
e
d
, e
n
er
gy
effi
ci
ent
,
a
n
d t
o
p
o
l
o
gy
awa
r
e
.
R
e
l
a
t
e
d w
o
r
k
s ha
ve
di
scuss
e
d c
o
n
g
e
st
i
on base
d o
n
t
h
e use
of t
h
e si
ze of que
ue
t
o
det
ect
t
h
e con
g
est
i
o
n. Si
t
u
at
i
ons exi
s
t
w
h
erei
n
the actual
que
ue size reac
hes
full buffe
r size
, eve
n
when
t
h
e avera
g
e
queue is below t
h
e
maxim
u
m
threshol
d
(MAXth).
In som
e
cases, pac
k
ets
will be
droppe
d because of
overflow [14],[15].
Th
e altern
ative p
a
th selectio
n sch
e
m
e
(DAlPaS), wh
i
c
h
is an
effectiv
e sch
e
m
e
th
at con
t
ro
ls
con
g
est
i
o
n
w
h
i
l
e
keepi
n
g
o
v
e
rhea
d t
o
a m
i
ni
m
u
m
,
has be
en re
p
r
ese
n
t
e
d
i
n
t
h
e
WSN
[
16]
.
The
o
p
era
t
i
on
of
th
is schem
e
is
b
a
sed
on
t
h
e
co
n
t
ro
l
of reso
urces i
n
st
ead of th
e co
n
t
ro
l
of th
e sen
d
i
ng
rate at th
e so
urce.
Congestion ca
n lead t
o
packet drops
and
re
transm
ission either at the
MAC
or u
ppe
r
l
a
y
e
rs, whi
c
h
a
r
e event
s
t
h
at
exh
a
ust
t
h
e al
ready
l
i
m
i
ted p
o
w
er
of
WS
Ns.
No
de
po
we
r ex
ha
ust
i
on ca
n res
u
l
t
i
n
r
out
i
n
g
hol
e
s
i
n
t
h
e
net
w
or
k,
whi
c
h can
ren
d
e
r
t
h
e net
w
or
k
u
n
abl
e
t
o
acc
o
m
pli
s
h i
t
s
ob
j
ect
i
v
e. Seve
ral
researc
h
w
o
r
k
s o
n
co
n
t
ro
lling
con
g
e
stio
n
i
n
WSNs
h
a
v
e
b
e
en
im
p
l
e
m
en
te
d
[16
]
,[17
]. The p
e
rform
a
n
ce o
f
DAlPaS
has b
een
evaluate
d a
g
ainst c
o
m
p
arable
schem
e
s and s
h
owe
d
prom
ising res
u
lts.
3.
PROP
OSE
D
SCHE
ME
Th
is section
illu
strates t
h
e math
em
at
ical
mo
d
e
l i
n
clud
ing ou
r
propo
sed
sch
e
m
e
th
rou
g
h
two
p
a
rts.
The fi
rst
part
expl
ai
n
s
h
o
w
t
o
preem
pt
i
v
el
y
det
ect
conge
st
i
on wi
t
h
t
h
e
hel
p
o
f
M
A
C
and
net
w
or
k
l
a
y
e
r
i
n
f
o
rm
at
i
on. Thi
s
i
s
represe
n
t
e
d as St
ep I an
d St
ep I
I
i
n
t
h
e
Fi
gure
2. T
h
e seco
nd pa
rt
sh
ows
ho
w t
o
di
s
c
ove
r
altern
ativ
e ro
u
t
es to
th
e d
e
stinatio
n
with
t
h
e h
e
lp
o
f
th
e TM
T. Th
is is rep
r
esen
ted
as Step III. First, it p
r
ed
ict
s
congestion
before it
occurs, a
n
d com
b
in
es t
w
o
di
f
f
ere
n
t
m
e
t
h
o
d
s t
o
di
sc
o
v
er t
h
e c
o
n
g
est
i
on.
Tw
o
pa
ra
m
e
t
e
rs,
nam
e
ly
, fai
l
u
re and q
u
e
u
e si
ze, from
t
h
e
M
A
C
and ne
t
w
ork layers, respectively are used. Sec
ond, it is
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
I
J
ECE
Vo
l. 6
,
N
o
. 5
,
O
c
tob
e
r
20
16
:
232
2
–
23
30
2
324
rep
r
ese
n
t
e
d
by
fi
ndi
n
g
t
h
e ne
w pat
h
base
d
o
n
t
h
e t
o
p
o
l
o
gy
of t
h
e
net
w
o
r
k. T
h
e sc
hem
e
i
s
an or
ga
ni
ze
d as i
n
th
e fo
llowing
step
s:
Step I
:
C
h
ec
ki
ng t
h
e
net
w
or
k
l
a
y
e
r t
o
di
sco
v
er t
h
e c
o
n
g
es
t
i
on be
fo
re i
t
occurs
(earl
y
co
nge
st
i
on
det
ect
i
o
n
)
,
and
se
ndi
ng
a
war
n
i
n
g m
e
ssage
onl
y
t
o
nei
g
hb
o
r
'
s
no
des t
o
co
ope
rat
e
a
n
d
avoi
d c
o
n
g
est
i
on
.
Step II
: Chec
king the MAC
layer param
e
ters to activate
an alternative
route find
er after
recei
ving
suc
h
war
n
i
n
g m
e
ssage.
Step III
:
Fi
ndi
ng
a
new
pat
h
base
d
on
t
h
e T
M
T.
Qu
e
u
e
‐
st
a
t
us
=
1
Ea
r
l
y
co
n
g
e
s
t
i
o
n
de
t
e
c
t
i
o
n
MA
C
la
y
e
r
Ne
t
w
o
r
k
la
y
e
r
Fa
i
l
u
r
e=
1
1
MIN
th
=
25
%
B
u
f
f
e
r
que
u
e
‐
si
ze
MA
X
th
=3*
MI
N
th
IF
In
st
a
n
t
a
n
e
o
u
s
qu
e
u
e
si
z
e
>=
MA
X
th
Se
nd
Wa
r
n
in
g
me
s
s
a
g
e
IF
Qu
e
u
e
‐
st
a
t
us
=
1
IF
Fa
i
l
u
r
e
=
1
1
Fa
i
l
u
r
e=
AC
K
Fa
i
l
ur
e
C
o
u
n
t
+
RT
S
Fa
i
l
u
r
e
C
o
u
nt
Ne
i
g
h
b
o
r
i
n
g
no
de
Se
n
d
e
r
no
de
ST
E
P
II
ST
E
P
II
I
ST
E
P
I
Di
s
c
o
v
e
r
al
t
e
rn
at
i
v
e
ro
u
t
e
Ne
t
w
o
r
k
to
p
o
l
o
gy
Se
ndi
ng
da
t
a
vi
a
Al
t
e
r
n
a
t
i
v
e
ro
u
t
e
TM
T
Fi
gu
re 2.
The
Pro
p
o
se
d
Sc
he
m
e
3.
1.
Early Congestion Detec
t
ion
The co
n
g
est
i
o
n i
n
wi
r
e
l
e
ss ad h
o
c net
w
o
r
k
s
pr
oba
bl
y
occ
u
rs i
n
ei
t
h
e
r
t
h
e M
A
C
or
net
w
o
r
k l
a
y
e
rs
,
and the
r
e a
r
e
differe
nt
ways t
o
detect it.
Howeve
r, the
cr
os
s t
w
o l
a
y
e
rs
i
n
t
e
rm
of chec
ki
ng
t
h
e
pa
ram
e
ters a
r
e
hel
p
ful
t
o
det
e
ct
any
earl
y
conge
st
i
on
bef
o
r
e
i
t
occurs. T
o
avoi
d s
u
c
h
co
nge
st
i
o
n
,
i
t
m
u
st
be pre
d
i
cat
ed fi
rs
t
and the
n
, t
h
ere shoul
d
be c
o
ope
r
ation to
find an a
lternative route. T
h
e cross
layers are: (a
)
Net
w
ork
co
ng
estion
and (b
)
M
A
C co
ngestio
n
as fo
llow
i
ng
.
Step I
:
→
(
a
)
N
e
two
r
k
congestio
n
:
A
ll nod
es h
a
v
e
a limited
b
u
f
f
e
r
qu
eu
e size. Th
er
efo
r
e, if
th
e
receive
d data size is greater than t
h
e actual
que
ue size of
the node, the
n
,
the data
will be droppe
d,
whi
c
h is
m
a
i
n
l
y
due t
o
t
h
e
occu
rre
nc
e o
f
ove
rl
oa
d
con
g
est
i
o
n.
Th
e net
w
o
r
k
l
a
y
e
r
has
param
e
t
e
rs t
o
m
easur
e t
h
e
current
que
u
e
size. E
quations 1 and
2 rep
r
esen
t th
e
p
a
rameters related to
n
e
two
r
k layer to
m
easu
r
e such
current
queue
s
i
ze.
Q
min
= 0.
25
×
B
u
f
f
e
r
queue-size
(1
)
Q
Threshold
= Q
max
=0.75×
B
u
ffer
queue-size
=
3× Q
min
(2
)
Whe
r
e
Q
min
de
not
es
t
o
t
h
e qu
art
e
r of
act
ual
que
ue si
ze
o
f
no
de w
h
i
c
h re
prese
n
t
s
t
h
e
m
i
nim
u
m
t
h
res
h
ol
d o
f
the actual que
u
e size.
And t
h
e
Q
Threshold
denotes t
o
the t
h
ree qua
r
ters
of act
ual queue
s
i
ze that repres
ents the
m
a
xim
u
m
t
h
reshol
d o
f
t
h
e a
c
t
u
al
que
ue si
z
e
.
Bu
ffer
queue-siz
e
re
prese
n
t
s
t
h
e act
ual
que
ue
si
ze of n
ode
. If t
h
e
current
queue
size of the
node
reac
hes
m
o
re or e
q
ual
Q
Threshold
th
en, a
warn
ing
messag
e
is sen
t
to
all
n
e
igh
boring
n
o
d
e
s Th
is ap
proach
is illu
strat
e
d
in STEP I
in
Fi
g
u
re
2
where it pred
icts
an
d d
e
tects if th
ere i
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
N
ode
C
o
ope
rat
i
on t
o
Av
oi
d
E
a
rl
y C
o
n
g
est
i
o
n
Det
ect
i
on
Ba
sed
o
n
C
r
oss .
.
.. (
A
b
d
a
l
r
az
ak
Tare
q R
a
hem)
2
325
an i
m
m
i
nent
con
g
est
i
o
n cr
os
si
ng
wi
t
h
M
A
C
param
e
t
e
rs too
(STE
P I
I
)
.
The q
u
e
u
e-
st
at
us wa
s ad
de
d t
o
al
l
no
des
.
B
y
defaul
t
,
que
ue
-st
a
t
u
s i
s
‘0’
.
Fo
r
i
n
st
ance, t
o
det
ect
an earl
y
congest
i
on
of a gi
ve
n l
i
n
k, t
h
e
in
stan
tan
e
ou
s q
u
e
u
e
is first ch
eck
ed
. In
stan
tan
e
ou
s qu
eu
e size refers to the curre
n
t que
ue size. If the curre
n
t
que
ue size
is e
qual
or greater than
three
qua
rters
of the act
ual que
u
e size,
th
en, a
warn
ing
m
e
ssag
e
is sen
t
t
o
t
h
e nei
g
h
b
o
ri
n
g
no
des
.
E
v
ery
n
ode
m
u
st
coo
p
erat
e
wi
t
h
eac
h
ot
he
r t
o
sen
d
t
h
i
s
m
e
ssage.
If a
n
y
no
de
rec
e
i
v
es
su
ch
warn
ing
messag
e
, th
en
t
h
e
q
u
e
u
e
-status fo
r t
h
at
p
a
rticu
l
ar
no
d
e
will
b
e
m
o
d
i
fied to
‘1’.
Step I
I
:
→
(
b
)
M
A
C
co
nge
st
i
on:
T
h
e m
obi
l
e
ad
hoc
net
w
or
k em
pl
oy
s a di
st
ri
b
u
t
e
d c
o
or
di
nat
i
o
n
function (DCF) for a m
e
dium
access. The DCF is a
ba
s
i
c channel acc
ess prot
ocol
for a
s
ync
h
ronous
dat
a
transm
ission i
n
the c
onte
n
tion pe
riod
base
d
on a Ca
rrier Sense M
u
ltiple Access
with Collision Avoidance
(CSMA/CA) m
echanism
.
T
h
ere are two
pa
ram
e
ters in IEEE 802.11 MAC layer
are
ACK
FailureC
ount
and
RTS
FailureCount
. Whe
r
e,
ACK
FailureCount
denot
es t
h
e n
u
m
b
er
of
Fai
l
u
re t
o
s
e
nd
D
A
T
A
an
d o
b
t
a
i
n
AC
K
,
an
d
RTS
FailureCount
denot
es t
h
e
num
ber
o
f
Fai
l
u
re t
o
obt
ai
n
f
r
ee m
e
di
a as s
h
ow
n
i
n
Fi
g
u
r
e
3
.
R
e
t
r
ansm
i
ssi
on o
ccur
s
only when a
n
ACK
or a CTS fram
e
is
not received from
the destination
node; thus, DATA
or RTS
fram
e
is
not
sent
.
Fi
gu
re
3.
M
A
C
fram
e
ret
r
a
n
s
m
i
ssi
on m
e
t
hod
Suc
h
si
t
u
at
i
o
n
l
eads t
o
d
r
op
pi
ng
o
f
pac
k
et
s
or
f
r
am
es i
n
t
h
e net
w
o
r
k
bec
a
use t
h
e
dat
a
i
s
sent
i
n
t
h
e
R
T
S/
C
T
S
pha
se t
o
o
b
t
a
i
n
t
h
e cha
n
nel
an
d
avoi
d t
h
e
hi
d
d
e
n/
ex
po
sed
n
o
d
e
pr
o
b
l
e
m
.
If
t
h
e s
e
n
d
er
do
es n
o
t
receive the CTS for a period ti
m
e
,
the sender node
retra
n
smits
an RTS until a CTS and holdi
ng channel are
o
b
t
ain
e
d
.
Th
erefore, th
e m
a
x
i
m
u
m retran
sm
i
ssio
n
fail
u
r
es
ov
er th
e
n
o
d
e
den
o
t
e th
e po
ssi
b
ilities th
at
th
e lin
k
i
s
possi
bl
y
an
d n
o
t
p
o
ssi
bl
y
con
g
est
i
o
n. E
v
ery
no
de
use
s
t
h
ese t
w
o va
ri
abl
e
s t
o
rep
r
esent
t
h
e
num
ber
of
retransm
issio
n
s
failu
re in
t
h
e MAC layer. Th
e defau
lt
max
i
m
u
m
v
a
lu
e o
f
RT
S
FailureCount
i
s
7 and t
h
e
defa
ul
t
m
a
xim
u
m
val
u
e of
ACK
Failure
Count
is 4 as sta
nda
rdization of MAC laye
r in IEEE
802.11 RTS/CTS [18],[19].
Eq
uat
i
on
3 re
f
l
ect
s t
h
e num
ber o
f
ret
r
a
n
sm
it
fai
l
u
re, w
h
e
r
e
F
Threshold
denot
es t
o
t
h
e m
a
xi
m
u
m
t
h
resh
ol
d
of
bot
h
DA
TA
an
d R
T
S
.
F
Threshold
= ACK
FailureC
ount
+ RTS
FailureCount
(3
)
The
n
u
m
b
er o
f
RTS
retra
n
s
m
issions re
fer
s
to t
h
e co
n
t
en
tio
n lev
e
l
o
f
th
e link
,
as
well as t
h
e
esti
m
a
t
i
o
n
of th
e lin
k
q
u
a
lity. If
F
Threshold
o
c
cu
rs
‘11’ ti
m
e
s, th
en
eith
er
con
g
est
i
o
n
bet
w
een n
odes or
som
e
o
t
h
e
r reaso
n
s
i
n
clud
ing
in
terferen
ce
will o
ccu
r. Th
eref
o
r
e, n
o
gu
aran
tee
is g
i
v
e
n
th
at co
ng
estion
o
c
curs b
y
counting the num
b
er of
F
Threshold
. Hence, i
t
m
u
st
be used as an i
ndi
cat
i
o
n o
f
co
nge
st
i
o
n det
ect
i
o
n wi
t
h
t
h
e
h
e
lp
fro
m
th
e n
e
two
r
k
layer.
In
t
h
is case,
nod
es
n
eed to
coo
p
e
rate with the qu
eu
e
statu
s
in
th
e
n
e
two
r
k
layer
as pre
v
i
o
usl
y
expl
ai
ne
d i
n
S
TEP I
(a).
Thi
s
wo
rk
co
m
b
in
ed
th
e MAC i
n
fo
rm
atio
n
wi
th
th
e ro
u
ting layer
pr
ot
oc
ol
t
o
det
ect
t
h
i
s
co
n
g
es
t
i
on.
To
detect such early congestion, E
q
uations
1, a
n
d
2
w
e
r
e
u
s
ed
to e
x
a
m
i
n
e
th
e in
s
t
an
ta
n
e
ou
s qu
eu
e
size and to se
nd wa
rning m
e
s
s
ages t
o
all nei
g
hbor
node
s wh
ile Equ
a
tion
3
was used
t
o
d
e
tect th
e MAC layer
in
fo
rm
atio
n
(Failu
re). Th
e
q
u
eu
e statu
s
in
t
h
e ro
u
ting
tab
l
e
u
tilizes co
ng
estio
n
p
r
ed
ictio
n. As an
ex
am
p
l
e, let
su
ppo
se if a giv
e
n
n
o
d
e
ch
eck
s
th
e
q
u
e
u
e
statu
s
con
d
ition
fo
r th
e
d
e
sti
n
atio
n
no
d
e
wh
ich
is‘1
’. Then
, th
e
second ste
p
is checki
ng
F
Thres
hold
. If th
e nu
mb
er
o
f
retransmit
F
T
h
reshol
d
for th
e MAC layer reach
e
s ‘1
1’, in
th
is
case, the
alternative route m
ech
ani
s
m
m
u
st
f
i
nd a
ne
w
r
out
e.
3.
2.
Alternati
v
e Route
Determinati
on
Al
t
e
rnat
i
v
e r
o
ut
es sh
oul
d b
e
fou
n
d
t
o
av
oi
d t
h
i
s
co
ng
est
i
on. T
h
u
s
t
h
e net
w
or
k t
o
pol
ogy
wa
s
rep
r
ese
n
t
e
d
usi
ng a
Tri
a
ng
ul
a
r
M
a
t
r
i
x
Ta
bl
e
(TM
T
) t
o
o
b
t
a
i
n
f
u
l
l
net
w
or
k t
o
pol
ogy
i
n
f
o
rm
at
i
on [
20]
.
The
Fig
u
re
4 illu
strates an ex
am
p
l
e to
sho
w
h
o
w th
e
TMT is
fi
lled
fro
m
n
e
two
r
k
t
o
po
log
y
.
Here, th
e triang
u
l
ar
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
I
J
ECE
Vo
l. 6
,
N
o
. 5
,
O
c
tob
e
r
20
16
:
232
2
–
23
30
2
326
m
e
t
r
i
c
dim
e
nsi
ons
are
eq
ual
t
o
t
h
e
num
ber
of
n
o
d
es.
T
hus
, eac
h
n
o
d
e
i
n
si
de
t
h
e
n
e
t
w
o
r
k
i
s
assi
gne
d
a
num
ber,
w
h
i
c
h
m
a
y
represe
n
t
t
h
e M
A
C
or
I
P
ad
dres
s. T
h
e
po
si
t
i
on
of t
h
e
no
de a
d
dress
i
s
rep
r
ese
n
t
e
d
o
n
t
h
e
di
ag
onal
o
f
t
h
e
TM
T, as
s
h
o
w
n i
n
Fi
g
u
r
e
4.
Fi
gu
re
4.
R
e
p
r
esent
T
r
i
a
n
gul
ar M
a
t
r
i
x
Tabl
e
Each
link
b
e
tw
een two
nodes r
e
p
r
esen
ts
b
y
“1
”
b
it in
si
d
e
TMT. O
t
h
e
r
w
ise, th
e ab
sen
ce
o
f
a li
n
k
b
e
tween
two
no
d
e
s will b
e
rep
r
esen
ted
b
y
“0
” b
it. Th
e po
sitio
n
in
sid
e
th
e TMT d
e
p
e
nd
s o
n
cro
s
sing
the row
no
de
wi
t
h
c
o
l
u
m
n
node
ad
d
r
e
ss.
To
fi
n
d
t
h
e
r
o
ut
es f
r
om
sou
r
ce n
ode
t
o
dest
i
n
at
i
on,
we
ass
u
m
e
d t
h
at
t
h
e
sou
r
ce
n
ode i
s
No
de
1 a
nd
t
h
at
t
h
e dest
i
n
at
i
on n
ode
i
s
No
de 4.
E
v
ery
bi
t
i
n
C
o
l
u
m
n
1 [1
, 1, 0,
1
,
0]
m
u
st
chec
k. The
co
rre
sp
on
di
n
g
connection rel
a
ted to Node i
s
(2,
3, 4, 5, a
nd
6).
Because “1” bit in the TMT
only represe
n
ts the links.
There
f
ore,
No
des
(2
,
3,
5
)
m
u
st
c
h
eck
i
t
,
i
f
any
o
n
e
has
ne
w l
i
n
k
wi
t
h
ot
hers
. Fi
rst
sa
ve
(
2
,
3,
5) i
n
t
h
e
vect
or
q
u
e
u
e
. M
o
r
e
ov
er, th
e un
i
q
ue f
u
n
c
tion
avoid
s
th
e in
ser
tio
n
o
f
an
y doub
le no
d
e
nu
mb
er
i
n
sid
e
th
e v
ector
que
ue.
Tec
hni
c
a
l
l
y
, t
h
e u
n
i
q
u
e
f
unct
i
o
n i
s
t
h
e key
t
o
e
n
su
re
l
o
o
p
-
f
r
ee.
If
t
h
e c
h
ec
k f
u
nct
i
on
d
o
es
n
o
t
fi
nd
t
h
e
d
e
stin
ation
nod
e, th
en
th
e first ele
m
en
t from th
e v
ecto
r
q
u
eue c
h
ec
ko
ut
,
m
eani
ng del
e
t
i
on f
r
om
vect
o
r
q
u
eue
,
and
bec
o
m
e
s the ne
xt step of the search
. He
re, the
vector queue
bec
o
m
e
s
(3, 5,
4
, 6)
.
At
t
h
i
s
poi
nt
,
t
h
e check
fu
nct
i
o
n fi
nd
s t
h
e
dest
i
n
at
i
o
n
no
de,
w
h
i
c
h
i
s
No
de
4.
T
h
e r
o
ut
e i
s
1
→
2,
→
4.
Acco
r
d
i
n
gl
y
,
i
f
any
n
o
d
e i
n
t
h
e net
w
o
r
k p
r
oba
bl
y
get
s
c
o
nge
st
ed, t
h
en t
h
e p
r
ocess o
f
f
i
ndi
n
g
ro
ut
e
sk
ip
s tem
p
o
r
ary fo
r t
h
at no
de b
y
pu
ttin
g
zero
b
it in
th
e
TMT for sh
ort ti
m
e
(5
ms
).
For i
n
stance
, let us
assum
e
t
h
e net
w
o
r
k
wi
t
h
N
o
de 2 a
b
o
u
t
co
n
g
est
i
o
n. T
h
en
put
“0
s” f
o
r N
ode
2 i
n
si
de T
M
T for
5
ms
to
a
v
o
i
d
t
h
i
s
no
de t
e
m
pora
r
y
t
i
m
e
. Howeve
r, N
o
de1
no
w has t
h
i
s
l
i
nks
(3
, 5)
. So l
i
k
ewi
s
e p
r
evi
o
us m
e
t
hod
(3
, 5) p
u
s
h
i
n
t
h
e que
ue
, t
h
en
pu
p 3 a
n
d
check t
h
e c
o
n
n
ect
i
o
n
s
wi
t
h
No
de 3
.
A
nd t
h
e ne
w co
nne
ct
i
ons f
o
r
No
d
e
3 t
o
que
ue
(5,4). Hence,
the ne
w alternative rout
e
is 1
→
3
→
4.
Fin
a
lly, con
c
lud
e
if th
e cong
estio
n
o
c
cu
rs in
Nod
e
2, t
h
en the
rout
e 1
→
3
→
4,
a
n
d 1
→
5
→
4 b
e
co
m
e
th
e altern
ativ
e
p
a
th.
4.
SIMULATION SET
U
P
Th
e m
a
in
g
o
a
l o
f
th
is sim
u
latio
n
is to
fin
d
altern
ativ
e p
a
th
s
wh
en
ev
er co
ng
estion h
a
pp
en
s is
im
m
i
nent
. The
r
ef
ore
,
a scena
r
i
o
was d
e
si
g
n
e
d t
o
sim
u
l
a
te th
is p
r
ob
lem
as sh
own
in
Fig
u
re 1
.
Th
e so
urce
n
o
d
e
s tran
sm
it
m
a
ssiv
e
d
a
ta traffic to
nod
e
4
0
b
y
using
the CBR d
a
ta traffic. Th
e sou
r
ce n
o
d
e
41
b
e
g
i
n
s
t
o
tran
sm
it d
a
ta traffic to nod
e
4
0
after
1
5
seco
nd
s. Th
en
, n
o
d
es 4
2
, 4
3
,
an
d 44
st
art
sen
d
i
ng
pac
k
et
s
a
f
t
e
r one
,
two
,
and th
ree second
s later,
resp
ectively. All
node
preparations
su
c
h
as
W
i
Fi, M
A
C,
Adhoc
W
i
f
iMac,
W
i
fiMacQueue, RtsCtsThres
hol
d,
DropTail
Que
u
e, a
nd DsssRate2Mbps
are
set according to Ta
ble
1.
The
ro
ut
i
n
g
pr
ot
oc
ol
s i
n
cl
ude
A
O
D
V
,
OLSR
,
an
d
DS
DV
p
r
ot
o
c
ol
s.
Tabl
e 1.
Sim
u
l
a
t
i
on par
a
m
e
t
e
rs
METR
ICS
VALU
E
Application protoc
ol
CBR
Nu
m
b
er
of nodes
45 nodes
Nu
m
b
er
of sour
ce
node
4 nodes
Sour
ce node I
D
Nodes 41,
42,
43,
4
4
Nu
m
b
er
of sink node
1 node
Sink node I
D
Node 40
W
i
-
F
i 802.
11
b
Packet size
128 By
te
T
r
ans
m
ission r
a
ng
e
250
m
dia
m
eter
Bandwidth link
2 M
bps
Sim
u
lation tim
e
30,
120 s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
N
ode
C
o
ope
rat
i
on t
o
Av
oi
d
E
a
rl
y C
o
n
g
est
i
o
n
Det
ect
i
on
Ba
sed
o
n
C
r
oss .
.
.. (
A
b
d
a
l
r
az
ak
Tare
q R
a
hem)
2
327
5.
NETWO
R
K PERFO
R
MA
NCE
Here
, the s
o
urce nodes
are c
o
nve
r
ge
nt, a
nd near
node
20. In these ci
rcum
s
t
ances, all source
node
s
sen
d
p
acket
s t
h
r
o
ug
h
on
e r
o
ut
e t
o
si
n
k
N
o
de
40
. T
h
e
sit
u
ation worsens if re
sour
c
e
s are
inc
r
ease
d
without
con
s
i
d
eri
ng t
h
e con
g
est
i
o
n t
y
pe, t
r
af
fi
c pat
t
e
rn
, and
net
w
or
k t
o
p
o
l
o
gy
. Th
e rout
i
ng p
r
ot
o
c
ol
sel
ect
s one
rout
e
t
o
sen
d
al
l
pac
k
et
s vi
a:
R
o
ut
e 1:
20
→
21
→
22
→
23
→
24
→
25
→
26
→
27
→
28
→
29. Technically, whe
n
appl
y
i
n
g
st
an
d
a
rd r
o
ut
i
ng
pr
ot
oc
ol
s, t
h
e r
o
ut
i
ng
pr
ot
oc
ol
doe
s not
co
nsi
d
er c
o
n
g
est
i
o
n
as wel
l
as t
h
e ot
her
no
des
.
H
o
wev
e
r, a
f
t
e
r a
ppl
y
i
ng
o
u
r
sc
hem
e
, t
h
e al
t
e
r
n
at
i
v
e ro
ut
e
wo
r
k
s
pr
o
p
erl
y
as s
h
ow
n i
n
Fi
gu
re
5. T
h
e
routing
protoc
ol forwards
th
e
pac
k
ets in an
efficient
path
because e
v
ery
node
posses
s
es t
h
e c
o
m
p
lete network
topology information. Nodes
coope
r
ate
with
each othe
r by sending warn
ing m
e
ssages, t
h
ere
b
y e
nha
nci
n
g the
ove
ral
l
net
w
or
k pe
rf
orm
a
nce.
The val
i
dat
i
o
n o
f
t
h
ese
o
b
s
e
rve
d
r
o
ut
es w
a
s achi
e
ve
d
by
W
i
res
h
ar
k s
o
f
t
ware
and
Net
A
ni
m
10
5.
R
out
e 1:
41
→
10
→
0
→
1
→
2
→
3
→
4
→
5
→
6
→
7
→
8
→
9
→
19
→
40
.
R
out
e 2:
42
→
10
→
11
→
12
→
13
→
14
→
15
→
16
→
17
→
18
→
19
→
40
.
R
out
e 3:
43
→
20
→
21
→
22
→
23
→
24
→
25
→
26
→
27
→
28
→
29
→
40
.
R
out
e 4:
44
→
30
→
31
→
32
→
33
→
34
→
35
→
36
→
37
→
38
→
39
→
40
.
Fig
u
re 5
.
Altern
ativ
e rou
t
e
The m
o
st
im
port
a
nt
m
e
t
r
i
c
t
h
at
can be u
s
ed t
o
m
easure net
w
o
r
k
pe
rf
orm
a
nce i
s
t
h
r
o
ug
h
put
.
Fi
g
u
re
6
sho
w
s t
h
e t
h
r
o
u
g
h
p
u
t
wi
t
h
45
n
odes a
n
d
sim
u
l
a
t
i
on t
i
m
e
of
12
0 s
.
T
h
e t
h
ro
u
g
h
p
u
t
m
easured
fo
r
wh
ol
e
network includes four nodes
sendi
ng
data traffic (Node
41 t
o
44). In
addition, one node receives
t
h
e
dat
a
tr
af
f
i
c, t
h
e sink
nod
e, nod
e
4
0
. G
e
n
e
r
a
lly, th
e th
ro
ugh
put
i
s
enhance
d
by
5
7
%, as a
resul
t
o
f
t
h
e r
e
duce
d
packet los
s
and the ne
w route to the
dest
i
n
at
i
on. F
u
rt
her
m
ore, t
h
e aver
age pac
k
et
del
i
very
rat
i
o
i
s
e
nha
nce
d
by
5
7
%
agai
ns
t
AO
D
V
pr
ot
o
c
ol
, as
sh
o
w
n
i
n
Fi
gu
re
7,
w
h
i
c
h al
s
o
s
h
o
w
s
t
h
e
resul
t
s
fo
r t
h
e
wh
ol
e
net
w
or
k.
Figure
6. Average T
h
roughput for s
o
urce
nodes
Figure
7. Average Pac
k
et
Del
i
very
Ratio fo
r sou
r
ce
no
des
The st
an
dar
d
r
out
i
n
g p
r
ot
oc
o
l
s AOD
V,
DS
DV
, an
d OLS
R
for
w
ar
d pac
k
et
s fr
om
Nod
e
s (4
1, 4
2
, 4
3
,
and 44) to
Node 40.
To
se
nd a
n
y data
pac
k
ets, the M
A
C la
yer m
u
st send
RTS and
recei
ve CTS fram
e
to hol
d
m
e
di
a, and t
h
e
n
sen
d
D
A
T
A
fram
e
and rece
i
v
e AC
K f
r
am
e. Thi
s
t
h
ree
-
way
ackn
o
w
l
e
dgem
e
nt
exha
ust
s
t
h
e
net
w
or
k.
T
hus
, t
h
e
be
ha
vi
o
r
o
f
ro
ut
i
n
g
pr
ot
oc
ol
s
doe
s
n
o
t
c
onsi
d
er
t
h
e cr
oss-l
a
y
e
r
bet
w
ee
n
net
w
or
k a
n
d
M
A
C
l
a
y
e
rs.
C
onse
q
uent
l
y
,
t
h
e t
h
r
o
ug
h
put
app
ears
ra
nd
o
m
for
N
odes
4
1
t
o
4
4
.
In t
h
e
seco
nd
scena
r
i
o
, t
h
e
272.3241
402.2381
399.0848
638.0421
0
200
400
600
800
Aodv
Olsr
Dsdv
CA
‐
TSG
Throughput
(Kbps)
Routing
Protocols
0
0.01
0.02
0.03
Aodv
Olsr
Dsdv
CA
‐
TSG
packet
delivery
ratio
Routing
protocols
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
I
J
ECE
Vo
l. 6
,
N
o
. 5
,
O
c
tob
e
r
20
16
:
232
2
–
23
30
2
328
sam
e
t
opol
o
g
y
has
bee
n
use
d
, e
x
ce
pt
t
h
e s
i
m
u
l
a
t
i
on t
i
m
e
i
s
3
0
sec
o
n
d
s
.
It
used
t
h
i
s
s
cenari
o
t
o
sh
o
w
t
h
e
beha
vi
o
r
of
t
h
e r
out
i
n
g
pr
ot
o
c
ol
i
n
sh
ort
pe
ri
o
d
s, e
s
peci
al
l
y
i
n
t
h
e
fi
rst
f
e
w sec
o
nds
. T
h
ere
f
o
r
e,
t
h
e
DS
D
V
pr
ot
oc
ol
i
s
not
i
n
cl
u
d
ed
i
n
t
h
e Fi
gu
res
gi
v
e
n i
t
s
l
o
ng
t
i
m
e req
u
i
r
em
ent
t
o
di
sco
v
e
r
t
h
e net
w
o
r
k a
n
d
b
u
i
l
d
ro
ut
i
n
g t
a
bl
es.
AO
D
V
ra
pi
dl
y
di
sco
v
ers
t
h
e ro
ut
e t
o
t
h
e
d
e
stin
ation
s
bu
t is no
t as efficien
t as th
e
OLSR
pr
ot
oc
ol
. T
h
e r
e
sul
t
sh
ow
s t
h
at
t
h
ro
u
g
h
p
u
t
f
o
r
pr
o
p
o
s
ed
pr
ot
oc
ol
i
s
agai
n
t
h
e best
resul
t
fo
r t
h
e s
h
ort
p
e
ri
o
d
of
pac
k
et
s se
n
d
i
n
g
fo
r
15
sec
o
n
d
s,
as s
h
ow
n
i
n
Fi
gu
re
8.
Fi
gu
re
8.
Th
r
o
ug
h
put
f
o
r a
v
e
r
age
4
so
u
r
ce
No
des
In Fi
gu
re
9, t
h
e t
h
r
o
u
g
h
p
u
t
f
o
r e
v
ery
s
o
u
r
c
e
no
de (
4
1,
42
, 4
3
, a
nd
4
4
) i
s
l
i
s
t
e
d. Si
m
i
l
a
rl
y
,
so
urc
e
no
des
41 a
nd
44
have
hi
g
h
t
h
r
o
ug
h
put
i
n
p
r
o
p
o
sed
pr
ot
oc
ol
. H
o
we
ve
r, f
o
r t
h
e
AO
D
V
and
OLSR
pr
o
t
ocol
s
,
onl
y
s
o
urce
n
o
d
e
41
ha
s
hi
g
h
t
h
r
o
ug
h
put
be
cause
of
t
h
e
t
h
ree-
way
ac
kn
o
w
l
e
d
g
em
ent
.
F
i
gu
res
10
,
11
s
h
o
w
th
e lo
st p
ack
et
s, th
e p
ack
et deliv
ery ratio
, resp
ectiv
ely. Fin
a
lly, it can
b
e
co
n
c
lud
e
d
th
at th
e th
ro
ugh
pu
t h
a
s
enha
nce
d
to
57.319% a
g
ains
t AODV
protocol. T
h
e r
ecei
ved
pac
k
ets enhance
d
to
57.348% agai
nst AODV.
The pa
cket del
i
very ratio e
n
hanced
t
o
57.35% against
AODV. The
pac
k
et
l
o
ss en
hanc
ed t
o
34
.9
6%
agai
nst
OLSR
pr
ot
ocol
.
Fi
gu
re 9.
Th
r
o
ug
h
put
f
o
r so
u
r
ce no
des 4
1
t
o
4
4
Fi
gu
re
1
0
. L
o
st
pac
k
et
s f
o
r
30
seco
n
d
s
Fig
u
r
e
11
. Pack
et
d
e
liver
y r
a
tio
fo
r
30
second
s
0
200
400
600
800
1000
1200
Aodv
Olsr
CA
‐
TSG
Throughput
(Kbps)
Routing
protocols
0
200
400
600
800
node
41
node
42
node
43
node
44
Throughput
(Kbps)
Routing
protocols
Aodv
Olsr
CA
‐
TSG
6760
6770
6780
6790
6800
6810
6820
Aodv
Olsr
CA
‐
TSG
packets
(pkt)
0.024
0.026
0.028
0.03
0.032
Aodv
Olsr
CA
‐
TSG
Packet
Delivery
Ratio
Routing
protocols
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
N
ode
C
o
ope
rat
i
on t
o
Av
oi
d
E
a
rl
y C
o
n
g
est
i
o
n
Det
ect
i
on
Ba
sed
o
n
C
r
oss .
.
.. (
A
b
d
a
l
r
az
ak
Tare
q R
a
hem)
2
329
6.
CO
NCL
USI
O
N
We pr
o
p
o
s
ed
a new sche
m
e
t
h
rou
g
h
whi
c
h earl
y
con
g
est
i
o
n can
be avoi
ded
base
d on t
h
e
i
n
f
o
rm
at
i
on fr
om
t
h
e M
A
C
and
net
w
or
k l
a
y
e
rs.
We ap
pl
i
e
d t
h
i
s
sc
hem
e
to t
h
e r
o
ut
i
ng
p
r
ot
ocol
wi
t
h
t
h
e hel
p
of net
w
ork topology, a
n
d s
u
c
ceeded in obtai
n
ing alternative routes
from
the s
o
urce
to the
destinati
o
n prior to
co
ng
estion
.
All n
o
d
e
s m
u
st p
a
y atten
tio
n
to
t
h
e warn
ing
m
e
ssag
e
and
ch
eck
th
e au
to
m
a
ti
c Failu
re
account
. If the
n
u
m
b
e
r reach
e
s 1
1
, th
en
t
h
e n
e
x
t
h
o
p
to
th
at n
o
d
e
will b
e
av
o
i
d
e
d. Con
s
eq
u
e
n
tly, ev
ery n
o
d
e can
reco
m
p
u
t
e
a n
e
w p
a
t
h
from th
e so
u
r
ce to
th
e d
e
stin
ati
o
n
to
tran
sfer
the data pac
k
ets. This pa
per
presente
d an efficient
and l
i
g
ht
wei
g
h
t
congest
i
on c
ont
rol
al
g
o
ri
t
h
m
.
The perf
o
r
m
a
nce charact
eri
s
t
i
c
s of t
h
e net
w
or
k i
n
cas
es of
th
ro
ugh
pu
t, lo
st p
ack
ets,
and packet delivery ratio were generally
enha
nced. Future efforts m
a
y enha
nce the
aspect
of p
o
w
e
r
con
s
um
pt
i
on by
di
st
ri
b
u
t
i
n
g
pow
er t
h
r
o
ug
ho
ut
t
h
e wh
ol
e
net
w
o
r
k
,
n
o
t
by
depe
n
d
i
n
g on
on
e
route alone.
ACKNOWLE
DGE
M
ENTS
Thi
s
wo
r
k
i
s
sup
p
o
rt
e
d
by
M
i
ni
st
ry
of
H
i
ghe
r E
d
ucat
i
o
n M
a
l
a
y
s
i
a
an
d m
i
ni
st
ry
of
sci
e
nce a
n
d
l
earni
n
g
,
un
de
r
Gra
n
t
No
. LR
GS/
T
D/
2
0
1
1
/
UKM
/
I
C
T
/
02/
02
. Al
s
o
Ira
qi
boa
r
d
of s
u
pre
m
e audi
t
Iraq/
B
a
gh
da
d
(G
ov
er
nm
ent
B
ody
)
.
A
u
t
h
or
s t
h
ank
s
Ira
qi
boa
r
d
of s
u
p
r
e
m
e audi
t
Iraq/
B
a
gh
da
d (G
o
v
e
rnm
e
nt
B
ody
)
whi
c
h
co
n
t
ribu
ted
effectiv
ely to
g
i
ve u
s
th
is
o
ppo
rt
u
n
ity fo
r pub
lish
i
ng
.
REFERE
NC
ES
[1]
M
.
I
.
A
.
A
.
R
.
T
.
R
a
h
e
m
,
et al.
, “
A
Com
p
arative and Anal
y
s
is
S
t
ud
y
of Vanet Routing Protocols,”
Journal of
Theoretical and
Applied
Informa
tion Technology,
vol. 66, pp. 691, 2014.
[2]
R. N. Devikar,
et al.
,
“
I
s
s
u
es
in
Routing M
e
chan
is
m
for P
acke
t
s
F
o
rwarding: A
S
u
rve
y
,”
Interna
tional Journal
o
f
Electrica
l
and
C
o
mputer Engin
e
ering (
I
JECE)
,
v
o
l. 6
,
pp
. 421-43
0, 2016
.
[3]
R. Havinal,
et al.
, “EASR: Graph-based Framework for Ener
g
y
Ef
fic
i
ent S
m
art Routing i
n
MANET using
Availab
ilit
y Zon
e
s,”
Internationa
l Journal
of Electrical and Comp
uter Eng
i
neering
,
vol. 5
,
2015
.
[4]
C. Cetinkay
a
, “Multi-ch
annel
cooperat
i
ve
MAC
protoco
l
for
wir
e
less LANs,”
Ad
Hoc Ne
twor
ks
,
vol. 28
, pp
. 17-3
7
,
2015.
[5]
E.
G.
Villegas,
et al.
, “A novel cheater
and jammer detection
sche
me for IEEE 802.11-ba
s
e
d wireless
LANs,”
Computer Netwo
r
ks,
vol. 86
, pp
.
40-56, 2015
.
[6]
S. M. Aghdam,
et al.
, “WCCP: A congestion co
ntrol proto
c
ol fo
r
wireless multimedia comm
unication in senso
r
networks,”
Ad H
o
c Ne
twor
ks
,
vo
l. 13
, Part B, pp.
516-534, 2014
.
[7]
C. Sergiou
,
et al.
, “Congestion control
in Wireless Sensor Networks
through d
ynamic alternat
iv
e path selection
,
”
Computer Netwo
r
ks,
vol. 75
, Part A, pp
. 226-238
,
2014.
[8]
A.
P.
Silva
,
et al.
, “A survey
on
congestion con
t
r
o
l for de
lay
and disruption
toler
a
nt
networks,”
A
d
Hoc Networ
ks
,
vol. 25
, Par
t
B
,
p
p
. 480-494
, 201
5.
[9]
L. Jun, “A cross-lay
e
r routing
optimizati
on method in Wireles
s
Mesh Networ
k,” in
Software Engineering and
Service S
c
ien
c
e
(
I
CSESS
)
, 2013
4th
IEEE Intern
ational Con
f
eren
ce on
, pp
. 357-3
60, 2013
.
[10]
G. Karbaschi
an
d A. Fladenm
u
ll
er, “A link-quali
t
y
and
cong
estio
n-aware cross lay
e
r m
e
tric for m
u
lti-hop wir
e
less
routing,” in
Mob
ile Adhoc and S
e
nsor Systems Conferen
ce, IEEE I
n
ternational Co
nference on
, pp
.
7, 655
, 2005
.
[11]
M. Haghpanah
i
,
et al.
, “Topolog
y
control
in lar
g
e-scale wireless sens
or networks: Between
information source an
d
sink,”
Ad H
o
c
N
e
twor
ks
,
vo
l. 11, pp. 975-990, 20
13.
[12]
A. Ghaffar
i
, “C
ongestion
contr
o
l mechan
isms
in wireless sensor networks: A s
u
rvey
,”
Journal of Network an
d
Computer Applications,
vol. 52
,
pp. 101-115
, 20
15.
[13]
K. Jaewon,
et a
l
.
, “TARA: Topo
log
y
-Awar
e
Res
ource Adap
tatio
n
to
Alleviate C
ongestion in
Sensor Networks,”
Parallel and Dis
t
ributed
System
s
,
IEEE Transactions on,
vol. 18
,
pp. 919-931
, 20
07.
[14]
G. Feng,
et al.
, “Modified RED gateway
s
under bursty
traff
i
c,”
Co
mmunications L
e
tt
ers, IEEE,
vol. 8, pp. 323-325,
2004.
[15]
S. Flo
y
d
and V.
Jacobson, “Rand
o
m early
d
e
tection gatew
a
y
s
for
congestion avoidance,”
N
e
tworking, I
EEE/ACM
Transactions on,
vol. 1, pp. 397-
413, 1993
.
[16]
C. Sergiou
,
et al.
, “Congestion control
in Wireless Sensor Networks
through d
ynamic alternat
iv
e path selection
,
”
Computer Netwo
r
ks,
vol. 75
, pp
.
226-238, 2014
.
[17]
N.
J.
Husse
in,
et al.
, “IR and
Multi Scal
e R
e
tin
ex im
age E
nha
ncem
ent for
Concealed Weapon Detection
,
”
Indonesian Jour
nal of Elec
trical Engineering
and
Computer Scien
ce,
vol. 1
,
pp
. 39
9-405, 2016
.
[18]
L. Jun, “A cross-lay
e
r r
outing op
timization meth
od in
Wire
le
ss Me
sh Ne
twork,
” in
IEEE Int
e
rnational Conf
erenc
e
on Software Eng
i
neering
and S
e
rvice S
c
ien
c
e ICS
E
SS
, pp
. 357-36
0, 2013
.
[19]
S. S. Kumaran, “Early
C
onges
tion Detection
and Self Cur
e
Routing in Manet,”
Computer
Networks and
Information Technologies,
pp. 5
62-567, 2011
.
[20]
A.
A.
R.
T.
Rahem,
et al.
, “A Triangular Matrix
Routing Table
Repres
entation f
o
r Efficient Routing in Manet,"
Journal of Theoretical
&
A
pplied Information Technology,
vol. 64
, 2014.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
I
JECE
Vo
l. 6
,
N
o
. 5
,
O
c
tob
e
r
20
16
:
232
2
–
23
30
2
330
BIOGRAP
HI
ES OF
AUTH
ORS
Abd Al-razak
T
a
req Rah
e
m
is current
l
y
a P
h
.
D
candida
te at Universiti
Keb
a
ngsaan
Mala
ysi
a
(UKM), the Dep
a
rtm
e
nt of
El
ectr
i
ca
l,
Ele
c
tron
ic
s
and S
y
stems Engineer
ing. He receiv
e
d th
e B.S
Computer Engin
eering
and Infor
m
ation Technol
og
y
from University
o
f
Technolog
y
,
Baghdad
,
Iraq,
in 2002
,
and the Master
of Technolog
y
de
gree in
Infor
m
ation Technolog
y
college of
Engineering fro
m BVDU, Pune
, Indi
a, in 2012. His current resear
ch in
ter
e
sts include wireless
networking and
mobile ad hoc network. Rou
ting Protocol. Network Performances. He was
consulting
in D
G
Pioneer M
a
gazine in
Iraq
(200
5-2008)
.
Mahamod Ismai
l
is currently
a professor in Communications
Engineering at the Universiti
Kebangs
aan M
a
la
ys
i
a
(UKM
). He received t
h
e B.S
c
. in El
ectr
i
ca
l and El
ectron
i
cs
from
University
of Strathcly
d
e, U
.
K.
in 1985, the
M.Sc. in Communi
cation E
ngineering and Digital
Electronics from UMIST, Manch
e
st
er U.K.
in 19
87, and
the
P
h
.D. from
Univers
i
t
y
of
Bradford
,
U.K. in 1996.
Professor Mahamod is an ex
ecutiv
e member of Communic
a
tions/Vehicular
Techno
log
y
S
o
c
i
et
y,
IE
EE M
a
l
a
y
s
i
a
chapt
e
r.
His
current
res
earch
int
e
res
t
s
i
n
clude M
obi
le
Communications and Wireless Networking.
Evaluation Warning : The document was created with Spire.PDF for Python.