TE
LKOM
NI
KA
Te
le
c
om
munica
tion,
C
omp
u
tin
g,
El
e
ctroni
cs and
Contr
ol
Vo
l.
18
,
No.
1
,
Febr
uar
y
2020
, pp.
3
7
~
50
IS
S
N: 16
93
-
6930, acc
red
it
ed
First G
ra
de by
Kem
enr
ist
ekd
i
kti, D
ec
ree
N
o: 21/E/
KP
T/
2018
DOI: 10.
12
928/
TELK
OMN
I
KA.v1
8i1
.
13639
37
Journ
al h
om
e
page
:
http:
//
jo
ur
nal.
uad.ac
.id
/i
nd
ex.
php/TE
LKOMNIKA
Perform
ance ana
lysis of
m
ulti
layer mu
lticast M
ANET C
RN
based on
ste
in
er
min
imal tree al
gorithm
Basma N
az
ar Nadhi
m,
Sa
r
ab K
am
al Mah
mo
od
El
e
ct
ri
ca
l
Eng
in
ee
ring
Depa
r
tment,
Co
ll
eg
e
of
en
gine
er
ing, Musta
nsiri
y
ah
Univ
ersity
,
Baghda
d
,
Ira
q
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
J
ul
1
9
, 2
019
Re
vised
A
ug
2
0
,
2019
Accepte
d
Se
p 7
, 2
0
19
In
thi
s
stud
y
,
the
m
ult
icast
m
obil
e
ad
hoc
(MA
NET)
CR
N
has
bee
n
deve
lop
ed,
whi
ch
invol
v
es
m
ult
i
-
hop
and
m
ult
ila
y
er
consid
era
t
ion
and
Stei
ner
m
ini
m
al
tre
e
(SM
T)
al
go
rit
hm
is
emplo
y
ed
as
th
e
rout
er
protoc
ol
.
To
enha
nc
e
the
n
etw
ork
per
form
anc
e
with
r
ega
rds
to
throughput
and
pac
k
et
del
iv
er
y
ra
te
(
PD
R),
as
cha
n
nel
assignm
ent
sche
m
e,
the
p
roba
bil
i
t
y
of
succ
ess
(PO
S)
i
s
emplo
y
ed
th
at
ac
count
s
for
th
e
cha
nn
el
av
ai
l
abi
lit
y
an
d
the
ti
m
e
nee
d
e
d
for
tra
nsm
issi
on
when
sele
ct
i
ng
the
best
cha
nnel
from
the
num
ero
us
av
ai
l
abl
e
cha
nn
el
s
for
dat
a
tra
nsm
ission
from
the
s
ourc
e
to
all
desti
nations
nodes
eff
ec
t
ive
l
y
.
W
it
hin
Ra
y
l
eigh
fad
ing
cha
n
nel
s
under
var
ious
ne
twork
par
amet
ers,
a
c
om
par
ison
is
do
ne
for
the
per
fo
rm
anc
e
of
SM
T
m
ult
ic
ast
(
MA
NET)
CRN
with
PO
S
sche
m
e
ver
sus
m
axi
m
u
m
dat
a
r
ate
(MD
R),
m
axi
m
um
ave
rag
e
spec
trum
av
ai
l
ability
(MA
SA
)
a
nd
ran
dom
cha
nne
l
assignm
ent
sche
m
es.
Based
on
the
sim
ula
ti
on
resul
t
s,
the
SM
T
m
ult
ic
ast
(MA
NET)
CRN
with
PO
S
sche
me
was
see
n
to
demons
tra
te
the
b
est
per
for
m
anc
e
ver
sus
o
the
r
sch
emes.
Also,
the
results
prove
d
th
a
t
the
throughput
and
PD
R
pe
rform
anc
e
are
improved
as
the
num
ber
the
pr
imar
y
ch
a
nnel
s
and
th
e
ch
anne
l’s
b
andwid
th
in
cre
as
ed
whi
le
dropp
ed
as
the
val
u
e
of
pac
ke
t
siz
e
D
i
ncr
ea
sed
.
The
n
et
work’s
per
for
m
anc
e
gre
w
with
r
ise
in
th
e
val
ue
of
idl
e
pr
obabi
lit
y
(
)
sinc
e
the
pr
imar
y
us
er’
s
(PU
)
tra
ffi
c
lo
ad is l
o
w when
th
e
va
lu
e
of
is h
igh.
Ke
yw
or
d
s
:
CRN
MANET
PO
S
Ra
yl
ei
gh
f
a
ding
SMT
This
is an
open
acc
ess arti
cl
e
un
der
the
CC
B
Y
-
SA
l
ic
ense
.
Corres
pond
in
g
Aut
h
or
:
Ba
s
m
a N
azar
Nadhim
,
Ele
ct
rical
En
gi
neer
i
ng D
e
par
t
m
ent, Co
ll
ege
of en
gin
ee
rin
g,
Mustansiriy
ah
Un
i
ver
sit
y, Ba
ghda
d,
Ir
a
q
.
Em
a
il
:
bes
m
a.n
azar
@yah
oo.
com
,
besm
a.n
azar@
uo
m
us
ta
nsi
riy
ah.
ed
u.
i
q
.
1.
INTROD
U
CTION
The
dem
and
f
or
ra
dio
sp
ect
ru
m
has
risen
co
ns
ide
rab
ly
because
of
th
e
rece
nt
s
pik
e
in
wireless
serv
ic
es
.
T
od
a
y’s
wireless
sy
stem
s
are
unde
r
the
regulat
io
n
of
a
fixe
d
s
pe
ct
ru
m
assignm
ent
po
li
cy
in
wh
ic
h
a
sp
eci
fied
s
pect
ru
m
ban
d
is
al
lott
ed
to
a
li
ce
ns
e
d
us
er
f
or
a
lon
g
te
rm
basis
and
f
or
a
wider
ge
ogra
ph
ic
al
locat
ion
.
T
he
us
e
of
sp
ect
rum
is
fo
c
us
ed
on
s
pecific
par
t
s
of
the
sp
ect
r
um
;
ho
we
ver,
a
co
ns
ide
rab
le
vo
l
um
e
of
the
s
pectr
um
sta
ys
un
util
iz
ed.
A
s
per
t
h
e
Feder
al
Com
m
un
ic
at
ion
s
Com
m
issi
on
(FC
C),
ge
ograph
ic
and
tem
po
ral
va
ria
ti
on
s
in
th
e
usa
ge
of
the
al
lo
tt
ed
sp
ect
r
um
fall
in
the
ra
nge
of
15%
t
o
85%
[
1
-
4].
Dy
nam
i
c
sp
ect
r
um
acce
s
s
(D
S
A),
al
so
cal
le
d
cogniti
ve
rad
io
net
wor
ks
,
is
reco
m
m
e
nd
e
d
f
or
a
ddre
ssing
th
ese
s
pe
ct
ru
m
ineff
ect
i
ven
es
s
issues
[
2,
5].
Cognit
ive
ra
di
o
(CR)
is
a
n
e
nab
li
ng
te
c
hnol
og
y
f
or
facil
it
at
ing
co
gnit
ive
us
er
s
(secon
dar
y
or
un
li
cen
sed
use
rs)
to
f
un
ct
io
n
on
the
vaca
nt
s
egm
ents
of
the
sp
ect
ru
m
al
lott
ed
to
li
cense
d
us
er
s
(prim
ary
us
ers
).
CR
is
broa
dly
te
rm
ed
as
a
ca
pab
le
te
c
hnology
for
ha
nd
li
ng
t
he
s
pe
ct
ru
m
scarcit
y
issue
trigg
e
red by t
he
ex
ist
in
g
in
fle
xib
le
s
pectr
um al
locat
ion st
ra
te
gy.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
1693
-
6930
TELK
OMN
IKA
Tel
ec
omm
u
n
C
om
pu
t El
Con
t
ro
l
,
V
ol.
18
,
No.
1
,
Fe
bruary
2
02
0:
37
-
50
38
This
is
dep
ic
te
d
in
Fig
ur
e
1.
I
t
is
adep
t
at
identify
ing
it
s
radi
o
env
i
ronm
ent,
and
a
dap
ti
vely
sel
ect
ing
tr
ansm
issi
on
pa
ram
et
ers
as
per
the
se
ns
in
g
ou
tc
om
es.
This
en
han
ces
t
he
perform
ance
of
t
he
co
gnit
ive
rad
i
o
syst
e
m
,
i
m
pr
ov
es
s
pectr
um
eff
ic
acy
an
d
aver
ts
inte
rf
e
r
ence
with
pri
m
ary
us
ers
[
6,
7].
Ba
sed
on
this
def
i
niti
on
,
tw
o
key
at
tribu
te
s
of
the
c
ogniti
ve
ra
di
o
ca
n
be
ou
tl
ine
d
as
f
ollows:
Cognit
ive
capa
bili
ty
,
i.e.
the
com
petence
of
the r
a
dio
te
chnolo
gy
to det
ect
info
rm
at
i
on
f
r
om
it
s
rad
io
en
vir
on
m
ent,
an
d
co
nf
i
gura
bili
ty
,
wh
ic
h
al
lows
t
he
ra
dio
to
be
dynam
ic
ally
pro
gr
am
m
ed
a
s
per
the
ra
dio
env
ir
onm
ent
[2
]
.
Mo
re
prec
ise
ly
,
the
co
gn
it
ive
r
adio
te
ch
nolo
gy
will
allow
th
e
us
ers
t
o
fin
d
ou
t
w
hic
h
seg
m
ents
of
the
s
pectr
um
are
avail
able,
sense
the
e
xistence
of
li
cens
ed
us
e
rs
wh
e
n
a
us
er
f
unct
io
ns
in
a
li
cense
d
ba
nd
(s
pectr
um
sensing
)
,
c
hoos
e
the
best
c
hannel
on
offe
r
(sp
ect
ru
m
m
a
nag
em
ent),
ha
rm
on
ie
s
this
channel’s
acc
ess
with
oth
e
r
us
e
r
s
(sp
ect
ru
m
sh
ar
ing), and
cl
ear
ou
t t
he
c
hannel
after
a li
censed
u
se
r
is sense
d
(s
pectr
um
mo
bili
ty
)
[2
]
. T
hus,
f
or
fu
t
ur
e
wireless
com
m
un
ic
at
ion
s,
it
can
be
c
on
si
der
e
d
as
a
po
te
ntial
t
echni
qu
e
to
m
ini
m
i
ze
sp
ect
r
um
sc
arcit
y
issue
[8
]
.
Figure
1. S
pectru
m
f
or P
U
a
nd S
U
i
n
the
n
et
work
Applic
at
ion
of
the
co
gnit
iv
e
pa
rad
i
gm
t
o
var
i
ou
s
sce
nar
i
os
per
ta
ini
ng
to
m
ulti
-
ho
p
wi
reless
netw
orks
can
be
done
.
The
c
ogniti
ve
ra
dio
ad
-
hoc
netw
ork
is
on
e
su
c
h
s
cenari
o
that
includes
CR
node
s
that
are
in
vo
l
ved
i
n
peer
-
to
-
pee
r
c
omm
un
ic
at
ion
a
m
on
gst
each
oth
e
r
via
a
d
-
hoc
connecti
ons
[
9,
10
]
.
R
ou
ti
ng
in
a
m
ul
ti
-
hop
co
gnit
ive
rad
i
o
ne
twork
(CR
N)
is
qu
it
e
a
diff
ic
ult
ta
sk
and
an
op
e
n
co
nc
ern.
Of
la
te
,
seve
ral
routin
g
protoc
ols
f
or
CR
N
s
hav
e
been
rec
omm
end
ed
a
nd
asse
ssed
.
C
omm
on
ly
,
these
prot
oco
ls
f
oc
us
on
ei
ther
sel
ect
in
g
the
best
qu
al
it
y
cha
nnel
or
th
os
e
channels
that
po
s
sess
the
m
axi
m
u
m
a
ver
a
ge
sp
ect
r
um
-
avail
abili
ty
tim
e.
I
n
a
CR
N,
bot
h
t
he
need
e
d
transm
issi
on
ti
m
e,
as
w
el
l
as
sp
ect
ru
m
-
avail
abili
t
y
tim
e, w
ere fo
und t
o
c
onside
ra
bly affect
r
ou
ti
ng and
netw
or
k
c
onnecti
vity
.
A
c
onsidera
bl
e
dec
rease
i
n
CR
N
perfor
m
ance
co
uld
resu
lt
du
e
to
sp
ect
r
um
-
avail
abili
ty
tim
e
,
par
ti
cula
rly
wh
en
t
her
e
is
a
s
m
al
le
r
aver
age
s
pectr
um
-
avail
abili
ty
tim
e
fo
r
an
as
sign
e
d
cha
nne
l
than
the
nee
ded
tra
ns
m
issi
on
tim
e
ov
e
r
that
cha
nnel
.
I
n
the
w
orst
case,
especial
ly
fo
r
m
ult
i
-
hop
CR
Ns
,
this
issue
beco
m
es i
m
po
rtant w
he
n
se
ve
ral li
nk
s ar
e involve
d.
Netw
ork
pe
rfor
m
ance can
b
e en
ha
nced
by
m
aking
u
se of
var
ie
d
cha
nn
el
qu
al
it
y
as
well
as
sp
ect
rum
avail
abili
t
y
eff
ic
ie
ntly
by
accou
nti
ng
for
co
gn
it
ive
routin
g
protoc
ol
desig
n
[11].
I
n
thi
s
researc
h
w
ork
,
the
m
ult
icast
ro
ute
r
pro
tocol
is
dev
el
op
e
d
by
em
pl
oyin
g
the
Stei
ner
m
i
nim
a
l
tree
(S
MT)
al
go
rithm
fo
r
m
ob
il
e
ad
-
hoc
CR
N
that
includes
m
ulti
-
hop
a
nd
m
ulti
la
ye
r
consi
der
at
io
n.
To
en
ha
nce
the
networ
k
pe
rfo
rm
ance
with
reg
ar
ds
to
the
th
r
oughput
as
wel
l
as
pack
et
delivery
rate
(PDR)
, p
r
ob
a
bili
ty
of
succ
ess
(
POS)
is
e
m
plo
ye
d
as
th
e
cha
nnel
assig
nm
ent
sc
hem
e
for
the
n
et
w
ork
a
fte
r
changin
g
the
netw
ork’s
rando
m
top
ology
to
SMT.
This
is
do
ne
to
s
el
ect
an
eff
ect
ive
channel
for
data
transm
issi
on
ba
sed
on
c
ha
nnel
a
vaila
bili
t
y
as
well
a
s
the
nee
ded
t
i
m
e
fo
r
tra
nsm
issi
on
a
nd
m
ake
a
com
par
ison
of
the
pe
rfor
m
ance
p
e
rtai
ning
to
m
ulti
cas
t
m
ul
ti
la
ye
r
m
u
lt
i
-
hop
ad
-
ho
c
CR
N
in
the
SMT
protoc
ol
as
we
ll
as
PO
S
sche
m
e
al
on
g
with
(MDR),
(MA
SA
)
a
nd
ra
ndom
chan
nel
assi
gn
m
ent
schem
es
at
var
i
ou
s
net
wor
k
par
am
et
ers
to
descr
i
be
eac
h
sc
hem
e
and
sel
ect
the
best
on
e
to
ac
hieve
the
bes
t
pe
rfo
rm
ance
for
the
m
ulti
ca
st m
ult
il
ay
er
m
ulti
-
hop ad
-
hoc
CR
N
in
the
S
MT p
ro
t
oco
l.
The
re
st
of
t
he
pap
e
r
is
str
uctu
red
as foll
ow
s:
s
ect
ion
2
outl
ines
the
m
ulti
ca
st
m
ob
il
e
ad
-
hoc
net
wor
k
(MA
NET).
Se
ct
ion
3
int
rod
uces
the
per
ti
nen
t
work
of
te
chn
iq
ues
a
nd
m
et
ho
dolo
gies
f
or
m
ulticasti
ng
netw
ork
cha
nnel
assignm
ent
and
the
r
ou
t
ing
pr
oto
c
ol
fo
r
un
directed
gr
a
ph.
The
St
ei
ner
al
gorith
m
for
m
ul
ti
cast
ing
ne
twork
is
dis
cusse
d
i
n
sec
ti
on
4.
T
he
syst
e
m
m
od
el
f
or
the
rec
om
m
end
ed
m
ulti
cast
protoc
ol
is
pr
e
sented
in
sect
i
on
5.
T
he
si
m
ulati
on
outc
om
es
and
co
nclusions
are
pr
e
sented
in
sect
ion
s
6
and 7,
resp
ect
i
vely
.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELK
OMN
IKA
Tel
ec
omm
u
n
C
om
pu
t El
Con
t
ro
l
Perf
orma
nce
analysis
of
mu
lt
il
ayer
mu
lt
ic
as
t
MANET CRN
ba
s
ed
on s
te
in
er
…
(
Basma
Naza
r
Na
dhim
)
39
2.
MU
LT
I
CAST
M
OBIL
E
A
D
-
HO
C NET
W
ORK (
M
AN
E
T)
The
ad
-
hoc
ne
twork
is
the
ki
nd
of
wireles
s
networ
k
that
com
pr
ise
s
a
cl
us
te
r
of
wire
le
ss
nodes,
wh
ic
h
are
a
de
pt
enough
to
c
om
m
un
ic
at
e
with
each
oth
e
r
in
the
abse
nce
of
infr
a
struct
ur
e
.
Ba
sic
al
ly
,
there
are
two
kinds
of
inf
rastr
uctu
re
-
le
ss
wi
reless
netw
orks:
st
at
ic
ad
-
ho
c
ne
twork
an
d
m
ob
il
e
a
d
-
hoc
netw
ork
(MA
NET)
[
12]
.
A
MA
NET
c
an
be
de
fine
d
a
s
an
interc
onne
ct
ed
syst
e
m
of
m
ob
i
le
ho
sts
and
does
no
t
in
cl
ude
a fix
e
d
inf
rastr
uctu
re.
I
n
M
A
NETs
, each
m
ob
il
e host
po
s
s
esses the a
bili
t
y fo
r
m
ulti
-
hop t
ran
sm
issi
on
. Also,
it
act
s
as
a
ro
uter
,
in
wh
ic
h
each
and
eve
ry
node
inv
ol
ved
in
t
he
MAN
ET
ne
eds
to
know
it
s
neighb
or
as
w
el
l
as
serv
e
as
a
r
oute
r
to
ad
va
nce
the
datag
ram
to
the
sp
e
ci
fied
desti
nation
[
13,
14]
.
Mult
ic
ast
ing
is
em
plo
ye
d
wh
e
n
t
her
e
is a
n
ee
d for a
pp
li
cat
ion
s t
o
se
nd the sam
e d
at
a
to d
i
ff
e
ren
t
des
ti
nations
sim
ultaneo
us
ly
.
The
m
ulti
cast
r
ou
ti
ng
proto
col
is
desig
ne
d
for
m
ob
il
e
ad
ho
c
net
wor
ks
(MA
NETs
)
to
pro
vide
su
pp
or
t
f
or
i
nfor
m
at
ion
disse
m
inati
on
f
r
om
a
se
nd
e
r
t
o
al
l
receive
rs
pert
ai
nin
g
t
o
a
m
ulti
cast
gro
up,
w
hile
al
so
em
plo
yi
ng
the
avail
able
bandw
i
dth
in
an
eff
ect
ive
m
ann
e
r
w
he
n
th
ere
are
fr
e
quen
t
top
ology
va
riat
ion
s.
Fo
r
M
A
NETs
,
va
rio
us
m
ulticast
routin
g
protoc
ols
hav
e
be
en
pu
t
f
orward.
M
ulti
cast
ing
c
omm
un
ic
at
i
on
i
s
consi
der
e
d
to
be
co
st
-
ef
fe
ct
ive
f
or
ap
pl
ic
at
ion
s
w
he
n
se
ndin
g
t
he
sam
e
data
to
va
rio
us
r
eci
pients
si
m
ultaneou
sly
.
In
c
ontrast
to
send
i
ng
data
thr
ough
m
ulti
pl
e
u
nicast
,
m
ult
ic
ast
ing
dec
rea
ses
the
co
nsum
ptio
n
of
li
nk
ba
ndw
idth,
delivery
delay
an
d
r
oute
r
pr
ocessin
g
[13,
15]
.
Th
e
current
m
ulti
c
ast
routin
g
pr
oto
c
ols
desig
ne
d
f
or
MANET
s
can
be
m
ajo
rly
se
gm
ented
into
m
esh
-
ba
sed
a
nd
tree
-
base
d.
These
protoc
ol
s
dif
fer
wit
h
re
gards
to
the
re
dunda
ncy
involve
d
in
the
pat
hs
be
tween
recei
vers
and
se
nd
e
rs.
Tree
-
based
protoc
ol
s
offer
just
a
sing
le
path
be
tween
receiv
ers
an
d
se
nd
ers,
wh
il
e
m
esh
-
base
d
pro
tocols
offer
m
ul
ti
ple
paths
[
13,
15]
.
Segm
entat
ion
of
routin
g
protoc
ols
is
do
ne
with
re
gards
to
m
essage
de
li
ver
y
sem
antic
s
as
un
ic
ast
,
bro
a
dc
ast
and m
ulti
ca
st, as
pr
ese
nted
in [1
4].
The
data
strea
m
is
ref
er
re
d
as
a
‘s
ource
’
or
sen
de
r
w
hil
e
the
e
nd
-
us
er
seeki
ng
to
re
cei
ve
the
data
stream
is
ref
erre
d
to
as
a
‘r
ecei
ver
’
.
If
a
sin
gle
receiver
no
de
is
pr
esent,
th
e
ro
utin
g
issue
d
is
cal
le
d
as
un
ic
ast
routin
g,
wh
ic
h
can
be
a
ddres
s
ed
with
c
om
puta
ti
on
of
the
s
hortest
pat
h
bet
ween
t
he
recei
ver
a
nd
the
s
ource
i
n
wh
ic
h
data
are
sent
from
on
e
source
to
t
he
receiver
.
In
th
e
m
ulti
cas
t
rou
ti
ng
,
a
s
ource
can
tra
ns
m
it
its
data
stream
to
a
cl
ust
er
of
hosts.
On
the
ot
her
ha
nd,
broa
dcast
r
ou
ti
ng
or
si
m
pl
y
broad
ca
s
ti
ng
ca
n
be
de
fine
d
as
transm
itti
ng
of
the
stream
from
the
so
urce
to
al
l
destinat
ion
s
node
co
nnect
ed
t
o
the
netw
ork.
M
ulti
cast
ing
involves
bro
a
dc
ast
ing
a
nd
un
ic
ast
ing
as
a
s
pecial
case,
a
nd
re
so
l
ves
the
issue
of
reachi
ng
of
the
strea
m
to
a
gro
up of
nodes
in
the
d
e
sti
nation [
16,
17]
.
3.
RELATE
D
W
ORKS
Ma
ny
ap
proac
hes
a
nd
m
et
hodo
l
og
ie
s
f
or
c
ha
nn
el
as
sig
nme
nt
an
d
c
onstr
ucting
the
rout
ing
tree
f
ro
m
ra
nd
om
un
dire
ct
ed
topolo
gy
are
dep
l
oyed
to
i
m
pr
ov
e
t
he
pe
rfor
m
ance
of
the
m
ulticast
cogniti
ve
rad
io
netw
ork
with
r
egards
t
o
the
t
hro
ughput
a
nd
var
i
ous
pa
ra
m
et
ers
[16].
T
he
m
ulti
cast
network’s
perform
ance
wh
e
n
tra
ns
m
itti
ng
data
to
m
ul
ti
ple
us
ers
reli
es
on
t
he
way
netw
ork
topolo
gy’s
c
onnecti
on
is
a
rr
a
ng
e
d.
In
[
18]
,
for
m
u
lt
ic
ast
netwo
rk,
introduc
ti
on
of
m
ini
m
u
m
S
te
iner
tree
al
gorithm
was
done
as
ro
utin
g
pr
oto
c
ol
and
presentat
i
on
was
done
f
or
t
he
m
et
ho
d
to
trans
f
or
m
un
di
rected
to
polog
y
to
Stei
ner
tree
in
[
19]
.
Be
sides
util
iz
ing
the
ef
fici
ent
tree
al
go
rithm
,
the
ch
oice
of
best
ch
ann
el
f
or
trans
m
itti
ng
data
in
a
m
ulti
-
ho
p
m
ulti
cast
cogniti
ve
ra
dio
netw
ork
is
a
vital
top
ic
f
or
i
m
pr
ovin
g
the
netw
ork’s
thr
ough
pu
t
perfor
m
ance.
The
m
ulti
cast
routin
g
sc
hem
e
is
cl
ubbe
d
w
it
h
the
al
l
oc
at
ed
cha
nnel
of
m
ulti
-
hop
m
ul
ti
cast
CRN
as
pr
e
sente
d
in
[
20
]
.
Mult
ic
ast
tree
is
e
m
plo
ye
d
fo
r
c
onstr
uc
ti
ng
m
ini
m
u
m
energy
direc
t
Stei
ner
tree
with
the
he
lp
of
low
-
com
plexity
approxim
ati
on
al
gorithm
for
CR
N
a
nd
the
im
pact
of
pri
m
ary
net
work
tra
f
fic
load
i
s
evaluate
d
to
d
e
te
rm
ine sp
ect
r
um
p
resen
ce
oppo
rtu
niti
es [
21]
.
The
su
m
rate
i
s
m
axi
m
ise
d
ta
ken
f
r
om
al
l
us
ers
of
CR
N
in term
s
of
trans
m
issi
on
rate
as
well
as
j
oin
t
channel
sp
ect
r
um
assign
m
ent
per
ta
inin
g
to
the
arr
a
nged
ac
cess
channel
f
or
m
ulti
u
ser
sing
le
-
tran
scei
ve
r
CR
N
opport
un
it
ie
s
as
sugg
e
ste
d
i
n
[
22]
.
CR
N’s
thr
oughput
perform
ance
can
be
co
ns
i
de
rab
ly
im
pr
ov
e
d
an
d
al
gorithm
can
be
em
plo
ye
d
t
o
decr
ea
se
the
end
-
to
-
en
d
del
ay
fo
r
cha
nnel
assignm
ent
as
well
as
fo
r
m
ulti
cast
routin
g
in
the
m
esh
netw
ork
per
ta
ini
ng
t
o
CR
.
This
a
lgorit
hm
con
s
iders
t
he
s
witc
hing
la
te
ncy
an
d
the
ch
an
nel’s
he
te
rogen
ei
ty
as
rec
omm
end
ed
in
[23].
In
cas
e
of
the
m
ulti
-
hop
CR
N,
t
he
pro
bab
il
ist
ic
routi
ng
schem
e
al
go
rithm
was
reco
m
m
end
ed
i
n
[
24]
.
This
al
gorithm
con
siders
t
he
cha
nnel
’s
a
vaila
bili
ty
time
an
d
essenti
al
trans
m
issi
on
tim
e.
The
CR
N
’s
be
st
throu
ghput
perform
ance
can
be
at
ta
ine
d
by
pr
e
ferrin
g
thi
s
schem
e
ov
er
oth
e
r
schem
es.
In
a
dd
it
io
n,
within
Ra
yl
ei
gh
fa
ding
cha
nnel
,
em
plo
yi
ng
of
m
axi
m
u
m
PO
S
routin
g protoc
ol w
a
s
done fo
r
m
ulti
-
ho
p C
RN [1
1].
Ov
e
rlap
ping
as
well
as
non
-
overla
pp
i
ng
cha
nn
el
a
ssign
m
ent
alg
ori
thm
s
fo
r
a
m
plifyi
ng
the
thr
ough
pu
t
perf
or
m
ance
are
rec
omm
en
ded
in
[
25]
.
T
he
cr
os
s
la
ye
r
routin
g
proto
col
proce
ss
w
as
pu
t
forw
a
r
d
in [
26]
to
choose th
e
best ch
a
nn
el
fr
om
a
m
on
gs
t t
he
ob
ta
ina
ble cha
nn
el
s
per
ta
ini
ng
to
m
ulti
cast
CR
N
to
obta
in
bette
r
vid
e
o
qual
it
y
by
al
l
recei
ver
s
’
no
des
w
it
hin
the
netw
ork.
I
n
[
17
,
27]
,
SPT
a
nd
MST
are
util
iz
ed
as
r
ou
ti
ng
protoc
ols
for
t
he
m
ultilay
er
m
ult
ic
ast
m
ulti
-
hop
C
RN
with
P
OS
schem
e
as
channel
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
1693
-
6930
TELK
OMN
IKA
Tel
ec
omm
u
n
C
om
pu
t El
Con
t
ro
l
,
V
ol.
18
,
No.
1
,
Fe
bruary
2
02
0:
37
-
50
40
assignm
ent.
T
he
net
wor
k’
s
pe
rfor
m
ance
is
i
m
pr
oved
with
reg
a
rds
to
the
thr
oughput
an
d
P
DR
com
pared
t
o
oth
e
r
sc
hem
es.
EXT
with
MS
T
an
d
SP
T
are
d
epl
oyed
as
r
outi
ng
al
gorith
m
s
fo
r
the
m
ulti
cast
m
ulti
-
ho
p
CR
N
.
The
P
OS sch
e
m
e w
as u
ti
li
zed as
c
ha
nn
el
a
s
s
ign
m
ent f
or s
el
ect
ing
an ef
fici
ent ch
a
nn
el
f
or
t
ran
sm
it
ti
ng
d
at
a in
accor
da
nce
wi
th
the
c
hannel
avail
abili
ty
a
nd
t
he
re
quisi
te
transm
issi
on
tim
e
as
reco
m
m
end
ed
in
[12].
Rou
ti
ng
prot
oc
ol
m
et
rics
in
m
ul
ti
ro
ute
and
one
-
way
rou
ti
ng
for
wirele
ss
and
co
gnit
ive
rad
i
o
netw
orks
are
com
par
ed
a
nd
analy
zed its c
ha
ll
eng
es as
r
ec
omm
end
ed
in
[28
]
.
4.
STE
INER AL
GORIT
HM
F
OR M
ULTIC
AS
TI
NG NET
WORK
The
Stei
ner
m
i
nim
a
l
tree
(SM
T)
al
gorithm
is
am
on
g
the
sever
al
al
gorithm
s
fr
om
the
gr
a
ph
the
ory
wh
ic
h
ha
ve
be
en
util
iz
ed
for
fixing
r
outi
ng
issues
in
a
wireless
m
ulticasti
ng
net
wor
k
[
16
]
.
T
he
S
te
iner
netw
ork
e
ncom
passes
two
ki
nd
s
of
ve
rtic
es:
te
rm
inals
(r
e
qu
i
red)
ver
ti
ces
an
d
no
nterm
i
nal
(S
te
ine
r)
ve
rtic
es
.
SMT
al
gorith
m
and
sp
a
nnin
g
tree
are
not
the
sam
e,
in
that
a
s
pa
nn
i
ng
tree
li
nks
al
l
ver
ti
ces
of
a
s
pecific
gr
a
ph,
w
hile
a
SMT
bi
nds
a
s
pecific
s
ubset
of
the
ve
rtic
es
(so
m
e
of
Stei
ne
r
ver
ti
ces
a
nd
te
rm
inals
ver
ti
ces
t
o
decr
ease
the
t
r
ee
cost)
[
16,
19]
.
A
key
iss
ue
with
SMT
is
how
t
o
fi
nd
a
t
ree
that
not
on
ly
has
m
ini
m
u
m
cost
bu
t al
s
o
i
nclu
de
s all
term
inal
ver
ti
ces as
w
el
l as any s
ubset
per
ta
ini
ng to n
on
te
rm
inal (Stei
ner
)
v
e
rtic
es.
Let
G
=
(
V,
E
,
w)
represe
nt
the
un
directed
gr
a
ph
al
ong
wi
th
a
set
of
ve
rtic
es
V
as
well
as
ed
ges
E
al
ong
with
nonn
e
gative
wei
gh
ts
w
.
I
n
Stei
ner
m
ini
m
al
tree
T,
t
he
needed
ver
ti
ces
set
is
(L
⸦
V)
,
in
w
hic
h
(T
⸦
G)
that
has
a
set
of
ve
rtic
es
W
t
hat
inclu
des
al
l
the
need
e
d
ve
r
ti
ces
L
as
wel
l
a
s
(
W
-
L
)
St
ei
ner
ver
ti
ces
[19,
29]
. T
he
ste
ps
t
o determ
ine SMT with m
inim
u
m
total
ed
ge
w
ei
ght
W
a
re
[19,
29
]
:
The
m
et
ric
cl
os
ur
e
G_L
on L
is de
velo
ped.
Krus
kal’s
al
gorithm
is e
m
plo
ye
d
to
determ
i
ne
m
ini
m
u
m
s
pannin
g
t
ree
M
ST
on
.
Stei
ner
ve
rtic
es
wer
e
inse
rted
betwee
n
a
pa
ir
of
te
rm
inal
ver
ti
ces
in
T_L
as
interm
e
diate
po
ints
t
o
determ
ine Stei
ner m
ini
m
al
tree T.
To
el
ucidate th
ese ste
ps, the
foll
ow
i
ng ex
am
ple is intr
oduc
ed [1
9]:
Figure
2
(
a
)
dem
on
strat
es
un
directed
grap
h
G
al
ong
with
nonterm
inal
ve
rtic
es
{u
1,
u2,
u3,
u4}
as
well
as
te
r
m
inal
ver
ti
ces
L
=
{v1
,
v2,
v3,
v4
,
v5}.
Th
e
sh
ort
est
path
le
ng
th
s
am
on
gst
al
l
L
ver
ti
c
es
are
determ
ined
to
bu
il
d
m
et
ric
clo
sure
on
L
,
a
s
pr
ese
nte
d
in
Figure
2
(
b
)
.
Krus
kal’s
al
go
rithm
is
us
ed
to
determ
ine
the
MST
as
de
m
on
strat
ed
in
Fi
gure
2
(
c
)
.
To
de
te
rm
ine
SM
T,
nonterm
inal
ve
rtic
es
{u
1,
u2}
are
placed
in
the
pa
th
betwee
n
te
r
m
inal
ver
ti
ces
{v1,
v4}
to
ke
ep
the
tree’s
to
ta
l
weigh
t
m
ini
m
u
m
as
pr
esen
te
d
in
Figure
2
(
d
)
.
Ba
sed
on
Figure
s
2
(
c
)
a
nd
2
(
d
)
,
it
can
be
seen
that
the
total
weigh
t
of
the
tree
is
22
pr
i
or
t
o
add
it
io
n
of
{u
1,
u2},
w
hile
the
tree
’s
t
otal
weig
ht
is
de
cr
eased
t
o
j
us
t
17
po
st
a
ddit
io
n
of
{
u1,
u2},
w
hich
ind
ic
at
es that t
he
Stei
ne
r p
oint
s cast an im
pact that dec
reas
e the total
c
os
t
of tree.
Figure
2.
Stei
ne
r
m
ini
m
a
l t
ree algorit
hm
: (
a)
undirecte
d gr
a
ph G,
(
b) the
m
et
ric cl
os
ure
G
L,
(
c)
t
he
MST
T
L
, d)
t
he
SMT
T
(a)
(b)
(c)
(d)
Evaluation Warning : The document was created with Spire.PDF for Python.
TELK
OMN
IKA
Tel
ec
omm
u
n
C
om
pu
t El
Con
t
ro
l
Perf
orma
nce
analysis
of
mu
lt
il
ayer
mu
lt
ic
as
t
MANET CRN
ba
s
ed
on s
te
in
er
…
(
Basma
Naza
r
Na
dhim
)
41
5.
SY
STE
M MO
DEL FO
R
P
R
OPOSE
D MU
LT
ICA
ST
CR
N
To
en
ha
nce
th
e
eff
ic
ie
ncy
of
total
sp
ect
ru
m
as
well
as
increase
the
ove
ra
ll
thro
ug
hput
of
the
CR
N,
nu
m
erous
a
ppr
oach
e
s
an
d
te
c
hn
i
qu
e
s
ha
ve
been
em
plo
ye
d.
At
m
os
t,
est
ablishm
ent
of
the
cu
rr
e
nt
pro
tocols
was
do
ne
on
ei
ther
the
pro
cess
of
path
s
el
ect
ion
or
t
h
e
channel
sch
e
m
e
assignm
e
nt
[12].
I
n
ad
diti
on,
the
sel
ect
io
n
of
cha
nnel
s
by
curre
nt
cha
nnel
assi
gn
m
ent
schem
es
reli
es
only
on
th
e
sp
ect
ru
m
’s
aver
a
ge
avail
abili
ty
su
ch
as
Ma
xim
um
Av
erag
e
Spec
trum
Av
ai
la
bili
ty
(MASA)
schem
e
or
ba
sed
on
the
c
ha
nn
el
’s
qu
al
it
y
su
c
h
as
Ma
xim
u
m
Data
Ra
te
(MDR)
sche
m
e,
or
sel
ect
s
the
c
hanne
l
rando
m
ly
with
no
lim
it
at
ion
[12,
17
]
.
T
his
paper
consi
de
rs
both
the
pro
ba
bili
ty
of
su
cces
s
(
P
OS
)
cha
nnel
as
sign
m
ent
sche
m
e
as
well
as
the
path
sel
ect
ion
pr
oc
ess
that
e
m
plo
ys
the
Stei
ner
m
ini
m
a
l
tree
(S
MT)
al
go
rithm
.
E
m
plo
yi
ng
POS
schem
e
as
chan
nel
as
sig
nm
e
nt
is
hel
pful
f
or
pe
rfo
rm
ance
enh
a
ncem
ent
of
the
m
ulti
la
yer
m
ulti
cast
m
ul
ti
-
hop
CR
N
en
vir
onm
ent
syst
e
m
with
re
gards
t
o
the
th
r
oughput
as
well
as
pack
et
deliver
y
rate
(P
DR
)
s
ubj
ect
t
o
var
i
ou
s
net
work
pa
ram
et
ers
as
this
sc
hem
e
co
ns
ide
rs
th
e
nee
ded
ti
m
e
for
tra
ns
m
iss
ion
a
nd
the
c
hanne
l
avail
abili
ty
wh
en
sel
ect
ing
th
e
best
cha
nn
el
s
for
eff
ect
i
ve
data
transm
is
sion
to
al
l
des
ti
nation
no
des
fr
om
the
source
.
I
n
t
his
resea
rch
w
ork,
for
vid
e
o
transm
issi
on
to
destinat
io
n
no
des
f
r
om
so
ur
c
e
node
ov
e
r
a
sing
le
session, m
ulti
l
ay
er m
ulti
cast
m
ul
ti
-
hop
CR
N
r
outi
ng
protoco
l i
s
em
plo
ye
d.
At
first,
ge
ne
r
at
ion
of
undi
re
ct
ed
gr
a
ph
is
done
with
a
se
t
of
nonte
rm
i
n
al
Stei
ner
(
Nnt
)
as
well
as
te
rm
inal
(
Nt)
ver
ti
ces
within
the
square
area
,
and
the
n
deter
m
inati
on
of
St
ei
ner
m
ini
m
al
tree
(S
MT)
is
done
e
m
plo
yi
ng
the
al
gorithm
as
mention
e
d
in
sec
ti
on
(
IV).
Be
tween
eac
h
de
s
ti
nation
no
de
a
nd
t
he
source
node
,
there
are
nu
m
e
rous
avail
a
ble
PU
cha
nnel
s
(
M)
and
the
Ma
rko
v
m
od
el
is
the
sta
tus
m
od
el
per
ta
inin
g
to
each
pr
im
ary
us
er
(
PU
)
cha
nnel
,
wh
ic
h
keeps
in
te
rch
a
ng
i
ng
be
tween
t
he
tw
o
sta
te
s
(idle
a
nd
busy).
The
idl
e
sta
te
su
ggest
s
that
SU
can
us
e
the
cha
nn
el
as
it
is
no
t
us
e
d
by
PU,
w
hile
bu
sy
sta
te
i
m
plies
that
P
U
us
es
the
c
hannel
a
nd
S
U
can
not
use
the
sam
e.
Fo
r
al
l
cha
nnel
s,
ide
ntica
l
ba
ndwi
dth
(
BW
)
i
s
fix
ed
.
Applyi
ng
of
the
P
OS
sc
he
m
e
is
done
to
t
he
netw
ork
as
channel
assi
gnm
ent
to
im
pr
ov
e
netw
ork
perform
ance
as
w
el
l
as
to
o
r
ga
nize
the
CR
N
transm
issi
on
s
so
as
to
m
ake
avail
able
com
m
on
con
t
ro
l
c
hannel
(CCC
)
[12,
17
]
.
The
pro
ba
bili
t
y
of
s
uccess
(
(
−
)
)
bet
ween
any
of
the
tw
o
node
s,
i
a
nd
j
,
in
SMT
tree
with
re
gards
to
channel
j
for
t
he
av
ai
la
ble c
ha
nn
el
(
C
)
of CR
N
ca
n
be
e
xpre
ssed
as
in
[11
]
:
(
−
)
=
(
−
(
)
(
−
)
)
(1)
wh
e
re
de
no
te
s
the
a
ver
a
ge
a
vaila
bili
ty
time
pe
rtai
ning
t
o
sp
ect
r
um
in
(
in
sec)
f
or
c
ha
nn
el
j
an
d
(
−
)
represe
nts
the
r
equ
i
red
tra
ns
m
issi
on
tim
e
in
(in
sec/
pack
et
)
to
transm
it
a
pa
cket
to
node
k
from
i
ov
er
ch
ann
e
l
j
, w
hich
ca
n be
expres
sed
as i
n [11]:
(
−
)
=
(
−
)
(2)
wh
e
re
D
si
gn
i
f
ie
s
the
pack
et
siz
e
(in
bits)
a
nd
(
−
)
de
no
te
s
t
he
data
rate
(i
n
bit/
sec)
betwee
n
i
a
nd
k
node
s
ov
e
r
c
ha
nn
el
j
,
as e
xpresse
d
i
n [11]:
(
−
)
=
(
)
2
(
1
+
(
)
(
−
)
∗
0
)
(3)
wh
e
re
0
de
note
s
the
the
rm
al
power
de
ns
it
y
in
(
W
at
t/
Hz),
re
presents
the
c
ha
nn
el
ba
ndwidt
h
a
nd
(
)
(
−
)
represe
nts the r
ecei
ved
powe
r t
o
recei
ver j f
r
om
tran
sm
it
te
r
i, as e
xpresse
d i
n
[
11]
:
(
−
)
=
(
4
)
2
(
(
)
(
−
)
)
(4)
wh
e
re
de
note
s
the
CR
’s
tra
nsm
issi
on
po
wer,
d
sig
nifies
t
he
distance
bet
ween
any
tw
o
nodes
,
n
r
ep
res
ents
the
path
l
os
s
e
xponent,
a
nd
(
)
(
−
)
i
m
pl
ie
s
the
cha
nn
el
p
owe
r
gai
n
bet
ween
i
a
nd
k
nodes
over
channel
j. W
it
h
reg
a
rds to
Ray
le
igh
fa
ding,
e
xpone
ntial
d
ist
ri
bu
ti
on
of
(
)
(
−
)
is done
with m
ean 1 [11
]
.
To
e
xp
la
in
ho
w
the
P
OS
ch
ann
el
assi
gnm
ent
schem
e
is
us
e
d
to
c
hoose
the
be
st
ch
ann
el
from
avail
able
cha
nn
el
s
f
or
pr
i
m
ary
and
seconda
ry
us
ers
to
transm
i
t
data
fr
om
so
urce
no
de
to
al
l
destinat
io
n
no
des
Fig
ur
e
3
is
us
ed.
We
hav
e
on
e
s
ource
node
(
v1)
a
nd
six
n
od
e
s
(fo
ur
no
des
are
the d
e
sti
nation
(v2, v
3, v4 a
nd
v5). T
hr
ee
av
a
il
able cha
nn
el
s
are use
d (CH
1, CH
2
a
nd CH
3).
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
1693
-
6930
TELK
OMN
IKA
Tel
ec
omm
u
n
C
om
pu
t El
Con
t
ro
l
,
V
ol.
18
,
No.
1
,
Fe
bruary
2
02
0:
37
-
50
42
Figure
3. Stei
ne
r
m
ini
m
a
l t
ree w
it
h
t
hr
ee
lay
ers
Fr
om
Fig
ur
e
3
one
ca
n
note
that
the
SMT
tree
has
t
hr
ee
la
ye
r
an
d
the
proces
s
of
P
OS
c
ha
nn
e
l
assignm
ent is a
pp
li
ed
to
e
ach
lay
er as fo
ll
ows
:
a. T
he
tra
ns
m
is
sion o
f
la
ye
r 1:
The
unic
ast
tra
ns
m
issi
on
is
us
ed
to
tra
ns
m
i
t
data
from
so
urce
node
(
v1)
to
node
(
u1)
a
nd
the
cha
nn
e
l
with
m
axi
m
u
m
value
of PO
S
is
the b
est
c
hannel
that use
d
f
or
d
at
a
tra
ns
m
i
ssion
(C
hannel
3
with POS=0
.
90
80
is
us
e
d
in
this case)
as
Table 1.
Note
that
th
e
CH
2
ha
s
(
POS=0)
this
m
ean
that
this
c
ha
nnel
is
occ
up
ie
d
by
P
U
and can
not be
us
e
d by SU.
Table
1.
POS f
or the tra
ns
m
is
sion o
f
la
ye
r
1
Ch
an
n
el
Receiv
er
CH1
CH2
CH3
POS f
o
r
b
est ch
an
n
el
v1
-
u1
0
.75
5
5
0
0
.90
8
0
0
.90
8
0
b.
The
transm
i
ssion o
f
la
ye
r 2
:
The
m
ulti
cast
t
ran
sm
issi
on
is
us
e
d
to
tra
ns
m
i
t
data
from
(u
1)
node
t
o
(
v2,
u2)
no
des.
Mi
nim
u
m
PO
S
value
is
ch
oos
ing
f
ro
m
each
channel
then
m
axi
m
u
m
val
ue
of
this
m
ini
m
u
m
values
is
choosin
g
f
or
data
transm
issi
on
(
Chan
nel
3
with
P
OS
=
0.885
3
is
us
ed
i
n
this
case)
as
Ta
ble
2.
Also,
it
sh
ould
be
note
d
th
at
CH2
cannot
be use
d f
or
t
ran
sm
issio
n beca
us
e
it
used
by P
U.
Table
2.
POS f
or the tra
ns
m
is
sion o
f
la
ye
r 2
Ch
an
n
el
Receiv
er
CH1
CH2
CH3
POS f
o
r
b
est ch
an
n
el
u1
-
v2
0
.72
0
5
0
0
.88
5
3
u1
-
v2
0
.81
6
5
0
0
.91
9
7
Min POS
0
.72
0
5
0
0
.88
5
3
0
.88
5
3
c. T
he
tra
ns
m
is
sion o
f
la
ye
r 3:
Part1:
The
unic
ast
transm
issio
n
to
tra
ns
m
it
data
from
no
de
v2
to
v5
node
is
us
e
d.
T
he
cha
nn
el
with
heig
ht v
al
ue of
POS is c
hoos
i
ng (
C
hannel2
with P
OS
=
0.8
661
is
u
se
d
i
n
t
his case
)
as i
n Table
3.
Part2:
T
he
m
ul
ti
cast
transm
is
sion
f
ro
m
node
u2
node
to
(v3,
v4)
no
des
is
us
e
d.
As
in
La
ye
r
2
m
ini
m
u
m
values
of
POS
f
or
eac
h
c
ha
nn
el
is
us
ed
then
f
or
data
transm
issi
on
m
axi
m
u
m
value
of
P
OS
f
r
o
m
m
ini
m
u
m
is cho
os
i
ng (
C
ha
nnel
3
with
P
OS
=
0.854
7
is
us
e
d i
n
this case
)
as
in Ta
ble 4.
Table
3.
POS f
or the tra
ns
m
is
sion o
f
la
ye
r 3
par
t
1
Ch
an
n
el
Receiv
er
CH1
CH2
CH3
POS f
o
r
b
est ch
an
n
el
v2
-
v5
0
.77
2
6
0
.86
6
1
0
0
.86
6
1
Table
4.
POS f
or
the tra
ns
m
is
sion o
f
la
ye
r 3
par
t
2
Ch
an
n
el
Receiv
er
CH1
CH2
CH3
POS f
o
r
b
est ch
an
n
el
u2
-
v3
0
.62
1
5
0
.84
3
6
0
.87
3
9
u2
-
u4
0
.62
1
5
0
.76
8
1
0
.85
4
7
Min POS
0
.62
1
5
0
.76
8
1
0
.85
4
7
0
.85
4
7
Evaluation Warning : The document was created with Spire.PDF for Python.
TELK
OMN
IKA
Tel
ec
omm
u
n
C
om
pu
t El
Con
t
ro
l
Perf
orma
nce
analysis
of
mu
lt
il
ayer
mu
lt
ic
as
t
MANET CRN
ba
s
ed
on s
te
in
er
…
(
Basma
Naza
r
Na
dhim
)
43
6.
COMP
UTER
SIMULATI
O
N AND
RES
U
LT
S
In
this
sect
io
n,
com
pu
te
r
s
i
m
ulati
on
s
are
carried
out
to
assess
the
perform
ance
per
ta
inin
g
to
m
ul
ti
la
ye
r
m
u
lt
i
-
hop
m
ulti
c
ast
m
ob
il
e
ad
hoc
CR
N
by
em
plo
yi
ng
th
e
SMT
al
go
rithm
fo
r
c
onstr
ai
ni
ng
the
netw
ork
with
P
OS
cha
nn
el
assi
gn
m
ent
schem
e.
Th
is
al
lows
sel
e
ct
ing
an
e
ff
ic
i
ent
cha
nn
el
f
or
data
transm
issi
on
base
d on
t
he
ch
ann
el
a
vaila
bili
ty
an
d
the
needed tim
e to car
r
y ou
t t
ra
ns
m
iss
ion
. A c
om
par
ison
i
s
carried
out
to
e
valuate
the
perform
ance
of
th
e
pu
t
f
orwa
rd
protoc
ol
al
ong
with
P
OS
sc
he
m
e
ver
su
s
m
axi
m
u
m
data
rate
(MD
R),
m
axi
m
u
m
aver
a
ge
sp
ect
r
um
avail
ability
(MASA)
an
d
r
andom
chan
nel
assignm
ent
sc
hem
e
s
with
re
gards
t
o
the
pac
ket
de
li
ver
y
rate
(PDR)
an
d
th
r
oughput,
sub
j
ect
ed
to
va
rio
us
netw
ork
pa
ra
m
et
ers
within
the
Ra
yl
ei
gh
fa
ding
c
ha
nn
el
m
od
el
.
We
ha
ve
acco
unte
d
f
or
t
hr
ee
idle
pro
bab
il
it
y
(i.e
.
=0.
3,
0.6,
0.9
)
to
assess
the
im
pact
cast
by
pr
im
ary
us
ers
traff
ic
load
on
the
perfor
m
ance
of
co
gnit
ive
rad
i
o
ne
twork
.
MATLAB
(R
2018a)
is
de
pl
oyed
for
co
m
pu
te
r
si
m
ula
ti
on
.
F
or
com
par
is
on
purpo
se,
the
syst
e
m
m
od
el
par
a
m
et
ers
are
dep
ic
te
d i
n Ta
ble 5.
Table
5.
Syst
em
p
ara
m
et
ers
Para
m
eter
Valu
e/Ty
p
e
Netwo
rk ar
ea
2
0
0
*
2
0
0
m
No
.
o
f
T
er
m
in
al n
o
d
es (Nt)
20
No
.
o
f
Non
ter
m
in
a
l no
d
es (Nnt)
10
Top
o
lo
g
y
tr
ee
SMT
tree
No
.
o
f
CR so
u
rce
On
e so
u
rce
No
of
pri
m
ar
y
cha
n
n
el (
M)
15
PU chan
n
el
m
o
d
el
Mar
k
o
v
m
o
d
el
Idle p
rob
ab
ility
P
I
[
0
.3 0
.6 0
.9]
Av
erage availab
ilit
y
ti
m
e
(
)
Ran
g
e f
ro
m
2
m
s t
o
45
m
s
Ban
d
wid
th
(
BW)
1
M
Hz
Pack
et size
(D)
4
KB
Tr
an
s
m
iss
io
n
po
wer
(P
t)
0
.1 W
att
Ch
an
n
el us
ed
Rayleigh
f
ad
in
g
ch
an
n
el
Path
los
s
ex
p
o
n
en
t
(
n
)
4
Ther
m
a
l no
ise p
o
wer (
0
)
10
−
8
W
/
Hz
6.1.
Perfo
r
ma
nce
eva
lu
ati
on
of
St
ei
ner
mi
ni
mal
tree
(
SMT
)
multic
ast
CRN
un
der
t
he
im
pa
c
t
of
channel
b
an
d
w
idth
As
pr
ese
nted
in
Fi
gures
4
(a
-
c)
a
nd
Fi
gure
5
(a
-
c
),
t
he
t
hro
ughput
as
well
as
P
DR
perform
ance
per
ta
ini
ng
t
o
m
ul
ti
la
ye
r
m
ul
ti
-
hop
m
ulti
ca
st
CR
N
is
co
m
par
ed
with
c
hannel
ba
ndwi
dth
possessin
g
var
i
ou
s
ty
pes
of
cha
nnel
assi
gn
m
ent
schem
es
as
well
as
three
values
of
I
dle
pr
ob
a
bili
ty
[
P
I
=
0
.
3
,
0
.
6
,
a
nd
0
.
9
]
,
resp
ect
ively
.
The
fig
ur
e
s
in
dicat
e
th
at
as
the
c
hannel’s
band
width
r
ose
,
the
pe
rfo
rm
ance
of
al
l
sc
hem
es
enh
a
nce
d
as
the
data
rate
(c
hannel
ca
pacit
y)
co
rr
es
ponds
to
th
e
c
hanne
l’s
ba
ndwi
dth
.
H
ow
e
ve
r,
bas
ed
on
Figures
4
(a)
and
5
(a
),
it
c
an
be
see
n
t
ha
t
the
ch
an
nels
are
bu
sy
m
os
t
of
the
tim
e
fo
r
sm
al
l
value
of
I
dle
pro
bab
il
it
y
(
P
I
=
0
.
3
)
du
e
to
high
pr
i
m
ary
us
er
tra
f
fic
load
.
Als
o,
unde
r
this
value
of
P
I
,
it
was
seen
tha
t
the
PO
S
’s
thr
ough
pu
t
pe
rfo
r
m
ance
was
sim
il
ar
to
that
o
f
MAS
A
sche
m
e
by
16
.4
%
,
wh
ic
h
ou
t
pac
ed
eve
n
MDR
an
d
RS
schem
es
by
53.
4%
a
nd
92.
7%
,
res
pecti
vel
y,
w
hile
with
reg
a
rds
to
PDR
,
it
was
by
15.
8%
,
124%
a
nd
168.5%
resp
ect
i
ve
ly
.
Fr
om
Figu
r
es
4
(
b
,
c
)
an
d
Figures
5
(
b
,
c),
it
can
be
s
een
that
as
the
idle
pro
bab
il
it
y
ro
s
e
(
P
I
=
0.6
an
d
0.9),
the
th
rou
ghpu
t
perform
ance
of
t
he
P
OS
schem
e
was
be
tt
er
than
MA
S
A,
MDR
an
d
SR
by
27.
9%,
88.
7%
a
nd
14
9.1
%
at
P
I
=
0.6
a
nd
by
35.
1%,
85.
1%
a
nd
14
8.5%
at
P
I
=
0.9.
F
or
PD
R
pe
rfo
rm
a
nce,
the
POS
s
chem
e
ou
tc
la
ssed
at
P
I
=0.6
by
18
.
7%,
213.
8%
an
d
29.16
%
and
at
P
I
=0.9
by
22.81%
,
217.5
%
an
d
326.0
2%
,
re
sp
ect
ivel
y.
This
c
ould
be
due
to
in
crease
in
P
I
val
ue
that
i
ncr
ea
ses
the
pro
ba
bili
ty
of
su
it
able
c
ha
nn
el
s
bein
g
a
vaila
ble
to
CR
us
ers
f
or
t
rans
m
issi
on
as
the
PU
tra
ff
ic
l
oad
is
at
a
lowe
r
le
vel.
6.2.
Perfo
r
ma
nce
eva
lu
ati
on
of
St
ei
ner
mi
ni
mal
tree
(
SMT
)
multic
ast
CRN
un
der
t
he
im
pa
c
t
of
pa
cke
t
siz
e
Figures
6
(a
-
c
)
an
d
Fig
ur
es
7
(a
-
c
)
prese
nt
the
thr
ough
pu
t
an
d
P
DR
perform
ance
per
ta
inin
g
t
o
m
ul
ti
la
ye
r
m
u
lt
i
-
hop
m
ult
icast
CR
N
wh
e
n
com
par
ed
with
pac
ket
siz
e
with
var
io
us
kinds
of
c
hannel
assignm
ent
schem
es
as
well
as
three
va
lues
pe
rtai
ni
ng
t
o
I
dle
pro
ba
bili
ty
[
P
I
=
0
.
3
,
0
.
6
,
a
nd
0
.
9
]
,
re
sp
ect
ively
.
T
he
fig
ures
in
dicat
e
that
the
PD
R
pe
rfor
m
ance
and
t
hroug
hput
droppe
d
as
the
value
of
pa
cket
siz
e
D
r
os
e.
T
his
ha
pp
e
ns
si
nce
to
tra
ns
m
it
the
pac
ket
dat
e
of
CR
us
e
r
e
ff
ect
ively
,
hi
gh
avail
abili
ty
of
ti
m
e
for
the
cha
nn
el
is
req
uire
d
fo
r
high
val
ue
of
D
,
w
hich
m
akes
it
diff
ic
ult
to
find
t
he
best
c
hannel.
Figures
6
(
a
-
c
)
in
dicat
e
that
t
he
e
nhance
d
t
hro
ughp
ut
gai
ns
f
or
P
OS
sche
m
e
as
agai
ns
t
MASA,
M
DR
and
RS
schem
es
at
P
I
=
0
.
3
,
0.6
a
nd
0.9
a
re
[
12.5%,
56.6%
and
10
2.1%],
[
16.9%,
100.5
%
an
d
155.3%
]
and
[
26.
2%
,
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
1693
-
6930
TELK
OMN
IKA
Tel
ec
omm
u
n
C
om
pu
t El
Con
t
ro
l
,
V
ol.
18
,
No.
1
,
Fe
bruary
2
02
0:
37
-
50
44
93.7%
and
170.8%],
res
pecti
ve
ly
.
Acco
r
ding
to
Figu
re
s
7
(
a
-
c),
the
e
nh
a
nc
ed
PD
R
gai
ns
fo
r
P
OS
sc
he
m
e
as
against
M
ASA
,
MDR
a
nd
R
S
sc
hem
es
at
P
I
=
0
.
3
,
0.6
a
nd
0.9
a
r
e
[
16.5
%
,
118.6%
a
nd
183.2
%]
,
[
14.3
%
,
213%
an
d
29
8.6%]
a
nd
[
23
.4
%
,
228.9%
and
367.2%],
resp
ect
ively
.
Also
,
with
ris
e
in
the
val
ue
of
P
I
,
the P
OS
s
chem
e’s per
form
ance i
m
pr
oves a
s t
he
P
U’s tra
ff
ic
load
is
lo
w
at
hi
gh
value
of
P
I
.
Figure
4. Th
r
ough
pu
t
vs cha
nnel
b
a
ndwi
dth
:
(a)
PI =
0.3, (b)
PI
=
0.6
,
(
c
)
PI
=
0.9
Figure
5. P
DR
vs
c
ha
nn
el
ba
ndwi
dth
:
(
a)
PI
=
0.3,
(
b). PI
=
0.6
,
(
c)
PI
=
0.9
(a)
(a)
(b)
(b)
(c)
(c)
Evaluation Warning : The document was created with Spire.PDF for Python.
TELK
OMN
IKA
Tel
ec
omm
u
n
C
om
pu
t El
Con
t
ro
l
Perf
orma
nce
analysis
of
mu
lt
il
ayer
mu
lt
ic
as
t
MANET CRN
ba
s
ed
on s
te
in
er
…
(
Basma
Naza
r
Na
dhim
)
45
Figure
6. Th
r
ough
pu
t
vs data
pack
et
siz
e
:
(
a)
PI
=
0.3,
(
b) P
I
=
0.6,
(
c
) PI
=
0.9
Figure
7. Th
r
ough
pu
t
vs data
pack
et
siz
e
:
(
a)
PI
=
0.3,
(
b)
PI
=
0.6,
(
c
) PI
=
0.9
6.3.
Perfo
r
ma
nce
eva
lu
ati
on
of
steiner
mi
nimal
tree
(SMT)
multic
ast
CRN
unde
r
th
e
eff
ect
of
incr
eased
t
he
prim
ar
y
ch
annel
s nu
m
ber
Figures
8
(a
-
c
)
and
Fig
ures
9
(a
-
c
)
pr
e
sent
th
e
throu
ghput
as
well
as
PD
R
perform
ance
of
m
ult
il
ay
er
m
ul
ti
-
hop
m
ulti
cast
CR
N
as
com
par
ed
wit
h
the
num
ber
of
pri
m
ary
chan
nels
al
ong
with
var
i
ou
s
ki
nd
s
of
channel
assi
gnm
ent
schem
es
as
well
as
th
re
e
values
of
I
dl
e
pro
bab
i
lity
[
P
I
=
0
.
3
,
0
.
6
,
a
nd
0
.
9
]
,
resp
ect
ively
.
The
PD
R
pe
rfor
m
ance
an
d
t
hro
ughput
e
nh
anced
as
t
he
num
ber
of
pr
im
ary
cha
nnel
s
r
os
e.
This
is
due
t
o
increase
i
n
the
chance
of
rai
sing
t
he
num
ber
of
a
vaila
ble
channel
per
ta
ining
t
o
CR
use
r
as
pr
ese
nt
ed
i
n
Figures
8
(a
-
c
)
an
d
9
(a
-
c
),
r
especti
vely
.
T
he
e
nh
a
nce
d
t
hro
ughput
gai
ns
for
PO
S
sc
hem
e
in
com
par
iso
n
to
MASA,
M
DR
an
d
RS
sche
m
es
at
P
I
=
0
.
3
,
0.6
a
nd
0.9
are
[
13
.6
%
,
14.4
%
an
d
47%]
,
[
23.
3%
,
26.1
%
an
d
68.0%]
and
[
24.7%,
32.5
%
and
73.
3%]
resp
ect
ively
.
The
e
nh
a
nce
d
PD
R
gains
for
P
OS
sc
he
m
e
in
com
par
ison
to
MAS
A,
M
D
R
an
d
RS
sch
e
m
es
at
P
I
=
0
.
3
,
0.6
and
0.9
are
[
13.
3%,
44.
96%
an
d
82.1%]
,
[13.9%,
68.8%
and
116.5%] a
nd [1
4.6%, 9
1.5% a
nd
138.7
%]
, r
es
pecti
vel
y.
(a)
(a)
(b)
(c)
(c)
(d)
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
1693
-
6930
TELK
OMN
IKA
Tel
ec
omm
u
n
C
om
pu
t El
Con
t
ro
l
,
V
ol.
18
,
No.
1
,
Fe
bruary
2
02
0:
37
-
50
46
Figure
8. Th
r
ough
pu
t
vs n
umber
of c
hannels
:
(
a)
PI
=
0.3,
(
b)
PI
=
0.6,
(
c
)
PI
=
0.9
Figure
9. P
DR
vs
nu
m
ber
of c
hannels
:
(
a)
PI
=
0.3,
(
b) P
I
=
0.6,
(
c
) PI
=
0.9
6.4.
Perfo
r
ma
nce
eva
lu
ati
on
of
steiner
mi
nimal
tree
(SMT)
multic
ast
CRN
unde
r
th
e
eff
ect
of
transmi
ssion
po
w
er
Figures
10
(a
-
c)
an
d
Fi
gure
s
11
(a
-
c)
pre
sent
the
th
r
oughput
as
well
as
PD
R
perf
or
m
ance
of
m
ul
ti
la
ye
r
m
ul
ti
-
hop
m
ulti
ca
st
CR
N
wh
e
n
com
par
ed
with
transm
issi
on
powe
r
al
ong
with
va
rio
us
ki
nd
s
of
channel
assignm
ent
sche
m
es
as
well
as
thre
e
values
pe
rtai
ning
to
idle
pr
ob
abili
t
y
[
P
I
=
0
.
3
,
0
.
6
,
an
d
0
.
9
]
,
res
pecti
vely
.
W
it
h
a
rise
i
n
t
he
tra
ns
m
is
sion
power,
the
PD
R
perform
ance
an
d
thr
oughput
en
han
ce
d.
T
his
is
becau
se
the
r
e
is
a
decr
ease
in
the
tim
e
need
e
d
f
or
data
transm
issi
on
,
thu
s
al
lowing
m
or
e
data
to
be
tra
ns
m
itted
ov
e
r
each
c
hannel.
Also
,
in
th
e
ca
se
of
hi
gh
val
ue
of
i
dle
pro
ba
bili
ty
,
there
is
e
nhanc
e
m
ent
in
the
pe
rfor
m
ance
of
PO
S
schem
es
as
m
entioned
i
n
the
earli
er
se
ct
ion
s.
The
en
han
c
e
d
thr
oughput
gai
ns
f
or
the
POS
schem
e
as
ag
ai
ns
t
MAS
A,
MDR
an
d
RS
schem
es
at
P
I
=
0
.
3
,
0.6
a
nd
0.9
a
re
[7.5
%
,
49
.
1%
and
74.
4%],
[15.1%,
56.
9%
and
110.2
%]
and
[
24.
9%
,
62.
1%
a
nd
148.9%],
re
s
pecti
vely
.
The
e
nh
a
nce
d
PD
R
gains
for
PO
S
schem
e
in
com
par
is
on
t
o
MA
SA,
MD
R
and
R
S
sch
e
m
es
at
P
I
=
0
.
3
,
0.6
and
0.9
are
[6.9
%
,
122.2
%
and
158.5
%]
,
[12.9%,
151.7%
a
nd
232.1%]
an
d
[20
.
8%,
165.7
%
an
d
292.7%],
res
pe
ct
ively
.
(a)
(a)
(b)
(b)
(c)
(c)
Evaluation Warning : The document was created with Spire.PDF for Python.