Int
ern
at
i
onal
Journ
al of Ele
ctrical
an
d
Co
mput
er
En
gin
eeri
ng
(IJ
E
C
E)
Vo
l.
10
,
No.
4
,
A
ugus
t
2020
,
pp. 370
2~37
14
IS
S
N: 2
088
-
8708
,
DOI: 10
.11
591/
ijece
.
v10
i
4
.
pp3702
-
37
14
3702
Journ
al h
om
e
page
:
http:
//
ij
ece.i
aesc
or
e.c
om/i
nd
ex
.ph
p/IJ
ECE
Dynami
c r
outing d
iscovery s
chem
e for hig
h mobil
ity in m
ob
ile
ad hoc wi
re
l
ess netwo
rks
Ha
ider
Alani
1
,
M
aha
Abdel
ha
q
2
,
Raed
A
l
sa
q
ou
r
3
1
IT
C
en
te
r
and
S
y
stem,
Ira
q
i
Min
istr
y
of
Elec
tr
ic
i
t
y
,
Baghd
ad, I
raq
2
Depa
rtment of I
nform
at
ion
T
ec
h
nolog
y
,
Co
ll
eg
e of
Com
pute
r and
Inform
at
ion
Sci
enc
es,
Prince
ss
Nourah
bint Abdulr
ahman
Univer
si
t
y
,
Saudi
Arab
ia
3
Depa
rtment of I
nform
at
ion
T
ec
h
nolog
y
,
Co
ll
eg
e of
Com
puti
ng
an
d
Inform
at
ic
s
,
Saudi
E
le
c
tronic
Univer
sit
y
,
Sau
di
Arabi
a
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
Dec
18, 201
9
Re
vised Jan
29
, 2020
Accepte
d
Fe
b 1, 2
020
An
innova
ti
v
e
te
chno
log
y
th
at
is
widely
use
d
in
m
an
y
app
li
c
at
ions
is
the
Mobil
e
Ad
-
hoc
Network
(
MA
NET)
.
Disc
over
y
and
m
aintena
n
ce
of
route
s
at
MA
NET
ar
e
importa
nt
issues.
W
it
hi
n
MA
NET,
broa
dca
st
ing
is
used
to
discove
r
a
pat
h
withi
n
on
-
demand
routi
ng
protoc
ols.
Esta
b
li
shing
and
m
ai
nta
ini
ng
a
r
oute
per
iod
ically
among
the
n
odes
is
the
challe
ng
e
that
req
uire
s
th
e
tra
n
sm
it
ti
ng
of
cont
r
ol
pac
k
et
s
ac
ross
a
net
work.
Th
is
stat
e
l
ea
ds
to
the
issue
of
broa
dca
st
ing
storm
s.
Broadc
asting
cont
rol
pa
ck
et
s
inc
re
ase
cont
rol
pa
cke
ts
over
hea
d
and
d
e
cre
ase
ne
twork
per
form
anc
e
.
In
thi
s
pape
r,
we
proposed
a
sche
m
e
ca
lled
AO
DV
-
Veloc
ity
and
D
y
namic
(
AO
DV
-
VD
)
for
eff
ec
t
ive
bro
adc
ast
con
trol
p
ac
ke
ts.
The
rout
ing
protoc
ol
for
th
e
ad
-
hoc
on
-
demand
dista
nce
v
ic
tor
(A
OD
V)
is
used
to
implement
t
he
proposed
AODV
-
VD
sch
eme.
AO
DV
-
V
D
sche
m
e
red
u
ce
s
both
th
e
ex
ce
ss
ive
rout
e
discove
r
y
contro
l
pac
ke
ts
and
ne
twork
over
hea
d
.
Network
sim
ula
tor
ver
sion
2.
35
(NS
2.
35)
was
used
to
co
m
par
e
the
pro
p
osed
AO
DV
-
V
D
sche
m
e
to
the
AO
DV
routi
ng
proto
col
in
te
rm
s
of
e
nd
-
to
-
end
l
aten
c
y
,
ave
rag
e
throughput
,
pa
ck
et
tra
nsm
ission ra
ti
o
and
over
he
a
d
ratio.
Ke
yw
or
d
s
:
AOD
V
Link b
reak
a
ge
MANET
Mob
il
it
y
Rou
ti
ng
disco
ve
ry
Copyright
©
202
0
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
:
Ma
ha Abdel
ha
q,
Dep
a
rtm
ent o
f Info
rm
at
ion
Te
chnolo
gy
,
Coll
ege
of
C
om
pu
te
r
an
d
I
nf
or
m
at
ion
Scien
ces
,
Pr
inces
s No
ur
a
h bint
Abd
u
lra
hm
an
Un
i
ver
si
ty
,
84428 R
iy
adh,
Saudi A
ra
bia
.
Em
a
il
: M
SA
bdel
haq@pn
u.
e
du.sa
, m
aha.uk
m
@g
m
ail.co
m
1.
INTROD
U
CTION
Mob
il
e
A
d
-
hoc
Networ
k
(M
AN
E
T)
is
a
m
ob
il
e,
in
fr
ast
ruct
ur
e
-
fr
ee
,
sel
f
-
co
nf
i
gurin
g,
i
nteracti
ve
an
d
m
ul
ti
-
hop
network
[
1].
M
A
NET
is
a
n
em
erg
e
ncy
netw
ork
with
a
ra
nge
of
m
ob
il
e
nodes
.
T
he
nodes
in
MANET
m
ov
e
ra
ndom
ly
wit
hin
t
he
netw
ork
du
e
to
ra
pid
and
f
reque
nt
to
po
l
og
y
c
ha
nge
s
in
t
he
MA
N
ET
[
2].
Su
c
h
m
ob
il
e
nodes
ca
n
f
un
ct
ion
sim
ultaneousl
y
as
hosts
an
d
r
ou
te
rs
since
they
can
travel
any
where
in
the
MA
NET
[
3].
I
n
M
AN
E
T
,
it
is
the
res
po
ns
ibil
it
y
of
t
he
routin
g
pr
oto
c
ol
to
de
velo
p
a
route
betwee
n
n
od
es
to
transm
it
pack
et
s,
w
hich
s
el
ect
s
the
op
ti
m
u
m
and
m
ain
ta
ins
the
r
ou
t
e.
W
he
n
a
source
no
de
wa
nts
to
transm
it
data,
t
he
r
ou
ti
ng
pr
oto
col
co
ns
t
ru
ct
s
a
path
betwee
n
a
source
node
and
a
desti
na
ti
on
node
.
MA
NET
routin
g protoc
ols ar
e
d
i
vid
e
d i
nto
th
ree clas
s
ific
at
ion
s:
po
si
ti
ve,
r
eact
ive
a
nd h
y
br
i
d
[
4].
In
M
A
NET
routing
pr
oto
c
ols,
a
broad
c
ast
in
g
sc
hem
e
is
necessary
w
he
n
pac
kets
are
se
nt
bet
wee
n
m
ob
il
e
no
de
s
to
m
ai
ntain
network
co
nnec
ti
vity
[5
]
.
Br
oa
dcasti
ng
is
usual
ly
def
i
ned
as
the
proce
ss
of
transm
itti
ng
a
pack
et
to
al
l
node
s
in
the
net
work
from
a
s
ource
no
de.
Br
oad
ca
sti
ng
is
m
or
e
fr
eq
uen
tl
y
us
ed
in
MA
NET
w
hen
the
s
ourc
e
node
broa
dc
ast
s
Rou
te
Re
qu
e
st
(RREQ
)
pac
kets
to
c
heck
f
or
a
r
oute
to
a
destinat
ion
node
,
usual
ly
in
path
disco
ve
r
y.
Broad
cast
in
g
is
of
t
en
us
e
d
in
ro
ute
m
ai
n
te
nan
ce
w
he
n
node
s
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N: 20
88
-
8708
Dyna
mic r
ou
ti
ng d
isc
overy
s
chem
e
for
h
i
gh mob
il
it
y in
mobile
ad hoc
wi
rel
ess n
et
work
s
(
Ha
ider
Ala
ni
)
3703
exch
a
nge
Hell
o
pa
ckets
for
gathe
rin
g
nei
ghbo
rin
g
in
for
m
at
ion
.
I
nter
m
ediat
e
nodes
in
MA
NET
assist
in
the
broadca
st
process
.
In
te
r
m
ediat
e
no
des
are
respon
si
bl
e
fo
r
f
orward
i
ng
the
pac
ket
to
oth
er
no
de
s
in
the n
et
wor
k
f
r
om
the sour
ce
node [
6].
The
flo
odin
g
r
esults
in
m
ultip
le
re
dunda
nt
transm
issi
on
s
on
the
net
wor
k,
i
n
w
hich
a
node
from
var
i
ou
s
nodes
can
get
the
payl
oad
of
th
e
sa
m
e
pack
et
.
This
eve
nt
is
known
as
a
broad
cast
i
ng
storm
pro
blem
[7
]
.
The
issue
of
broa
dcasti
ng
s
torm
s
le
ads
to
regular
c
on
te
ntion
a
nd
pac
ket
colli
sion
s
,
wh
ic
h
increases
net
w
ork
over
hea
d
t
raffic
.
I
n
the
r
ou
te
m
ai
ntena
nce
ph
a
se,
t
he
broa
dcast
st
orm
pr
oble
m
occu
rs
,
durin
g
w
hich
r
ou
te
s
a
re
upda
te
d
by
act
ivati
ng
ne
w
r
ou
te
di
scov
e
ry
reque
sts
to
rep
la
ce
b
roke
n
r
ou
te
s.
I
n
this
pap
e
r,
we
s
ug
gested
a
sche
m
e
called
AODV
–
Velocit
y
and
Dynam
ic
(AODV
-
V
D)
f
or
ef
fici
ent
broad
ca
st
con
t
ro
l
pack
et
s.
The
res
ults
ob
ta
ine
d
from
evaluati
ng
th
e
AOD
V
-
VD
schem
e
are
ve
ry
posit
ive
an
d
s
how
the ef
fecti
ve
ne
ss of the
propo
sed
sc
h
em
e in mi
t
igati
ng
t
he pr
oble
m
o
f
the
broa
dcast sto
r
m
.
The
rem
ai
nd
er
of
the
pap
e
r
is
struct
ur
e
d
acc
ordin
g
t
o
t
his.
We
pro
vid
e
a
descr
i
ption
of
the
researc
h
backg
rou
nd
a
nd
w
ork
relat
ed
to
it
in
Sect
ion
2.
Sect
io
n
3
descr
i
bes
the
ne
w
schem
e
AO
D
V
-
V
D.
Sec
ti
on
4
descr
i
bes
t
he
env
i
ronm
ental
sim
ulatio
n
a
nd
the
pe
rfo
rm
ance
m
et
ric.
Sect
ion
5
a
ddre
sses
the
fin
dings
a
nd
evaluati
ons a
nd Secti
on
6 su
m
m
arizes ou
r
re
search
and
ou
r
f
uture
w
ork.
2.
BACKG
ROU
ND AN
D REL
ATED W
O
RK
2.1
.
Ad
-
ho
c
on
-
de
man
d
dist
an
c
e v
ec
to
r
routi
ng
pro
tocol
AOD
V,
t
hat
Per
kin
s
a
nd
Roye
r
bu
il
t
[
8]
,
is
a
reacti
ve
r
outi
ng
prot
oco
l,
an
d
i
ts
m
ulti
-
hop
routin
g
protoc
ol
an
d
disc
overy
react
on
dem
and
.
In
the
AOD
V
routing
pr
oto
c
ol,
the
stre
ng
t
hs
of
Desti
natio
n
Se
qu
e
nce
d
Dista
nce
Vecto
r
Ro
uting
(
DSDV
)
and
Dy
nam
ic
Sour
ce
R
outi
ng
(
DS
R
)
proto
cols
ar
e
com
bin
ed
[8]
.
The
s
ource
no
de
broa
dcasts
a
r
ou
te
request
(RREQ
)
pac
ke
t
thr
oughout
t
he
MA
NE
T
nodes
i
n
the
route
disc
overy
pr
ocess
of
the
A
O
DV
routing
proto
col
on
M
ANET
an
d
set
s
a
tim
er
t
o
wait
f
or
the
r
eply
.
The
R
RE
Q
pa
cket
co
ntains
r
ou
ti
ng
i
nfor
m
at
ion
,
i
nclu
ding
the
I
P
ad
dress
of
ori
gin
at
or,
ID
of
br
oad
cas
t
and
seq
uen
ce
num
ber o
f desti
nation.
Each
inte
rm
ediat
e
node
recei
ves
the
RR
EQ
pac
ket
an
d
holds
t
wo
oper
at
ion
s
run
ning
the
re
verse
path
to
the
s
ource
node.
Fir
st
,
the
interm
e
diate
node
che
cks
if
it
has
previo
us
ly
received
the
RR
E
Q
pack
et
with
the
sam
e
or
i
gin
at
or
IP
a
ddress
a
nd
bro
adcast
ID,
an
d
then
determ
ine
wh
et
he
r
the
RR
EQ
pac
ket
sh
oul
d
be
re
j
ect
ed
or
acce
pted.
Se
cond,
if
t
he
RR
EQ
pac
ket
is
acce
pted
the
interm
e
diate
node
will
check
the
destinat
io
n
sequ
e
nce
num
ber
sto
re
d
in
it
s
routing
ta
ble.
The
interm
ediat
e
no
de
uni
-
c
ast
the
Rou
te
Re
ply
(RREP)
pack
e
t
to
the
so
ur
ce
node
if
the
num
ber
of
the
seq
uen
ce
is
gr
eat
er
than
or
equ
al
to
the
num
ber
con
ta
ine
d
i
n
the
RR
EQ
.
I
f
no
i
nterm
ediat
e
node
has
a
su
f
fici
ently
f
resh
pat
h
to
t
he
de
sti
nation
node
(n
e
w
destin
at
ion
se
quence
nu
m
ber
)
,
the
n
the
RR
EQ
pa
cket
m
us
t
con
ti
nu
e
na
vig
at
ing
unti
l
it
reaches
the
destinat
io
n
node.
Fig
ure
1
show
s
the
source
node
(S)
that
transm
it
s
RR
EQ
p
ac
kets
ov
e
r
the
netw
ork
t
o
it
s
neighb
or
i
ng
no
des
un
ti
l
the
RR
EQ
pa
cket
reac
hes
the
destin
at
ion
node
(
D).
Figure
2
al
s
o
s
hows
the d
e
sti
nation
node (
D)
w
hich replie
s to
the
sour
ce
no
de w
it
h
an
RR
E
P.
Figure
1.
A
ODV br
oa
dcast
s
R
REQ
pack
et
Figure
2.
A
ODV
r
epl
ie
s
RR
E
P p
ac
ket
In
AO
DV, ev
e
ry n
ode in the
netw
ork
al
so
s
ends H
el
lo p
ac
ket p
eri
od
ic
al
ly
to
keep
it
s r
outi
ng
table at
the
one
-
hop
ne
ighbor.
A
H
e
ll
o
pack
et
is
us
e
d
to
deter
m
ine
if
the
ne
ighborin
g
c
onnecti
on
is
sti
ll
al
ive.
The
no
de
sen
ds
Hell
o
pac
ket
with
a
tim
e
in
te
rv
al
cal
le
d
he
ll
o
-
interv
al
to
it
s
neigh
bor
node
to
detect
br
oke
n
li
nk
s
betwee
n
the
no
des.
Ever
y
node
s
ends
Hell
o
pa
ckets
to
it
s
neig
hbors
an
d
they
receive
their
ackno
wled
gm
e
nt.
I
f
a
no
de
sen
ds
Hell
o
pa
ckets
to
a
ne
ighbor
twic
e
a
nd
ha
s
not
rec
ei
ved
a
m
essa
ge
of
ackno
wled
gm
e
nt for i
t, the
n
t
he node i
niti
at
es the
bro
ken li
nk pr
ocess.
2.2
.
Rela
ted
w
ork
In
[
9]
,
the
a
ut
hors
pro
posed
a
r
ou
ti
ng
pro
tocol
cal
le
d
Mob
il
it
y
Aw
a
re
an
d
D
ual
-
P
hase
Ad
-
hoc
On
-
de
m
and
Di
sta
nce
Vect
or
with
Ad
a
ptive
Hell
o
Me
ssa
ge
s
(MA
-
DP
-
A
O
DV
-
A
HM)
.
It
fo
c
us
es
pri
m
ar
il
y
on
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
10
, No
.
4
,
A
ugus
t
2020
:
3702
-
3714
3704
dev
el
op
i
ng
rou
te
s
that
con
sid
er
the
sp
ee
ds
a
nd
directi
on
of
m
otion
of
vehi
cl
es
with
reg
a
rd
to
s
ource
ve
hicle
s,
there
by
pro
vidi
ng
m
or
e
ef
fici
ent
r
ou
te
s
a
nd
redu
ci
ng
r
oute
br
ea
ks
.
The
a
uthors
ha
ve
s
ugge
ste
d
a
n
ada
ptiv
e
pack
et
noti
ficat
ion
syst
em
,
w
hich
is
dir
ect
ly
connecte
d
to
the
re
gula
r
hello
m
es
sages
so
l
ution,
w
i
t
h
the
s
peeds
of
the
ve
hicle
s,
re
su
lt
ing
in
a
s
ubsta
ntial
re
duc
ti
on
i
n
the
am
ou
nt
of
ov
e
r
hea
d
c
on
t
r
ol
an
d
ne
twork
congesti
on.
T
he
protoc
ol
was
evaluated
us
i
ng
Q
ualNet
sim
ula
tor
in
te
r
m
s
of
over
hea
d
co
ntr
ol,
e
nd
-
to
-
e
nd
delay
and
ene
r
gy
con
s
um
ption
.
Re
su
lt
s
of
t
he
si
m
ulati
on
su
m
m
arized
that
MA
-
DP
-
A
O
DV
-
A
HM
eff
e
ct
ively
con
t
rib
utes
to
m
itigati
ng
net
work
i
ns
t
abili
ty
by
pr
od
ucin
g
reli
able,
sec
ur
e
r
oute
s
an
d
reducin
g
co
nnect
ion
fail
ur
es
. I
t
has
al
so
s
how
n
it
s
su
pe
rio
rity
ove
r
A
O
DV an
d M
DA
-
A
O
DV.
In
[
10]
,
t
he
auth
or
s
intr
od
uced
a
Qu
al
it
y
of
Ser
vice
(QoS
)
al
gorithm
in
ad
ho
c
netw
orks
.
The
pr
opos
e
d
al
gorithm
entitled
Mod
i
fied
Ligh
t
Weig
ht
distrib
uted
the
QoS
Al
gorith
m
,
ta
kin
g
int
o
account
the
posit
ion
a
nd
relat
ive
vel
oc
it
y
of
the
node
s
and
est
im
ati
ng
th
e
i
m
po
rta
nce
of
li
nk
i
ng
reli
abili
ty
,
choosin
g
a
path
with
l
onge
r
du
rati
on,
reducin
g
li
nk
br
ea
ks
a
nd
e
nhancin
g
Q
oS.
T
he
aut
hors
m
e
asur
e
d
the
al
gorithm
'
s
eff
ic
ie
ncy
i
n
pa
cket
delivery
rati
o,
pac
ket
de
li
ver
y
la
te
ncy
,
ave
ra
ge
e
nd
-
to
-
e
nd
delay
over
hea
d
r
out
in
g
a
nd
pack
et
s
lost.
The
sim
ulatio
n
res
ults
sh
ow
that
the
Mod
i
fied
Lig
ht
Weig
ht
distrib
uted
Q
oS
Algorithm
enh
a
nces
the
Q
oS
pa
ram
et
ers
of QoS i
n
a
d h
oc netw
orks.
In
[
11]
,
the
auth
or
s
propo
sed
two
pr
e
di
ct
ive
li
nk
avail
abili
ty
te
chn
iq
ues
pe
rform
ed
du
ri
ng
the
co
ntr
ol
pa
c
ket
r
ou
ti
ng.
T
he
first
te
ch
niqu
e
ref
e
rr
e
d
t
o
a
s
zo
ne
-
base
d
e
stim
ation
div
id
es
each
node'
s
area
of
t
ran
sm
issi
on
int
o
th
ree
a
r
eas,
nam
el
y
inn
er
,
m
idd
le
,
a
nd
oute
r,
ba
se
d
on
the
si
gn
a
l
stren
gth
obta
ined.
Her
e
,
nodes
ly
ing
in
the
i
nn
e
r
and
oute
r
z
on
e
are
pro
hib
it
ed
from
transm
i
tt
i
ng
the
co
ntr
ol
pack
et
as
t
hey
hav
e
extrem
el
y
hig
h
and
lo
w
sign
a
l
streng
th
r
ecei
ved,
res
pecti
ve
ly
.
No
des
in
th
e
m
idd
le
reg
io
n
are
only
perm
itted
inside
the
loca
li
ty
fo
r
the
co
ntr
ol
pack
et
fl
ow.
The
seco
nd
te
chn
i
qu
e
is
an
i
m
pr
ov
em
ent
ov
e
r
zo
ne
-
base
d
est
i
m
ation
,
in
wh
ic
h
no
des
are
co
ns
ide
re
d
f
or
r
ou
te
se
le
ct
ion
in
the
ou
te
r
zo
ne
a
nd
a
re
re
ferre
d
to
a
s
segm
ent
-
base
d
est
i
m
at
ion
.
It
cal
culat
es
the
l
ink
a
vaila
bili
ty
rati
o
of
the
no
de
for
eac
h
a
dj
acent
li
nk
base
d
on
the
nei
ghbor'
s
current
posit
ion
,
it
s
ad
j
ace
nt
segm
ent,
and
the
sect
or
in
w
hich
t
he
nei
ghbor
is
pr
ese
nt
withi
n
the
no
de'
s
tran
sm
it
reg
ion.
S
uch
inf
orm
ation
assist
s
in
the
trans
fer
of
c
ontr
ol
pac
kets
within
t
he
ou
te
r
re
gion.
The
sim
ulati
on
wa
s
pe
rfo
rm
ed
ov
e
r
DS
R
and
A
ODV
protoc
ols
in
te
r
m
s
of
con
t
ro
l
ov
e
r
head,
ba
ndwi
dth,
nu
m
ber
of
c
ontrol
pack
et
s
,
le
ng
t
h
of
the
r
oute
,
a
nd
nu
m
ber
of
r
ou
te
e
rror
s
ge
ne
rated
within
t
he
net
work.
The
res
ults
show
that
knowl
edg
e
of
the
li
nk
sta
tus
durin
g
the
ro
utin
g
ph
ase
helps
to
fi
nd
m
or
e
sta
ble
ro
utes
with
reduce
d o
verhea
d
c
on
tr
ol
an
d t
hus i
ncr
e
ases the
overal
l perfo
rm
ance o
f
the
netw
ork.
In
[12]
,
a
nei
ghbori
ng
in
f
orm
at
ion
-
base
d
broa
dcasti
ng
s
chem
e
is
being
intr
oduce
d
to
m
ini
m
iz
e
la
te
ncy
fo
r
a
d
ho
c
wireless
netw
orks.
In
the
schem
e,
few
Hell
o
m
e
ssages
a
re
interch
a
nged
t
o
gathe
r
inf
or
m
at
ion
about
on
e
-
ho
p
neig
hbors.
T
he
in
form
at
io
n
c
ollec
te
d
is
us
e
d
to
cal
c
ulate
the
de
nsi
ty
of
the
nei
ghbor,
the
rati
o,
a
nd
t
he
nu
m
ber
of
expose
d
neig
hbors
with
on
e
-
hop,
on
wh
ic
h
the
pr
ob
a
bili
ty
and
delay
of
retra
nsm
issi
on
are
m
od
ifie
d
.
T
he
way
the
pr
ob
a
bili
ty
and
dela
y
of
re
broa
dca
sti
ng
are
desc
r
ibed
in
the
neig
hbori
ng
kn
ow
le
dg
e
-
base
d
broa
dca
st
schem
e
decr
eases
the
ov
e
rh
ea
d
tra
ns
m
is
sion
a
nd
e
ff
ec
ti
vely
restrains
the
a
ggre
gation
of
t
raffic
.
A
fter
t
ha
t,
a
m
et
ho
d
f
or
t
he
velocit
y
-
base
d
de
li
ve
r
y
of
data
is
pr
opos
e
d
and
a
pp
li
e
d
to
a
neig
hbori
ng
knowle
dge
-
base
d
br
oad
c
ast
schem
e
to
fu
rt
her
re
duc
e
la
te
ncy,
creati
ng
neig
hbori
ng
kn
ow
le
dg
e
a
nd
sp
eed
-
based
bro
adcast
schem
e.
It
is
sp
eci
fied
that
few
higher
-
vel
ocity
no
de
s
are
us
e
d
to
retra
nsm
it
the
i
nco
m
i
ng
m
essage
with
a
gr
eat
er
prob
a
bili
ty
.
The
si
m
ulati
on
determ
ines
the
sc
hem
es
eff
ic
ie
ncy
unde
r
dif
fer
e
nt
network
c
onfig
ur
at
ion
s.
T
he
res
ults
sh
ow
that
the
new
sc
hem
e
ou
tpe
rfo
rm
s
curren
t
broa
dcast
sche
m
es
in
te
r
m
s
of
ov
e
r
head,
and
i
n
pa
rtic
ular,
in
the
a
ver
a
ge
en
d
-
to
-
e
nd
delay
.
The
pro
pos
e
d
schem
e
lowers
the
over
hea
d
by
88.
4
pe
r
cent
relat
ive
t
o
fl
oodi
ng
an
d
the
ave
ra
ge
en
d
-
to
-
e
nd
de
la
y
by
a m
axi
m
u
m
o
f
88.9 pe
rce
nt.
In
[13]
,
the
a
uthors
sug
gested
a
novel
di
sco
ver
y
sc
he
m
e
within
MANE
T
cal
le
d
the
Dyn
am
ic
Pr
oba
bili
sti
c
Rou
te
(
DP
R)
.
The
DP
R
sc
hem
e
fo
rw
a
rded
the
pac
ket
s
with
a
dynam
ic
al
l
y
determ
ined
pro
bab
il
it
y
known
as
a
fo
r
w
ard
i
ng
pr
ob
a
bi
li
ty
(F
P)
to
the
neighb
or
in
g
node
.
The
li
k
el
ihoo
d
of
f
orwa
r
ding
is
dynam
ic
al
l
y
determ
ined
to
de
pend
on
t
wo
factors;
the
fir
st
is
the
den
sit
y
of
nei
ghbori
ng
l
ocal
node
s
and
the
seco
nd
is
the
total
num
ber
of
it
s
nei
ghbors
protect
ed
by
broa
dcast.
Using
the
DPR
schem
e,
networ
k
perform
ance is h
ig
he
r
t
ha
n ot
her p
ro
t
oco
ls
.
In
[
14
]
,
t
he
auth
or
s
intr
od
uced
a
new
pro
bab
il
ist
ic
broa
dcasti
ng
sc
hem
e
fo
r
M
AN
E
T
s
th
at
ov
e
rc
om
es
the
co
ns
trai
nts
of
existi
ng
br
oadcast
ing
sc
hem
es.
T
hroug
h
de
ta
il
ed
si
m
ulatio
ns,
it
is
sho
wn
that
the
ne
wly
pro
po
s
ed
sc
hem
es
in
diff
e
re
nt
operati
ng
co
ndit
ion
s
a
nd
sce
na
rios
out
perfor
m
pr
evio
us
broad
ca
st
schem
es.
Un
li
ke
previ
ou
s
w
orks,
the
p
r
op
os
e
d
broa
dcast
schem
e
strat
e
gy
is
based
on
inf
or
m
at
ion
fro
m
the
node
vel
oc
it
y
vector
to
a
dju
st
the
retra
ns
m
issi
on
pro
bab
il
it
y
and
ca
te
gorize
the
re
li
abili
ty
of
the
node
s
accor
dingly
.
T
he
aut
hors
ap
plied
this
vel
oc
it
y
vector
an
d
eval
uated
it
s
eff
ic
ie
ncy
with
res
pect
to
va
rio
us
i
m
po
rtant
m
et
r
ic
s
su
ch
as
li
nk
sta
bili
ty
and
ov
e
rh
ea
d
RR
EQ
pac
kets.
T
he
key
be
nef
it
of
this
sc
hem
e
is
to
avo
i
d
the
tra
diti
on
al
AOD
V'
s r
oute
r
e
-
disco
ve
ry pro
ce
ss,
e
s
pecial
ly
at high m
ob
il
it
y nod
es.
A
novel
sc
hem
e
for
C
hannel
Ad
a
ptiv
e
Proba
bili
sti
c
Broadcast
(CAP
B
)
is
pro
posed
i
n
[
15]
,
w
hich
is
i
m
ple
m
ented
in
the
AOD
V
routin
g
prot
oc
ol
to
re
place
pure
fl
ood
-
bas
ed
br
oad
cast
.
Unde
r
the
sc
hem
e,
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N: 20
88
-
8708
Dyna
mic r
ou
ti
ng d
isc
overy
s
chem
e
for
h
i
gh mob
il
it
y in
mobile
ad hoc
wi
rel
ess n
et
work
s
(
Ha
ider
Ala
ni
)
3705
the
prob
a
bili
ty
of
retra
ns
m
i
tt
i
ng
RR
EQ
pa
ck
et
s
is
re
gu
la
te
d
by
t
he
c
urre
nt
sign
al
i
nterf
e
re
nce
plu
s
no
ise
rati
o
(S
I
NR)
a
nd
ne
ighborin
g
de
ns
it
y
of
t
he
node
s.
C
om
par
ed
to
t
he
oth
e
r
th
ree
br
oad
c
ast
ing
sc
hem
e
s
w
h
ic
h
include
AOD
V
,
fixe
d
pro
ba
bili
sti
c
sche
m
e
AOD
V
-
P
,
an
d
the
dynam
ic
no
ise
-
dep
e
ndent
prob
a
bili
sti
c
schem
e
(DNDP)
, th
e
CAPB sc
hem
e p
erfor
m
s b
et
te
r i
n
te
rm
s o
f o
ve
rh
ea
d routi
ng, t
hro
ughput a
nd
end
-
to
-
en
d del
ay
.
Energy
E
ff
ic
ie
nt
Nei
ghbor
Cov
e
ra
ge
protoc
ol
(E
ENCP
)
w
as
pro
pose
d
by
[
16
]
to
r
oute
disco
ver
y
i
n
MANET
.
U
nc
ov
e
re
d
nei
ghbor
a
nd
nei
ghbor
hood
node
densi
ty
are
tw
o
m
et
ric
factor
s
co
ns
i
der
e
d
unde
r
the
EENC
P
pr
oto
c
ol
to
re
duc
e
the
f
orwardi
ng
of
the
RR
E
Q
pac
ket.
T
he
forw
a
r
ding
of
the
RR
EQ
pac
ket
by
the
node
reli
es
on
it
s
batte
ry
li
fe
and
nei
ghbor
hood
de
ns
it
y,
wh
ic
h
forwards
wh
e
n
the
r
e
is
adeq
uate
e
nergy
and
w
hen
the
node
densi
ty
su
r
passes
a
pa
rtic
ular
th
reshol
d.
T
he
EE
NC
P
prot
oco
l
ca
n
lim
i
t
reb
r
oa
dc
ast
ing
with
pac
kets
m
or
e
than
AODV
a
nd
Neig
hbor
Co
ver
a
ge
-
base
d
Prob
a
bili
sti
c
Re
br
oad
c
ast
(N
CPR
)
pr
oto
c
ols
.
A
ne
w
pro
ba
bili
sti
c
vector
ve
locit
y
schem
e
was
dev
el
op
e
d
in
[
14]
for
rou
te
disco
ver
y
i
n
MANE
T.
Ba
s
ed
on
the
AOD
V
r
outi
ng
protoc
ol,
this
schem
e
cat
egorizes
al
l
m
ob
il
e
nodes
into
reli
able
nodes
a
nd
unr
el
ia
ble
nodes
in
te
rm
s
of
se
nder
a
nd
r
ecei
ver
no
d
e
ve
locit
y,
and
as
sign
s
a h
ig
h
ret
ran
sm
issi
on
pr
ob
a
bili
ty
fo
r
re
li
able
nodes
a
nd
a
l
ow
retra
ns
m
is
sion
pro
ba
bili
t
y
for
un
reli
able
nodes
.
T
his
ty
pe
of
schem
e
helps
to
dis
cov
e
r
the
ste
adiest
and
m
os
t
eff
i
ci
ent
routes,
there
by
enhan
ci
ng
r
oute
di
sco
ver
y
ef
fici
ency.
This
sc
hem
e
dem
on
strat
es
the
RR
EQ
pac
ket'
s
su
per
i
or
it
y
in
te
r
m
s
of
bo
t
h
over
hea
d
and
co
nnect
ion
sta
bili
ty
.
A
novel
al
gorithm
,
Re
liabili
ty
Fact
or
-
base
d
Ro
uting
Pr
ot
oc
ol
(RFB
RP)
,
was
intr
oduce
d
by
[17]
to
disc
ov
e
r
a
r
el
ia
ble
route
in
MA
N
ET.
It
is
desig
ned
base
d
on
r
eact
ive
routing
protoc
ols
su
c
h
as
AODV
a
nd
DS
R.
It
is
pos
it
ion
ed
on
the
reli
abili
ty
factor
(RF
)
betwee
n
t
he
in
te
rm
ediat
e
no
de
s
to
sel
ect
a
dep
e
ndable
r
o
ute.
RF
is
a
pr
i
m
ary
m
et
ric
that
lim
it
s
the
r
ou
ti
ng
fail
ur
e
a
nd
t
he
nu
m
ber
of
r
oute
disc
ov
e
ry
re
qu
e
sts,
reducin
g
the
ove
rh
ea
d
cost
s
of
eac
h
node
i
n
the
r
ou
te
dis
cov
e
ry
pr
ocess
.
I
n
te
rm
s
of
over
hea
d,
delive
ry
of
a
pack
et
,
and
en
d
-
to
-
e
nd
delay
,
the RFBR
P al
g
or
it
hm
p
er
form
s b
et
te
r
t
han AOD
V
a
nd sem
i
-
proacti
ve
AO
DV (SP
-
A
ODV)
prot
oco
ls
.
Hanji
a
nd
S
hetta
r
[
18]
intr
oduce
d
an
Im
pr
oved
A
ODV
(I
-
A
O
D
V)
proto
col
i
n
MA
NET
.
The
route
discov
e
ry
in
t
he
I
-
A
ODV
pro
tocol
was
bas
ed
on
the
l
oc
a
ti
on
a
nd
ene
rg
y
of
the
nodes
i
n
the
netw
ork.
T
he
source
node
sel
ect
s
the
in
te
rm
ediat
e
no
de
locat
ed
in
the
com
m
un
ic
at
i
on
ra
nge,
w
hic
h
has
le
ss
distance
than
the
oth
e
r
interm
ediat
e
no
des
to
the
de
sti
nation
node
a
nd
has
ene
rg
y
gr
eat
er
t
han
th
e
oth
e
r
interm
ediat
e
n
od
e
s
to
i
ncr
ea
se
the
li
feti
m
e
of
t
he
path.
The
I
-
AOD
V
protoc
ol
pr
ov
i
des
e
ff
ic
ie
nt
pa
cket
transm
issi
on
,
e
lim
inate
s
ov
e
r
head
a
nd
im
pr
ov
e
s
the
li
feti
m
e
of
the
pat
h
that
le
ads
to
a
sta
ble
path
c
om
par
ed
to the A
ODV p
ro
t
oco
l.
Kok,
et
al
.
[19]
propose
d
an
eff
ect
ive
a
d
-
ho
c
wireless
broa
dcasti
ng
sc
hem
e,
kn
own
as
Im
pr
ov
e
d
Partia
l
D
om
in
ant
P
runin
g
(
IP
D
P),
base
d
on
Pa
rtia
l
Do
m
inant
P
runin
g
(PDP).
I
t
us
es
nei
ghbor
hood
inf
or
m
at
ion
to
reduce
re
dund
ancy
of
broad
c
ast
ing
an
d
ens
ur
i
ng
that
al
l
no
de
s
in
a
netw
ork
are
reac
ha
ble
to
pack
et
s
.
To
c
om
par
e
their
effi
ci
ency
in
two
sam
ple
set
s,
IP
DP
a
nd
P
DP
w
ere
buil
t.
The
f
irst
sa
m
ple
in
sta
ti
c
scen
ari
os
w
hile
the
seco
nd
se
t
of
data
in
m
ob
il
e
scenari
os
exam
ines
the
schem
es.
The
I
PD
P
is
ve
ry
ef
fecti
ve
in
sta
ti
c
scena
rios
in
re
du
ci
ng
the
redu
nd
a
ncy
of
broa
dc
ast
wh
il
e
m
ain
ta
inin
g
a
vaila
bili
ty
to
al
l
no
des
in
a
netw
ork.
From
the
m
ob
il
e
scena
rio,
al
l
var
ia
nts
of
I
PD
P
pe
rfor
m
ed
at
lo
w
to
hi
gh
node
m
ove
m
ent
sp
ee
d
cl
os
e
P
DP
in
te
rm
s
of
reac
hab
il
it
y.
A
ne
w
A
ODV
(NAOD
V)
in
MANET
was
su
ggest
e
d
by
[
20
]
in
the
r
ou
ti
ng
protoc
ol.
It
re
sol
ves
the
AOD
V
r
ou
ti
ng
pro
tocol
issue
s
by
accounti
ng
local
congesti
on
of
the
netw
ork
w
hile
sel
ect
ing
the
r
ou
te
.
Ba
se
d
on
netw
ork
congesti
on,
the
nu
m
ber
of
ho
ps
is
identifie
d
wh
il
e
s
el
ect
ing
the
ne
xt
node
t
o
li
m
it
the
su
r
vi
va
l
tim
e
of
fo
r
ward
a
nd
rev
e
rse
r
oute
s,
an
d
reduce
the
nu
m
ber
of
RR
EQ
pac
ket
r
et
ran
sm
issi
on
s,
reducin
g
over
head
c
os
ts.
Mo
hs
in
,
et
al
.
[2
1]
exam
ine
the
eff
ect
s
of
the
de
ns
it
y
and
m
ob
il
i
ty
n
od
e
s
of
di
ff
e
re
nt
m
arit
i
m
e
tr
aff
ic
m
od
el
based
on
the
cap
abili
ti
es
of
thr
ee
MANET
r
outi
ng
protoc
ols
that
com
pr
ise
A
ODV,
A
OMD
V
a
nd
D
SDV
.
Va
rio
us
node
de
ns
it
ie
s
an
d
m
ob
il
it
y
int
erpre
t
the
traff
ic
m
od
el
s
that
will
be
m
ai
nly
fo
un
d
i
n
the
m
ari
ne
en
vir
on
m
en
t.
The
pe
rform
ance
assess
m
ent
of
the MA
NET
pr
oto
c
ols is
j
ud
ge
d
in
term
s o
f packet
delive
ry r
at
io.
Zo
hr
a
a
nd
Ra
hm
an
[22]
int
rod
uced
t
he
dy
nam
ic
prob
a
bili
sti
c
br
oa
dc
ast
al
gorithm
in
MA
NET.
This
al
gori
th
m
est
i
m
at
es
t
he
re
broa
dcast
prob
a
bili
ty
by
util
i
z
ing
the
forw
a
rd
i
ng
pro
ba
bili
ty
of
a
node
.
The
al
gorithm
introdu
ce
d
a
log
ic
al
m
od
el
to
identify
the
exp
ect
ed
nu
m
ber
of
forw
a
rd
i
ng
nodes
th
at
are
essenti
al
to
ac
com
plish
a
br
oad
ca
st
in
a
s
el
f
-
pr
un
i
ng
al
gorithm
.
It
em
plo
ys
neig
hbor
kn
ow
le
dg
e
to
li
m
it
the
unnecessa
ry
retran
sm
issi
on
s
in
MA
N
ET.
It
ex
hi
bits
su
pe
rio
r
perform
ance
co
m
par
e
d
to
the
sta
ti
c
pro
bab
il
ist
ic
al
gorithm
and
the
prob
a
bili
sti
c
ad
j
us
t
broa
dc
ast
al
gorithm
.
A
dynam
ic
par
ti
ti
on
i
ng
sc
hem
e
(D
P
S)
dep
e
nde
nt
on
nodal
ve
hicle
s
was
sug
gested
by
Ra
ye
ni,
et
al
.
[23]
.
The
DPS
m
od
el
com
pu
te
s
par
ti
ti
on
siz
ing
an
d
qua
ntit
ie
s,
after
wh
ic
h
it
com
pu
te
s
a
siz
e
and
br
oa
dcast
tim
et
a
ble
for
each
pa
rtit
ion
.
DPS
sc
hem
e
perform
s
reli
ably
an
d
ti
m
e
-
e
ff
ic
ie
ntly
in
a
n
em
erg
ency
br
oad
ca
st
pro
gra
m
acro
ss
m
ultip
le
ho
ps
in
ei
ther
l
ow
or
high
-
de
ns
it
y
traff
ic
sta
te
s.
This
m
et
ho
d
su
it
ably
app
li
es
to
var
ie
d
s
cenari
os
feat
uri
ng
sca
rce,
m
i
d
an
d
high
-
de
ns
it
y
traf
fic.
Sim
ulati
on
s
c
onfirm
that
DP
S
pe
rfor
m
s
ou
tst
and
i
ng
ly
with
low
la
te
ncie
s
an
d
high
reli
abili
ty.
Kim
ur
a,
et
al.
[24]
pr
ese
nted
a
den
sit
y
-
a
w
are
pro
bab
il
ist
ic
ro
utin
g
m
od
el
featur
in
g
ad
aptive
pro
bab
il
ist
ic
fo
r
wardin
g
ope
rati
on,
w
he
rein
the
f
orwardi
ng
pr
ob
a
bili
ti
es
decli
ne
as
m
essages
a
re
re
di
rected.
Su
c
h
de
ns
it
y
-
s
ensiti
ve
pr
ob
a
bili
sti
c
ro
utin
g
m
od
el
s
c
on
tr
ol
the
s
pee
ds
of
ci
rcu
la
ti
ng
cop
ie
s
of
m
ess
ages
i
n
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
10
, No
.
4
,
A
ugus
t
2020
:
3702
-
3714
3706
high
-
de
ns
it
y
areas.
The
res
ult
s
of
the
pe
rfo
r
m
ance
a
ssessm
ent
pr
esent
t
he
ada
ptive
pr
ob
a
bili
sti
c
fo
r
wardin
g
m
od
el
in
a
high
-
de
ns
it
y
noda
l
reg
io
n.
The
On
-
de
m
and
,
T
ree
-
base
d
Ro
ut
ing
P
r
oto
c
ol
(
OTRP)
was
sugg
e
ste
d
by
Al
Aam
ri,
e
t
al
.
[2
5]
fo
r
M
AN
E
T.
The
O
TRP
with
AODV
pr
oto
c
ol
al
lowe
d
it
to
sca
le
up
and
al
s
o
le
ssen
ov
e
r
head
s
i
n
MANET
r
ou
ti
ng
disco
ve
ry,
by
pairi
ng
t
he
con
ce
pt
of
hop
-
to
-
ho
p
r
ou
t
ed
tra
ns
m
issi
o
ns
with
the
r
ou
ti
ng
discov
e
ry
al
gorithm
s
kn
own
a
s
Tr
ee
-
base
d
O
pti
m
iz
ed
Floodin
g
(T
OF).
T
he
pri
m
ary
el
e
m
ents
wh
ic
h
gove
r
n
OTRP
pe
rfo
r
m
ance
wer
e
hy
po
theti
cal
ly
exam
ined
and
assessed
in
te
r
m
s
of
nodal
branc
h
counts
an
d
lo
cat
ion
s,
a
nd
c
ounts
of
RR
EQ
r
ou
ti
ng
ex
pi
rati
on
s.
At
va
rio
us
le
vels
of
nodal
de
ns
i
ty
and
mo
bility
,
OTR
P
wa
s
c
on
t
ras
te
d
to
A
O
DV,
OL
SR,
a
nd
dynam
ic
MANET
O
n
-
dem
a
nd
(
D
YMO
)
r
ou
ti
ng
protoc
ols.
T
he
fin
dings
in
dicat
e
that
OTRP
m
echan
ism
s
per
f
or
m
ou
tst
an
dingly
with
m
uch
-
le
sse
ned
ne
twork
ov
e
r
head an
d d
ecl
ining
t
raffic
loads.
Zha
ng,
et
al
.
[
26
]
s
uggested
Neig
hbori
ng
Cov
e
ra
ge
-
der
i
ve
d
Pro
bab
il
ist
ic
Re
broad
ca
sti
ng
(
NCPR)
protoc
ols
for
MA
NET.
T
he
m
et
ho
d
dy
nam
ic
al
ly
talli
es
reb
r
oa
dc
ast
ing
delay
s
an
d
re
broadca
sti
ng
pro
bab
il
it
ie
s.
Re
broad
ca
sti
ng
delay
s
are
us
e
d
to
res
ol
ve
the
f
orwardi
ng
or
der
s
an
d
to
e
ff
ic
ie
ntly
util
i
z
e
neig
hbori
ng
co
ver
a
ge
in
form
at
ion
to
de
rive
fu
rt
her
a
nd
m
or
e
e
xa
ct
co
ve
rag
e
rati
os
.
Co
nn
ect
io
n
fact
ors
are
e
m
plo
ye
d
in
NCPR
to
dete
rm
ine
reb
ro
a
dc
ast
ing
pro
ba
bi
li
t
ie
s,
to
m
ai
n
ta
in
networ
ked
connecti
vity
,
and
to
decr
ease
incid
ences
of
redu
ndantly
retra
nsm
itted
traff
ic
.
Re
search
on
r
ou
ti
ng
request
s
in
r
ou
ti
ng
di
sco
ver
y
me
tho
ds
c
on
t
r
ast
ed
NCPR
with
Dynam
ic
al
ly
Pr
ob
a
bili
sti
c
Rou
te
discov
e
ry
protoc
ol
s
(D
PR
),
an
d
AOD
V
protoc
ols.
NC
PR
ge
ner
at
es
le
ss
traff
ic
re
broa
dcasts
tha
n
fl
ooding
protoc
ols
as
a
c
on
s
eq
ue
nce
of
few
e
r
redu
nd
a
ntly
reb
r
oa
dcast
traffi
c,
wh
ic
h
decre
ases
incide
nc
es
of
pack
et
c
olli
sion
s
a
nd
netw
ork
co
nte
ntion,
raises the
prop
or
ti
on
of p
ac
ke
t deli
ver
ie
s
, a
nd
dim
inishes a
ver
a
ge
e
nd
-
to
-
end lat
encies.
3.
PROP
OSE
D
AODV
V
EL
O
CITY A
N
D D
YNA
MIC
(
A
ODV
-
VD)
SCHE
ME
The
desi
gn
a
nd
i
m
ple
m
entation
of
the
pr
opose
d
AODV
-
VD
al
go
rithm
is
based
on
the
li
te
ratur
e
pro
po
se
d
A
dv
a
nce V
el
ocity
A
war
e
Pro
bab
il
i
sti
c
(
A
VAP
)
-
A
ODV
sc
hem
e
by
[
14]
,
A
O
D
V
-
VD
sc
hem
e
m
akes
us
e
of
pr
ob
a
bi
li
sti
c
velocit
y
vector
t
o
init
ia
te
ro
ute
disc
ov
e
ry.
Eac
h
m
ob
il
e
no
de
i
s
cl
assifi
ed
as
ei
ther
a
Re
li
able
No
de
(
RN
)
or
an
Un
Re
li
able
N
ode
(
URN
)
.
Thi
s
cl
assifi
cat
ion
is
based
on
th
e
send
e
r
an
d
r
ecei
ver
nodes
’
velocit
y
vecto
r.
Re
li
able
no
des
a
re
assig
ned
a
hig
h
re
broa
dcast
pro
ba
bili
ty
value,
wh
il
e
unr
el
ia
ble
nodes
are
give
n
a
low
re
br
oad
ca
st
proba
bili
ty
value.
The
schem
e
pr
op
os
ed
by
K
anak
a
ris,
et
al
.
[27]
,
is
consi
dered
a
dynam
ic
den
si
ty
-
dr
ive
n
r
ou
te
request for
wardin
g
sc
hem
e.
More
ov
e
r,
t
his
m
ade
it
,
po
ssi
ble
f
or
each
no
de
to
use
it
s
neighb
orhoo
d
de
ns
it
y
as
a
basis
f
or
forw
a
rd
i
n
g
t
he
RR
EQ
m
essage.
The
nei
ghbor
hood
’
s
node de
ns
it
y pl
ay
s a v
it
al
role
in d
et
e
rm
ining
the r
e
broa
dcast
proba
bili
ty
[15
]
.
The
pro
po
se
d
AOD
V
-
VD
s
chem
e
transm
i
t
s
an
RR
EQ
pa
cket
w
hile
di
sco
ver
in
g
a
route
bet
wee
n
the
no
des,
A
ODV
-
V
D
sc
hem
e
us
es
the
node
’s
dynam
ic
densi
ty
and
velocit
y
vect
or.
It
ai
m
s
to
reduc
e
the
un
necessa
r
y
broa
dcasti
ng
of
r
oute
co
ntr
ol
pac
kets
a
nd
ident
ify
t
he
r
el
ia
ble
interm
e
diate
node
s
e
xi
sti
ng
betwee
n
the
source
an
d
destin
at
ion
nodes
.
A
ODV
-
V
D
sc
he
m
e
m
akes
us
e
of
seve
ral
ad
ju
sta
ble
pa
ram
eter
s
t
o
determ
ine
wh
e
ther
retra
ns
m
is
sion
is
per
m
it
t
ed
or
f
orbi
dd
e
n.
The
in
f
or
m
a
ti
on
obta
ined
f
ro
m
the
no
de
de
ns
it
y
w
it
hin
th
e
nei
ghbor
hood
ca
n
be
util
i
z
ed
to
rebro
a
dcast
R
REQ,
sel
ect
ive
ly
.
If
the
neig
hbour
hood
has
m
any
nodes
,
then
th
e
aim
wo
uld
be
to
decr
ease
t
he
pro
ba
bili
ty
of
ret
ran
sm
issi
on
to
a
void
possible
re
dund
ancy.
In
this
case
,
no
des
t
hat
are
on
ly
on
e
ho
p
a
w
ay
are
def
i
ned
as
bei
ng
“a
dj
a
cent”.
T
he
refo
re,
if
a
node
se
nd
s
an
RR
EQ
pa
cket,
retra
ns
m
issi
on
ca
n
on
ly
be
perform
ed
by
a
subset
of
no
des
in
it
s
nei
ghbo
rho
od.
T
he
num
ber
of
nodes
al
l
owed
t
o
pe
r
for
m
retran
sm
issio
n
ca
n
var
y
dy
nam
ic
al
ly
de
pendin
g
on
th
e
neig
hborh
oo
d
no
de.
Table
1
s
hows
the
par
am
et
ers
us
e
d
in
the
A
O
DV
-
V
D
sc
hem
e.
The
A
O
DV
-
VD
sc
hem
e
flow
cha
rt
can b
e
seen
in
Fig
ur
e
3
W
hen
a
ny
no
de
(
Fi
=
1,2,3,..,
n
)
receives
th
e
RR
EQ
pac
ket,
t
he
pr
opos
e
d
A
ODV
-
VD
sc
he
m
e
can
be per
form
ed
and the
pac
ket
is
proce
ssed
as
fo
ll
ows
Fig
ur
e
3.
Table
1
.
A
ODV
-
VD
pa
ram
eter
s
s
pecifica
ti
on
Para
m
eters
Descripti
o
n
n
to
tal nu
m
b
e
r
o
f
no
d
es p
resent
ed
in
the n
etwo
rk
Fi
set o
f
no
d
es th
at h
av
e r
eceived
the R
REQ
m
ess
ag
e
i
rang
e f
ro
m
1 to
n
n
u
m
b
er
of
no
d
es
ƞ
β
i
n
u
m
b
er
of
no
d
es ad
jacent to
no
d
e i
P
i
p
acket f
o
rwar
d
in
g
pro
b
ab
ility
b
ased
on
βi
d
n
o
d
e adjacen
cy
C
f
co
n
trol f
acto
r
b
etween (0,1
)
ad
ju
sted
u
sin
g
Pi
R
rand
o
m
nu
m
b
e
r
in
th
e r
an
g
e [
0
,
1
0
0
]
θth
p
redef
in
ed
ang
le t
h
resh
o
ld
θ
Co
sin
e ang
le between
trans
m
itt
er
an
d
r
eceive
r
to
d
ecid
e if
the rec
eiv
er
is RN
or UR
N
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N: 20
88
-
8708
Dyna
mic r
ou
ti
ng d
isc
overy
s
chem
e
for
h
i
gh mob
il
it
y in
mobile
ad hoc
wi
rel
ess n
et
work
s
(
Ha
ider
Ala
ni
)
3707
Figure
3.
A
ODV
-
VD
sc
hem
e
flo
wch
a
rt
1.
Wh
e
n
a
s
ource
node
(S)
trie
s
to
disc
over
a
new
r
ou
te
,
it
s
velocit
y
vecto
r
is
placed
i
n
th
e
RR
EQ
pa
cke
t
head
e
r.
2.
Wh
e
n
this
RR
EQ
pack
et
is
obta
ined
by
a
ny
node
receive
r
(R)
wit
hin
the
broa
dcast
ra
nge
of
S,
the
c
os
i
ne
ang
le
θ is cal
c
ulate
d.
It the
n m
akes a
re
broa
dcast
decisi
on
us
in
g
t
he
f
ollo
wing
process:
-
If
the
node’s
cosine
a
ng
le
value
is
θ
>
θth,
then
t
he
receiv
er
node
is
cat
e
gorized
as
a
U
RN.
It
th
e
n
assigns a l
ow re
broa
dcast pr
obabili
ty
.
-
If
the
no
de’
s
c
os
ine
an
gle
val
ue
is
θ
<
θth,
t
hen
the
recei
ve
r
node
is
cat
egorized
as
RN.
Then
go
to
STEP
3.
3.
If
βi ≤
d t
he
n
Fo
r
wa
rd the R
REQ
pack
et
el
se
Ca
lc
ulate
f
orw
ard
i
ng proba
bili
ty
Pi at node Fi
Pi
=
100
β
i
∗
(
d
∗
C
f
;
For
0<C
f
≤1
4.
Gen
e
rati
on
of
a Ran
do
m
N
um
ber
(
R)
betw
een
(0,10
0)
5.
If
R
< Pi t
hen
Fo
r
wa
rd the R
REQ
pack
et
el
se
Ign
or
e a
nd
Dro
p
the
RR
EQ
p
a
cket
6.
En
d
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
10
, No
.
4
,
A
ugus
t
2020
:
3702
-
3714
3708
The
pro
po
s
ed
A
ODV
-
V
D
s
chem
e
fo
r
M
AN
E
T
ai
m
s
t
o
im
pr
ove
the
netw
ork
pe
rfor
m
ance
a
nd
reso
l
ve
the
iss
ue
of
t
he
f
requen
t
li
nk
br
ea
kag
e
.
T
he
sc
hem
e
sel
ect
s
the
reli
able
no
de
durin
g
t
he
r
ou
t
e
disco
ver
y
proc
ess
to
av
oid
t
he
li
nk
brea
k
an
d
el
i
m
inate
redunda
nt
retra
nsm
issi
on
to
achi
eve
the
lo
west
value
of
co
ngest
ion
, t
her
ef
ore,
redu
ce the
ov
e
r
hea
d
in
the
netw
or
k
[28]
.
4.
SIMULATI
O
N
EN
VI
R
O
N
MENT
We
us
e
d
NS
-
2.35
to
m
eas
ur
e
the
A
O
D
V
-
VD
sc
hem
e
'
s
per
form
ance.
The
node
s
are
m
ov
ing
accor
ding
to
t
he
ra
ndom
waypo
int
m
ob
il
ity
m
od
el
.
In
th
e
rando
m
waypo
i
nt
m
ob
il
ity
m
od
el
,
node
s
m
ov
e
fr
eel
y
an
d
ra
ndom
ly
without
bor
der
re
stric
ti
on
in
this
m
ob
il
it
y
m
o
de
l.
CB
R
traff
ic
is
gen
erat
ed
by
the
ap
plica
ti
on
la
ye
r
at
the
node.
F
or
e
ve
ry
node
,
the
t
ran
sm
issi
on
ra
ng
e
is
25
0
m
.
The
pa
us
e
ti
m
e
for
the
node
is
c
onsta
nt
at
0
t
o
a
ll
ow
no
de
m
ov
e
m
ent
at
al
l
t
i
m
es.
Netw
ork
scenari
o
no
de
sp
ee
d
was
sel
e
ct
ed
at
5,
10,
20,
30
,
40,
a
nd
50
m
/
s,
an
d
al
l
sce
nari
os
set
the
num
ber
of
no
des
at
50.
All
nodes
are
set
in
a
10
00
m
sq
ua
re a
rea
of
about
1000 m
. Tab
le
2 displa
ys t
he param
eter
s in
the
sim
ul
at
ion
.
Table
2
. Si
m
ul
at
ion
par
am
et
e
rs
Para
m
eters
Valu
e
Un
it
Si
m
u
latio
n
ti
m
e
300
S
Netwo
rk ar
ea size
1
0
0
0
×
1
0
0
0
M
Nu
m
b
e
r
o
f
no
d
es
20
-
100
No
d
es
Data pack
et siz
e
512
Bytes
Ban
d
wid
th
2
Mbp
s
No
d
es sp
eed
5
–
50
m
/
s
Pau
se ti
m
e
0
S
Data traf
f
ic
CBR
Perfo
rm
ance
m
et
rics
are
use
d
in
the
c
omparati
ve
st
ud
y
to
assess
the
pe
rfor
m
ance
of
the
pro
posed
AOD
V
-
VD
sc
hem
e.
The
fol
lowing
c
rite
ria
are
us
e
d
for
e
valuati
ng
t
he
pro
po
se
d
AOD
V
-
VD
sc
hem
e
in sce
nar
io
s.
a.
Av
e
ra
ge
e
nd
-
to
-
e
nd
(
E2E
)
d
e
la
y
The
a
ve
rag
e
e
nd
-
to
-
e
nd
delay
m
et
ric
is
the av
era
ge
tim
e
it
t
akes to
su
ccess
fu
ll
y rela
y t
he
data p
acke
t
from
the
so
ur
ce
thr
ough
th
e
netw
ork
t
o
destinat
io
n.
T
his
delay
incl
ud
e
s
m
any
sm
al
le
r
networ
k
delay
s,
includi
ng
al
l
pote
ntial
delay
s
caused
by
bu
f
fer
in
g
la
te
ncy
durin
g
r
ou
te
di
s
co
ver
y,
que
uin
g
at
router
int
erf
ace
qu
e
ue,
delay
s
in
MAC
ret
ran
s
m
issi
on
,
pr
op
a
gation
a
nd
tra
nsm
issi
on
tim
e.
The
a
ver
a
ge
da
ta
pack
et
dela
y
for
the E
2E
ca
n be
d
et
erm
ined us
ing
t
he follo
wi
ng for
m
ula:
2
=
∑
(
−
)
−
1
(
1
)
w
he
re
Ri
is
the
total
p
ac
ket r
e
cei
ved
,
Si
is
th
e total
p
ac
ket s
ent, a
nd
n
is t
he
num
ber
of
da
ta
p
ackets
.
b.
Av
e
ra
ge
th
rou
ghput
(
A
vg
-
T
hro
ughp
ut)
The
a
ver
a
ge
t
hro
ughput
m
etr
ic
is
the
ave
rag
e
of
receiv
ed
s
uccessful
data
pac
kets
t
o
the
t
otal
si
m
ulati
on
tim
e
per
io
d.
T
he
total
through
put
is
expresse
d
in
kilo
bits
pe
r
seco
nd
(kb
ps
)
,
an
d
m
eas
ur
es
the
routin
g
pr
oto
c
ol'
s
eff
ic
ie
ncy
and
reli
abili
ty
in
receivi
ng
da
ta
pack
et
s
t
hro
ugh
destinat
io
ns
.
T
he
form
ula
us
e
d
to calc
ulate
the
ave
rag
e
thro
ughp
ut is as
f
ollow
s:
ℎ
ℎ
=
∑
−
∗
8
1000
(2)
c.
Packet
deliver
y rati
o
(
PD
R
)
The
PD
R
m
et
ri
c sh
ows the total
n
um
ber
o
f
r
ecei
ved
d
at
a pa
ckets b
y desti
nations divi
de
d
by the total
nu
m
ber
of
data
pack
et
s
se
nt
by
the
so
urces. Th
is
m
et
ric
pr
esents
how
a p
r
oto
c
ol
su
ccess
f
ully
delivers
pa
ckets
from
the
so
urc
e
to
the
destin
at
ion
.
A
high
pack
et
del
iver
y
rati
o
ind
ic
at
es
good
re
su
lt
s
,
w
hich
repres
ent
the
wholene
ss
an
d
correct
ness
of
the
routin
g
pr
oto
c
ol.
T
he
pa
cket
delive
ry
r
at
io
is
com
pu
te
d
us
i
ng
t
he
f
ol
lowing
form
ula:
=
∑
∑
∗
100
(3)
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N: 20
88
-
8708
Dyna
mic r
ou
ti
ng d
isc
overy
s
chem
e
for
h
i
gh mob
il
it
y in
mobile
ad hoc
wi
rel
ess n
et
work
s
(
Ha
ider
Ala
ni
)
3709
d.
Packet
ov
e
r
hea
d rati
o
(
PO
R
)
Rou
ti
ng
ove
rhead
rati
o
m
et
ri
c
is
the
rati
o
of
the
total
nu
m
ber
of
routin
g
pac
kets
sent
to
the
total
nu
m
ber
o
f
rout
ing
p
ackets
se
nt
and
the d
at
a p
ackets
se
nt.
T
his
m
et
ric
giv
es
an
idea
a
bout
the
extra b
a
ndwidth
the ove
rh
ea
d u
ses to ha
nd
le
da
ta
traff
ic
.
T
he
over
hea
d
r
outi
ng is cal
culat
e
d usin
g form
ulae b
el
ow:
=
+
∗
100
(4)
e.
Av
e
ra
ge of
c
ol
li
sion
r
at
e
(
ACR)
The
ACR
sig
ni
fies
the
total
nu
m
ber
of
RR
EQ
pac
kets
th
at
hav
e
bee
n
dro
pped
a
nd
w
hich
di
d
not
s
uccess
fu
ll
y
ar
rive
at
the
no
des
du
rin
g
the
si
m
ulati
on
pe
rio
d.
T
he
f
or
m
ula
belo
w
s
hows
ho
w
the
A
CR
is
cal
culat
ed:
(
/
)
=
∑
(5)
f.
Link st
abili
ty
Link
sta
bili
ty
per
ta
ins
t
o
the
total
nu
m
ber
of
bro
ken
route
s
that
are
fou
nd
duri
ng
data
transm
issi
on
betwee
n
no
des
durin
g
the
sim
ula
ti
on
tim
e.
Rou
te
br
ea
k
m
easur
em
ents
are
do
ne
de
s
pite
var
yi
ng
ne
twork
densi
ti
es and nod
e
s
peed
s
. Ca
lc
ulati
on
of the
li
nk
sta
b
il
it
y fo
ll
ows the
f
ollow
i
ng
f
or
m
ula
:
Link
stabi
li
t
y
(l
in
k/
rout
e)
=
to
t
a
l nu
mbe
r of rou
t
e b
re
a
k
a
g
e
(6)
5.
RESU
LT
S
AND DI
SCUS
S
ION
S
5.1.
Results
b
as
ed
on
th
e
num
be
r of n
od
es
The
netw
ork
de
ns
it
y
is
con
si
der
e
d
to
be
a
s
ign
ific
a
nt
an
d
decisi
ve
par
am
et
er.
T
he
num
ber
of
no
des
unde
r
dif
fer
e
nt
network
de
nsi
ti
es
was
set
to
20,
40,
60,
80
,
a
nd
100
durin
g
the
sim
ula
ti
on
,
an
d
fo
r
al
l
the
netw
ork
pa
ram
et
ers,
ever
y
node
was
s
et
to
a
m
axi
m
um
sp
eed
of
20
m
/s.
Figu
re
4
show
s
the
aver
a
ge
thr
oughput
ve
r
su
s
the num
ber
of
n
odes
.
Fig
ure
4
s
hows
t
here
is
a
s
m
al
l
decr
ease
in
the
a
ve
rag
e
t
hroug
hput
of
the
pro
po
se
d
AOD
V
-
VD
sc
hem
e
wh
en
th
e
nu
m
ber
of
node
s
is
40
c
om
par
ed
to
the
m
od
ifie
d
AVA
P
-
A
O
DV
and
th
e
or
i
gin
a
l
AO
D
V
pr
otoc
ols.
I
n
a
de
nse
networ
k,
the
netw
ork
bec
om
es
m
or
e
sta
ble.
Ther
e
fore,
t
her
e
is
an
increase
in
t
he
ave
rag
e
th
r
oughput
in
a
de
ns
e
net
work
a
nd
c
onve
rg
e
nt
rati
o
in
dif
fer
e
nt
protoc
ols.
F
i
gure
4
al
so
s
hows
tha
t
the
AOD
V
-
V
D
has
ex
per
ie
nc
ed
a
sta
ndar
d
increase
in
ave
rag
e
th
rou
ghpu
t
with
t
he
dec
r
ease
in
inte
ns
it
y
of
dro
pp
i
ng
no
de
s.
T
his
ca
n
be
du
e
to
the
ra
ndom
dep
loym
ent
of
no
des
t
hat
change
f
ollow
i
ng
it
s
nu
m
ber
i
n
a
ne
twork
.
Figure
5
dem
on
st
rates
the
var
yi
ng
a
ver
a
ge
de
la
y
of
E
2E
by
co
ntras
ti
ng
the
c
urre
nt
A
ODV
-
VD
schem
e
with
the
init
ia
l
A
O
DV
prot
oco
ls
and
A
VAP
-
A
ODV.
The
fig
ur
e
sho
ws
a
c
on
si
der
a
ble
dif
fer
e
nce
betwee
n
th
os
e
protoc
ols
in
the
ave
rag
e
E2
E
delay
.
Com
par
e
d
to
the
ori
gin
al
A
O
DV
and
AVAP
-
AODV
protoc
ols,
the
aver
a
ge
E
2E
la
te
ncy
in
the
pro
posed
AOD
V
-
VD
is
r
ed
uc
ed.
F
or
e
xam
ple,
in
a
de
ns
e
ne
twor
k
with
100
node
s,
the
ave
rag
e
E2E
d
el
ay
f
or
AOD
V,
A
V
A
P
-
A
O
DV
,
an
d
AOD
V
-
VD
is
0.118
98
s
,
0.0
7972
7s
,
and
0.0
640505
s,
res
pecti
vely
.
Thi
s
m
igh
t
be
the
cause
be
hin
d
t
he
dec
reas
e
in
pac
ket
dro
p
ove
r
the
networ
k.
This
occurs
be
cause
the
de
nse
netw
ork
,
the
routes
betwee
n
s
ource
a
nd
destinat
io
n
no
des
is
a
sta
ble
route.
And
the
sta
ble
route
yi
el
ds
to
le
sser
route
breakage
s
an
d
le
sser
re
-
e
sta
blis
hm
ent
that
ta
ke
le
ss
delay
t
i
m
e
and
an
inc
rease i
n
t
he
a
ver
a
ge
t
hroug
hput
,
a
s
observ
e
d
i
n
Fi
gur
e
4
.
The
P
OR
res
ults
show
n
in
F
igure
6
im
ply
that
the
pro
posed
AOD
V
-
VD
sc
hem
e
has
ena
bled
a
reducti
on
of
PO
R
necess
a
ry
to
al
locat
e
the
routes
in
var
io
us
s
our
ces
and
destinat
ions,
f
ollo
wed
by
the
A
VAP
-
A
ODV
a
nd
or
i
gin
al
A
O
DV
protoc
ols.
T
he
con
tr
ol
fact
or
in
the
fee
ding
net
work
with
pri
or
knowle
dge
of
t
he
no
de
ad
j
ace
ncy
dro
pped
i
n
diff
e
re
nt
node
s
has
play
ed
a
sign
ific
a
nt
r
ole
in
the
re
duct
io
n
of
the
P
OR.
As
obser
ve
d
in
F
i
gure
6
in
the
hi
gh
-
den
sit
y
are
a,
the
re
is
an
increase
in
P
O
R
beca
us
e
of
hav
i
ng
m
any
ro
utes
le
adin
g
to
the
gen
e
rati
on
of
m
or
e
con
trol
pack
et
s
.
AOD
V
-
VD
te
ch
nique
,
in
the
pro
po
s
e
d
schem
e,
gen
e
r
at
es f
ewe
r
c
ontrol
pac
kets.
Figure
7
il
lustrate
s
the
e
ff
ec
ts
of
t
he
di
ff
e
ren
t
protoc
ols
on
the
ACR
com
par
iso
n.
The
resu
lt
s
ind
ic
at
e
that
w
hen
m
ov
in
g
f
r
om
a
low
-
de
nsi
ty
reg
ion
to
a
high
-
de
ns
it
y
area,
the
ACR
in
creases
dram
atical
ly
.
This
is
becau
s
e
as
the
nu
m
ber
of
nodes
in
creases,
s
o
do
es
the
nu
m
ber
of
possible
f
orwa
rd
e
d
no
des
.
W
he
n
the
num
ber
of
node
s
incre
ased
pr
ogressi
vely
,
the
pro
po
s
ed
AOD
V
-
VD
sc
hem
e
l
ed
to
le
ss
c
olli
sion.
The
res
ult
has
est
ablished
t
he
eff
ic
ie
ncy
of
dy
nam
ic
den
sit
y
-
dri
ve
n
r
ou
te
r
equ
e
sts
to
decre
ase
the
colli
sion
by
su
pp
or
ti
ng
a
peer
-
to
-
pee
r
c
onnecti
on
between
te
rm
inals
.
But
AVAP
-
A
ODV
a
nd
the
ori
gin
al
A
O
D
V
app
ea
re
d
to
po
se
appr
ox
im
ately
si
m
il
ar
ACR,
wh
ic
h
is
great
er
than
the
pro
po
se
d
AODV
-
V
D.
A
dd
it
ion
al
ly
,
al
lowing th
e lo
west am
ou
nt
of A
CR
i
n
t
he n
et
work w
ould
r
es
ult i
n
le
ss
po
wer co
nsum
pti
on.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
10
, No
.
4
,
A
ugus
t
2020
:
3702
-
3714
3710
Figure
8
sho
w
s
that
the
pro
pose
d
A
ODV
-
V
D
le
d
to
a
reduced
num
ber
of
li
nk
br
ea
ka
ge
s
com
par
ed
to
the
A
VAP
-
AOD
V
an
d
th
e
AODV
proto
cols.
T
he
num
ber
of
bro
ken
l
ink
s
decr
ease
s
wh
e
n
the
nu
m
ber
of
nodes
i
ncr
ease
s
beca
us
e
the
netw
ork
has
t
he
pr
op
e
ns
it
y
to
be
sta
ble
in
a
de
ns
e
netw
ork.
As
obser
ved
i
n
F
igure
8,
the
AOD
V
-
VD
sc
hem
e
has
le
ss
er
brok
e
n
li
nk
s
in
dif
fer
e
nt
nodes
de
ns
it
ie
s,
due
to
t
he
t
echn
i
qu
e
us
e
d
to
sel
ect
the
reli
able
neig
hbor
node
.
This
stu
dy
c
le
arly
sh
ows
tha
t
offer
i
ng
a
protoc
ol
to
r
educ
e
the
pr
ob
a
bili
ty
of
r
oute
break
age
would
ena
ble
en
ha
ncem
e
nt
of
the
r
oute
eff
ic
ie
ncy,
w
hich
helps
t
o
c
onsu
m
e
le
ss pow
e
r
a
nd
g
e
ner
at
es
fe
w
er c
on
tr
ol
pac
ke
ts use
d
in
the
netw
ork.
Figure
4.
A
vg
t
hro
ughput
v
s
num
ber
of
node
s
Figure
5.
A
vg
E2E dela
y
v
s
num
ber
of
node
s
Figure
6. P
OR
vs
nu
m
ber
of
node
s
Figure
7
.
ACR
vs
num
ber
of
nodes
Figure
8
.
Lin
k st
abili
ty
vs
the
nu
m
ber
of no
de
s
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N: 20
88
-
8708
Dyna
mic r
ou
ti
ng d
isc
overy
s
chem
e
for
h
i
gh mob
il
it
y in
mobile
ad hoc
wi
rel
ess n
et
work
s
(
Ha
ider
Ala
ni
)
3711
5
.2
.
Results
b
as
ed
on
th
e
speed
of nodes
Figure
9
sho
ws
the
va
ryi
ng
av
erag
e
t
hro
ughp
ut
by
c
om
par
ing
the p
r
opose
d
AOD
V
-
VD
sc
hem
e
with
the
ori
gin
al
AOD
V
an
d
t
he
A
V
AP
-
A
O
DV
prot
oco
ls.
Figure
9
s
hows
that
the
pro
po
se
d
sc
he
m
e
has
apprecia
bly
outpaced
both
A
VAP
-
AOD
V
a
nd
the
earli
er
AOD
V
,
as
e
vid
ence
d
i
n
t
he
data,
par
ti
cula
r
ly
as
node
sp
e
ed
rati
os
rise
pro
gr
e
s
sively
.
This
is
du
e
to
the p
r
op
os
e
d
schem
e
s
el
ect
ing
the
reli
able
node
an
d
le
sser
con
t
ro
ls
pac
ket
leadin
g
to
the
trans
fer
of m
or
e d
at
a
pac
kets.
Figure
10
il
lustrate
s
the
av
erag
e
E
2E
d
el
ay
of
pac
ket
data
trans
fer
s
for
the
r
ou
ti
ng
disc
ov
e
r
y
protoc
ol,
acco
r
ding
to
the
vari
ou
s
no
de
sp
ee
ds
.
A
s
node
s
peed
inc
rease
s
,
the
aver
a
ge
E2E
d
el
ay
of
pack
e
t
data
tra
ns
fe
rs
increases
pr
opor
ti
onat
el
y.
T
his
occurs
be
cause
a
s
node
sp
ee
d
i
ncr
ea
s
es
al
ong
the
r
ou
te
s
betwee
n
sou
rc
e
and
destinat
i
on
nodes
,
rec
urre
nt
breaka
ge
s
and
re
-
e
sta
blishm
ent
in
the
li
nk
occur.
Re
s
ults
for
the
A
ODV
-
V
D
sc
hem
e
sh
ow
that
it
yi
el
ds
co
ns
ide
rab
ly
few
e
r
delay
s
in
com
par
iso
n
to
A
VAP
-
A
O
DV
a
nd
the earli
er
AO
DV.
Figure
9.
A
vg
t
hro
ughput
v
s
s
peed o
f n
od
es
Figure
10.
A
vg
E2E
d
el
ay
v
s
sp
ee
d of
node
s
Figure
11
sho
ws
the
three
prot
oco
ls
i
n
c
om
par
ison
acc
ordi
ng
t
o
the
P
OR
an
d
t
he
diff
e
ren
t
s
pee
ds
.
The
ac
quired
r
esults
in
dicat
ed
a
si
gn
i
ficant
disp
a
rity
in
P
OR,
s
howi
ng
it
s
lowest
pea
ks
in
t
ho
se
case
s
w
here
the
pro
po
s
ed
AODV
-
V
D
schem
e
was
run
ning
at
diff
e
re
nt
spe
ed
ranges.
Co
m
par
ed
with
that
of
AVAP
-
A
O
DV
and
the
ea
rlie
r
A
O
DV,
the
new
schem
e
pro
vid
e
d
few
e
r
PO
R.
T
his
is
due
to
t
he
A
O
D
V
-
VD
schem
e
'
s ab
il
ity t
o
co
ntr
ol loc
al
area
den
sit
ie
s whil
e RR
EQ i
s selec
ti
vely
r
ebroa
dcast.
Figure
12
sho
ws
com
par
at
iv
e
resu
lt
s
for
the
ACR
of
packet
transf
ers
acc
ordin
g
to
the
node
s
peeds.
The
re
su
lt
s
ac
qu
i
red
rev
eal
t
hat
the
earli
er
AOD
V
prot
oc
ol
ge
ner
at
e
d
hi
gh
e
r
rates
of
c
olli
sion
s,
f
ollo
wed
by
the
AVAP
-
A
O
DV
protoc
ol.
The
le
ast
valu
e
fo
r
ACR
wa
s
recorde
d
w
he
n
the
pro
pose
d
A
ODV
-
V
D
schem
e
was
ex
ecute
d
at
var
io
us
s
pe
eds.
T
his
is
a
tt
ribu
ta
ble
to
t
he
ef
fici
ency
of
t
he
A
ODV
-
V
D
m
echan
ism
in
determ
ining
th
e
sta
te
f
or
d
ec
r
easi
ng
the nodes
in
t
he
neig
hbor
hood,
es
pec
ia
ll
y
in
the
eve
nt
of h
ig
h
pro
ba
bili
ty
conditi
ons.
T
he
low
e
r n
um
ber
of
c
o
ll
isi
on
s
in
the
n
et
wor
k helps
to
c
onsum
e less powe
r.
Figure
11
. PO
R vs
s
pee
d of
nodes
Figure
12
.
AC
R
vs
s
pee
d of
nodes
Evaluation Warning : The document was created with Spire.PDF for Python.