TELKOM
NIKA
, Vol. 13, No. 4, Dece
mb
er 201
5, pp. 1233
~1
241
ISSN: 1693-6
930,
accredited
A
by DIKTI, De
cree No: 58/DIK
T
I/Kep/2013
DOI
:
10.12928/TELKOMNIKA.v13i4.2341
1233
Re
cei
v
ed
Jul
y
16, 201
4; Revi
sed O
c
tob
e
r 24, 201
5; Acce
pted No
vem
ber 5, 20
15
A Novel Approa
ch to Optimize Cognitive Radio Network
Utilization using Cascading Technique
H Venka
t
es
h
Kum
a
r*
1
, MN Giriprsad
2
1
Nagarj
u
n
a
Col
l
eg
e of Engi
ne
erin
g and T
e
ch
nol
og
y, De
part
m
ent of ECE,
Bang
alor
e, Indi
a
2
JNT
U
College
of Engine
eri
n
g
Anatapur, D
e
p
a
rtment of ECE, Andhr
a Pra
desh, Indi
a
e-mail: ve
nkate
s
hkumar.h
@g
mail.com
A
b
st
r
a
ct
T
he co
gnitiv
e
r
adi
o ex
per
ienc
es l
a
ck of
cha
nne
ls. T
h
e
Un
li
cense
d
cl
ie
nts
(UU) ac
qu
ire c
han
ne
l
s
from the l
i
ce
ns
ed users (
L
U). But, w
hen the LUs ne
ed
the
m
, the UUs n
e
ed to stop the
i
r
transmiss
io
n an
d
han
dle
the
ch
ann
el
back
to
the L
U
s. T
h
is
outco
me
is
de
l
a
y i
n
tra
n
s
m
is
sion
of
infor
m
ation
by
UUs.
W
e
prop
osed
a
n
a
l
gorith
m
bas
ical
ly ce
nt
erin
g to
me
et the
d
e
lay
,
by uti
l
i
z
i
n
g
a
data c
a
sca
din
g
strategy w
her
e
the inf
o
rmatio
n
of the
UUs
are store
d
in
the i
n
ter
m
e
d
i
a
te n
odes
so
that eve
n
if th
e trans
miss
io
n
is
interru
pted, th
e
data
is
not l
o
st
. T
he ex
peri
m
e
n
tal r
e
su
lts de
mo
nstrate
th
at prop
os
ed
fra
m
ew
ork is s
uper
i
o
r
to other existin
g
framew
orks.
Ke
y
w
ords
: co
gnitiv
e
radi
o ne
tw
ork, delay, data cascad
i
n
g
Copy
right
©
2015 Un
ive
r
sita
s Ah
mad
Dah
l
an
. All rig
h
t
s r
ese
rved
.
1. Introduc
tion
A co
gnitive
radio
is an
intelligent
ra
dio
that ca
n
be cu
stomi
z
ed
an
d co
nfigure
d
dynamically. Here the
rad
i
o distin
gui
sh
es the
chan
nel by itself
and a
c
co
rdin
gly cha
nge
s
its
para
m
eters [
1
]. Paramete
rs li
ke
routin
g, freque
nc
y
etc. are bala
n
ce
d by the
cog
n
itive radi
o a
s
requi
rem
ent. Cog
n
itive rad
i
os continu
o
u
s
ly cont
ro
l
s
its own pa
rticu
l
ar a
c
tion an
d
focus the
ra
dio
freque
ncy e
n
v
ironme
n
t, chann
el condi
tions, lin
k p
e
rform
a
n
c
e,
and
so forth
and alte
rs
the
setting
s to d
e
liver the
req
u
i
red
quality of
se
rvice
subj
ect to a
n
a
p
p
r
op
riate
com
b
ination
of u
s
er
requi
rem
ents,
ope
rational li
mitations, a
n
d
re
gulato
r
y con
s
trai
nts. It has th
e
cap
a
city to pe
rce
i
ve
the out
side
e
n
vironm
ent a
nd u
s
e
s
a
r
tificial intelli
gen
ce to
lea
r
n from envi
r
onm
ent, and
ma
kes it
s
inner state
a
d
just to th
e
meas
urable
cha
nge
s of
receive
wi
rele
ss si
gnal
s by
cha
ngin
g
so
me
para
m
eters i
n
ord
e
r to
ackno
w
le
dge hi
gh reli
able
i
n
any pla
c
e a
nd the p
o
we
rful utilization
of
spe
c
tru
m
re
source
s [2]. A
Cog
n
itive Ra
dio co
nsolida
t
es num
ero
u
s sou
r
ces of d
a
ta, decid
es i
t
s
pre
s
ent
wo
rki
ng setting
s,
a
nd colla
borates with othe
r
cog
n
itive ra
di
os
(CR) in
a
wirel
e
ss [3].
The
promi
s
e of cognitive radios is
enhanced utilization of spectrum
resources, redu
ced engineeri
n
g
and
plan
ning
time, an
d a
daptation
to
curre
n
t op
era
t
ing conditio
n
s
.
Qualities of cognitive radio
sy
st
em
s:
-
Determine th
eir own environment an
d react.
-
Pr
o
v
id
es
r
o
bu
s
t
s
e
r
v
ic
es
-
Operational
state langua
ge
s influen
ce
g
eneral network archite
c
ture
s.
In a cognitiv
e
network th
e ra
dio
spe
c
t
r
um
sp
a
c
e
s
i
s
really
scare and
a
r
e n
e
ed to b
e
reu
s
ed. Th
e use
r
s
with the cha
nnel
s are calle
d as
licensed u
s
er
(LU) a
nd tho
s
e without lice
n
se
is
c
a
lled
th
e un
lic
en
se
d
u
s
er
(UU
)
. Th
e
un
lic
en
se
d
u
s
er
w
a
its fo
r
the c
h
a
n
n
e
l
to
be
fr
ee
fr
om th
e
licen
se
d u
s
e
r
. One
s
it i
s
f
r
ee th
e
cell t
hat is re
sp
on
sible
for
su
p
p
lying
cha
n
n
e
ls to
the
UUs
inform them
about its avai
lability [4]. The cell
s un
dergo group fo
rmation unt
il it achieve
s
Na
sh
stable
con
d
ition [5]. When
it reache
s this condi
tion it
can b
e
come i
ndep
ende
nt enou
gh to se
rve
its UUs with t
he chan
nel
s. But duri
ng t
he chan
nel
s
being
used b
y
the UU
s, if
the LUs
app
ear
again
then, t
he
UUs
hav
e to
swit
ch f
r
om th
e
li
cen
s
ed
chan
nel
to othe
r unli
c
ensed
ch
ann
el.
Hen
c
e th
e co
ntinuou
s data
transmissio
n
is interrupt
e
d
and
re
sults in delay [6] [7]. In this pa
pe
r
we have p
r
op
ose
d
a model
that will meet t
he delay by mean
s of ca
scadi
ng techni
que
s.
The rem
a
inin
g pape
r is de
scribe
d as fol
l
ows
.Section
2 describe
s
the literature
survey
followe
d by t
he p
r
op
osed
system
in
se
ction
3.
Secti
on 4
tells ab
out the
syste
m
mod
e
l follo
we
d
by the experi
m
ental re
sult
s in se
ction 5.
Finally the paper i
s
co
ncl
uded by secti
on 6.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
A Novel Ap
proach to Optim
i
ze Cogniti
ve Radi
o Net
w
ork
Utilizatio
n
…
(H Ve
nkatesh Kum
a
r)
1234
2. Literature
Sur
v
e
y
In [1], intrin
sic prope
rtie
s a
nd
cu
rre
nt
re
sea
r
ch
chall
enge
s
o
f
the CRAHNs are
pre
s
ente
d
. F
i
rst, novel
spectrum m
a
nagem
ent fu
nc
tion
alities su
ch as
sp
ectru
m
sen
s
i
ng,
spe
c
tru
m
sh
aring, a
nd
spectrum d
e
ci
sion, a
nd
sp
ectru
m
mobil
i
ty are introd
uce
d
from t
he
viewpoi
nt of
a net
work
re
quirin
g
di
stri
buted
coo
r
di
nation. A p
a
r
ticula
r e
m
ph
asi
s
i
s
given
to
distrib
u
ted coordi
nation b
e
twee
n CR use
r
s th
rou
g
h
the establi
s
hme
n
t of a commo
n co
ntrol
cha
nnel.
In [2], they have identified
and a
ddressed th
re
e
fun
damental ch
alleng
es en
countered
spe
c
ifically b
y
mobile SUs. First they m
odele
d
chan
n
e
l availability
experie
nced
by a mobil
e
SU
as a two
-
sta
t
e continu
o
u
s
-time M
a
rko
v
chain (C
T
M
C) a
nd v
e
r
i
f
y
it
s accu
ra
cy
v
i
a in-de
p
t
h
simulatio
n
. Seco
ndly to protect prim
ary
/
incumb
ent
communi
catio
n
s from S
U
interferen
ce, they
introdu
ce
d gu
ard di
stan
ce i
n
the sp
ace d
o
main a
nd fin
a
lly derive th
e optimal gu
a
r
d di
stan
ce th
a
t
maximize
s th
e spatiotem
p
oral spe
c
trum
opportu
ni
ties available to mobile cogniti
ve radio
s
.
Fan [6] expl
o
r
es the
overh
eard
info
rmat
ion,
in
cludin
g
data
req
u
e
s
t and
data
re
ply, to
optimize
ca
che pla
c
eme
n
t and ca
ch
e di
scovery. To t
he be
st of kn
owle
dge, this is the first work
that co
nsi
ders the
overhe
aring
p
r
ope
rt
y of wi
rele
ss com
m
uni
cat
i
ons in
data
ca
chi
ng. Th
is
prop
osed sy
stem wa
s able
to redu
ce the
message
co
st and al
so th
e acce
ss d
e
la
y.
In our p
r
evio
us
wo
rk,
we
have fo
cu
sed
on
gr
o
up fo
rming
schem
e
in
cog
n
itive radio fo
r
s
p
ec
tr
um
sen
s
ing. At the point when the
data shoul
d
be tran
smitte
d con
s
eq
uent
ly then in such
cases cogniti
ve radio sy
st
em is utilized. The transmitters in
the subjective sy
stems are signal
ed
su
ch that the
y
can
con
s
e
quently re
co
gnize
the
a
c
ce
ssi
ble cha
nnel
s
and pa
ss sig
nal
throug
h
t
hem.
I
n
su
ch
a ci
rc
umst
a
n
c
e a
c
ce
ssi
bili
t
y
of
c
han
nel
s a
s
sume
s th
e esse
ntial p
a
rt. As
cha
n
n
e
ls
are
gen
erally limited
so th
ey nee
d to
b
e
shar
ed
am
ong th
e p
a
rti
c
ipatin
g u
s
e
r
s. Th
e u
s
e
r
who
own
ch
ann
el
s they
could
call thei
r o
w
n
are
name
d
a
s
Li
cen
s
e
d
Users
(L
U), a
n
d
those do
es
not
posse
ss cha
nnel
s they
co
uld call thei
r
own, n
eed
to
utilize
ch
ann
els
claim
ed b
y
the autho
ri
zed
client
s are
ca
lled unlicen
se
d client
s (UU). Cells ca
n
gi
ve restri
cted
sup
port of
its UUs, in light of
the fact that
a sin
g
le
cell
is capa
ble ju
st for a
little
numbe
r of L
U
s. So to
give exact d
a
ta
in
rega
rdi
ng to t
he
cha
nnel
a
c
cessibility th
e cells
nee
d
to take
in the
insi
ghts
of the
con
s
ide
r
a
b
le
numbe
r
of L
U
s. So
the
r
e
is
a requi
re
ment for cell
s to
gro
up f
o
rmin
g
with
one
anothe
r
and
stru
cture a ga
thering to sha
r
e the dat
a of
the con
s
ide
r
able num
be
r of LUs.
3. Proposed
Work
The co
ntrib
u
tions a
nd si
gni
fican
c
e of this work are:
a) Cha
nnel
Sen
s
ing
b) Grou
p
formin
g
c)
Nash stability
Cha
nnel avail
ability is characteri
ze
d as t
he lik
elih
ood
of an authori
z
ed cha
nnel b
e
ing acce
ssib
le
for the
corre
s
po
nde
nces
of unli
c
en
se
d
clie
nts.
C
h
a
nnel
availabil
i
ty is a
key p
a
ram
e
ter for
an
effective de
si
gn of
ch
ann
el
determinatio
n metho
d
s
an
d ad
ditionally
routin
g in
co
gnitivetive ra
dio
system
s. In
static situ
ation
s
, the
availa
bilit
y of a
ch
annel
de
pen
ds
only o
n
th
e p
r
ima
r
y u
s
er
׳
s
activity. In mobile scen
ari
o
s, the
availa
bility of a channel dynami
c
ally varies in time due to the
cha
nge
s of the use
r
s
׳
relati
ve position
s
. Figure1 sho
w
s the wo
rkflow of cha
nnel
availability [8].
For fin
d
ing
the p
r
ofit-lo
ss state i
n
cog
n
itiv
e
radi
o spectrum se
n
s
ing, a
g
r
o
u
p
formin
g
approa
ch i
s
p
r
esented
he
re in
Figu
re
2. Thi
s
a
pproa
ch is ba
se
d o
n
the ‘g
ro
up’
concept. Initiated
cog
n
it
iv
e rad
i
o cr
eat
e
s
st
ruct
u
r
e diffe
rent ‘gro
up
s’
as pe
r so
me crite
r
ia.
Every gathering
comp
ri
se
s of
a he
ad
nod
e, som
e
g
a
te
way no
de
s a
nd no
rmal
no
des.
Du
ring
detectio
n
sta
ge,
node
s
sen
s
e
cha
nnel
s a
nd arra
nge t
he chan
nel
and
slot a
s
signment; du
ri
ng informatio
n
transmitting stage, every n
ode h
a
s th
e capa
city co
rre
s
po
nd
with its neigh
bor
no
des i
n
a
ssi
gn
ed
cha
nnel
s du
ri
ng given time
slots [9]. Thu
s
, node
s
can
be asso
ciate
d
togethe
r to frame a sy
ste
m
without b
a
si
c
cha
nnel
s a
n
d
the sy
st
em n
e
twork i
s
ma
de st
ride
s. Th
is stu
d
y intro
duces th
e g
r
o
u
p
forming p
r
o
c
edure incl
udi
ng neigh
bor
node dete
c
ti
on, node co
mmuni
cation,
group form
a
t
ion
,
time openin
g
task a
nd time
slot assignm
ent.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 13, No
. 4, Decem
b
e
r
2015 : 123
3 – 1241
1235
Figure 1. Availability of cha
nnel
Figure 2. Gro
up formin
g in cog
n
it
ive radi
o for profit –lo
ss
state
Figure
3 represents achieving
the
Nash stability usi
ng group
formation. Thi
s
situation
utilizes a non-cooperative
game
model
to concentrate on the
opposition among
primary
users
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
A Novel Ap
proach to Optim
i
ze Cogniti
ve Radi
o Net
w
ork
Utilizatio
n
…
(H Ve
nkatesh Kum
a
r)
1236
with the Nash Equilibri
um
(NE) as the best ar
rangement. The
primary users first
set the bid
techni
que
s in
cludi
ng the
spectrum
si
ze
and
the
co
st, and after that present t
heir
bid
s
to t
h
e
auctio
nee
r. A
fter gath
e
rin
g
all th
e d
a
ta f
r
om th
e
prim
ary u
s
e
r
s,
the
au
ctione
er gi
ves fee
dba
ck to
every prim
ary user. An it
erative calcul
ation is
utilized which permits every primary user to
maximize th
ei
r reve
nue by
updatin
g their bid from th
e joining of the
NE. The
co
st of the NE poi
nt
can b
e
viewe
d
as the tru
e
spe
c
tru
m
value.
Figure 3. Nash Stability
The Netwo
r
k is used by its use
r
s in trans
mitting
an
d re
ceiving o
f
data amon
g them
throug
h th
e
chann
els. A
chann
el i
s
a
route from
a t
r
an
smitting
u
s
er to
a
re
cei
v
ing u
s
e
r
. All
the
use
r
s d
o
not have a cha
n
n
e
l of their own, they
need to be sha
r
ed.
The ch
ann
el holde
r users can
any time use
the ch
ann
el
s to tra
n
smit
and rece
ive
their data.
The u
s
e
r
s
wi
thout their o
w
n
cha
nnel
s hav
e to borrow t
he ch
ann
els
from the ch
a
nnel hold
e
rs.
In our discu
ssi
on we have
named th
e chann
el hold
e
r use
r
s
as th
e
licen
se
d us
e
r
s
(L
U) a
nd t
he u
s
ers
with
out ch
ann
els
as
the unli
c
en
se
d users (UU).
One
s
the
ch
annel i
s
fre
e
from the u
s
e
of the ch
ann
el hold
e
r, it can
be u
s
ed
by th
e non
-cha
nne
l holde
rs. Each cell
is
re
sp
onsi
b
le fo
r se
rvicing
a fe
w
numbe
r of
no
n-
cha
nnel
hold
e
rs.
The
cell
s provid
e info
rmation to the
non
-chan
nel
holde
rs ab
o
u
t the availa
b
ility
and non
-avail
ability of the
cha
nnel
s. Any cell is
capa
ble of providi
ng informatio
n only about few
of the
cha
nne
l holdi
ng u
s
e
r
s. In o
r
d
e
r to
obtain th
e inf
o
rmatio
n of
al
l the u
s
e
r
s th
e cells ne
ed t
o
sha
r
e the
kno
w
led
ge of oth
e
r cells too. F
o
r that t
he ce
ll tries formi
n
g gro
up with
other
cell
s. It
is
the pe
rsonal
deci
s
io
n of
other cells whe
t
her to
a
llo
w
the cell to
group
with it
or not. Fo
r
ea
ch
grou
ping
the
cell
s ma
ke
sure th
at the
a
ddition
of
the
ne
w
cell
sho
u
ld n
o
t bri
ng
it to any
kind
of
loss. Fo
r eve
r
y groupin
g
t
he p
a
rticular
cell th
at s
eek friend
shi
p
from othe
r
cell
s ma
ke
sure
that
its payoff increases e
a
ch time it group
s
with othe
r cel
l
s. So whe
n
e
v
er a ch
ann
el is free the cell
tells its re
spe
c
tive non-ch
a
nnel
hol
ders
about its avail
ability.
If the information is true th
en the non
-chann
el use
r
use
s
the cha
nnel for tra
n
smitting its
data. If the in
formation i
s
false the
n
the non-ch
ann
el use
r
tran
smit
s its data an
d
the informati
on
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 13, No
. 4, Decem
b
e
r
2015 : 123
3 – 1241
1237
gets lo
st. So the cell keep
on formin
g group
s until an
d unle
ss it wil
l
rea
c
h a situ
ation wh
ere t
h
e
cell
will be having the inf
o
rmation regarding all
the users
having channel
s. Thus
a
situation
arrive
s wh
ere
the cells be
come self
-dep
ende
nt
and a
r
e able to se
rve its UUs
wi
thout any ones
help. Thi
s
ultimate situatio
n is called a
s
the Na
sh stable state.
One
s
the cel
l
is Na
sh
sta
b
le
coalitio
n form
ation is en
de
d and
the
cell
is a
b
le fo
r
su
pplying of th
e
ch
annel
s to
the UUs
witho
u
t
grou
ping
with
other cells [1
0].
The
UUs whi
l
e u
s
ing
the
cha
nnel
s
suff
er
delay
s in
transmitting t
heir data
du
e to th
e
appe
ara
n
ce
of lice
n
sed
u
s
ers. So
in
orde
r to
avoi
d the
dist
urb
ance in
the
cha
nnel
traffi
c,
whe
never an
y licen
sed
u
s
er a
gain
nee
ds to
acce
ss
the chann
el
spectrum th
en
the unli
c
e
n
sed
use
r
s
have t
o
vacate th
e
spe
c
tru
m
an
d let it free for the li
cen
s
ed u
s
er to
use the chann
e
l
for
transmitting its data [11].
For u
s
in
g the
cha
nnel
always the li
cen
s
ed
user
g
e
ts the first pre
f
eren
ce b
e
ca
use th
ey
are th
e regi
stered
u
s
ers fo
r the
ch
annel
s o
r
they
o
w
n
the chan
nel.
Since
be
cau
s
e of the
su
dd
en
appe
ara
n
ce
of the lice
n
sed u
s
e
r
the
unlicen
sed
u
s
er ha
s to
surrend
er the
cha
nnel fo
r the
licen
se
d u
s
e
r
. In doin
g
su
ch th
e o
ngoi
ng d
a
ta tra
n
s
missio
n by t
he u
n
licen
se
d u
s
er lea
d
s to
some
ki
nd
of interruptio
n.
As a
re
sult t
he
contin
uou
s flo
w
of th
e
data i
s
l
o
st.
The l
o
ss
of
the
continuous data transmi
s
sion resu
lts in
some ki
nd of
delay. Thus
t
he transmi
s
si
on del
ay will
be
signifi
cantly increa
sed. Th
e delay
co
nst
r
aints in
cog
n
i
tive network are toug
h to meet. Our m
a
in
con
c
e
r
n i
n
th
is p
ape
r i
s
to
meet thi
s
d
e
l
ay co
nstraint
. Delay
with
respe
c
t to th
e fadin
g
is th
at
whe
n
a
sign
al travel
s i
n
all di
re
ctions
& m
eets the
receiver ante
nna, the
a
n
te
nna
will
get t
he
same
sig
nal
s from differen
t
paths. So a
s
a
re
sult of
this some
sig
nal which are
re
ceived
after
some
fra
c
tion
of se
co
nd,
will have
so
me del
ay. So
all the
comb
ined affe
ct is the cau
s
e of
the
sign
al fading.
Those that receive the si
gnal
s lately are ca
ll
ed the delayed cand
idates. To av
oid this
delay difficult
y we rep
r
esent the o
c
cu
rre
nce of
th
e LUs in
a
contin
uou
s
chain. Applyin
g
ca
scadin
g
wil
l
result in con
t
rolling the da
ta access del
ay.
4. Sy
stem Model
Our cognitive
network con
s
ist
s
of
s
e
t of total
numb
e
r of licen
se
d
users an
d
s
e
t
contai
ning
numbe
r of unli
c
en
se
d users.
Let
be the set of cascadi
ng node
s. Th
e data is
tran
smitted from
the sou
r
ce to the
receiver th
ro
ugh the
ca
scading
node
s.
The tra
n
smission
co
st is
g
i
ven as
. Where
is the
traffic
rate,
is the transmi
ssion g
r
ap
h an
d
is the total edge
s in
.
The entry cost is given as
(1)
whe
r
e
is th
e data entry prob
ability for node
2
,
is the round trip cost from
2
to a
ca
scadin
g
no
de and
is the roundt
rip del
ay from node
2
to the casca
d
ing no
de ne
xt to it.
For a no
de
2
the delay con
s
traint i
s
given by
(2)
Whe
r
e
is the entry delay for node
2
.
is the thresh
old for
delay, and
is
the level of trus
t.
The delay co
nstrai
nt pro
b
l
e
m can b
e
mi
nimize
d by minimizin
g
the total delay i.e.
(3)
whe
r
e
is a cost ratio, and
focu
sing o
n
f
o
r all nodes
.
Dela
y
approach:
Let
be the sh
ortest path b
e
t
ween a
n
y tw
o node
s. The
data entry del
ay from
1
to
its next casca
d
ing no
de is
2
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
A Novel Ap
proach to Optim
i
ze Cogniti
ve Radi
o Net
w
ork
Utilizatio
n
…
(H Ve
nkatesh Kum
a
r)
1238
The time ta
ken by the
so
urce to rece
ive the re
que
st
is the
sho
r
te
st path,
leng
th.The
disa
bled no
d
e
s are usele
ss d
u
rin
g
the
transmi
ssio
n
of the reque
st from the reque
ster to the
sou
r
ce. Supp
ose
we have
to transmit a
requ
est from
node 1 to no
de 6 wh
ere n
ode 3 an
d 4 has
no di
re
ct
cha
nnel
s
sin
c
e th
ose
chan
nel
s are b
e
in
g
u
s
ed by th
e
LUs. So
nod
es 3
and
4
be
com
e
s
usel
ess, sin
c
e it does not
help in the
tra
n
smi
ssi
on to rea
c
h to the n
ode 6.
Figure 4. Links between th
e use
r
nod
es
durin
g tran
sm
issi
on
The vario
u
s
states from n1
to reach n6 a
r
e as follo
ws:
State1
n1, n2
State2
n1, n4
State3
n1, n3
State4
n1, n4,
n2
State5
n1, n2, n5
State6
n1,
n3, n6
State7
n1,
n2, n5, n6
The sh
orte
st time for which the reque
st
is r
eached a
t
the node 6 is calle
d the sho
r
test len
g
th
path. Let
us consi
der the
shorte
st path
l
ength from th
e re
que
ste
r
n
ode
n1 to
the
so
urce
nod
e
n6
be
.
m
i
n
0
:
0
(4)
If
is a functio
n
of
su
ch a
s
(5)
1
1
⁄
(6)
⁄
(7)
Whe
r
e
is the (1, n) th ele
m
ent in
th power of
th element.
To find the
from the requ
ester to diffe
rent ca
sc
adin
g
node
s, i.e. if the interme
d
iate
node
n5 i
s
a
ca
scadin
g
no
de an
d if n
5
and the
sou
r
ce
node
n6
b
o
th re
ceive
s
t
he requ
est from
the re
que
ster, then it is e
nded
and th
en the
stat
e
for the
sou
r
ce n
ode
will
be obtai
ned
by
mergi
ng all th
e states into
one final stat
e.
State1
n1, n2
State2
n1, n4
State3
n1, n3
State4
n1, n4,
n2
State5
n1,
n2, n5, n6
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 13, No
. 4, Decem
b
e
r
2015 : 123
3 – 1241
1239
The shortest
path will giv
e
us the value
of the entry delay because Entry
Delay
=
2(sho
r
test pat
h).
Thus th
e entry delay redu
ces to
2
(8)
whe
r
e
is the
delay thresh
old limit and
is
the level of trus
t
2
⁄
(9)
5. Experimental Re
sult
In the propo
sed model
we
have con
s
id
ered
a n
e
two
r
k topol
ogy inclu
d
ing 15 li
cen
s
e
d
use
r
s tran
smi
tters a
nd
15 l
i
cen
s
e
d
u
s
e
r
receivers a
n
d
about
72 u
n
l
i
cen
s
e
d
u
s
e
r
s. The
r
e
are
9
cell
s in
clude
d
whi
c
h
sha
r
e
the co
gnitive
netwo
rk.
He
re after the
sp
ectru
m
optimi
z
ation i
n
ord
e
r
to deal
with
the d
e
lay effe
ct we h
a
ve u
s
ed
t
he
ca
scading
techniq
ue. Th
e resul
t
s obtai
ned
from
the new p
r
op
ose
d
tech
niq
ue are pl
otted
as sh
own in Figure 5.
Figure 5. The
delay respon
se
s for the pr
opo
sed
syste
m
vs. the existing system
The Fig
u
re
6
descri
b
e
s
a
bout the del
ay ch
a
nge
compa
r
ison in
the Existing
system
without the cascadi
ng techniqu
e with the pro
p
o
s
ed
system whe
r
e the data ha
s bee
n savin
g
in
the casca
d
in
g no
de
s. Th
us f
r
om
the
grap
h it
i
s
seen th
at the
syst
em
whe
r
e th
e
ca
sca
d
in
g
techni
que i
s
involved suffe
rs lo
w delay t
hen the on
e without the cascadi
ng techniqu
e
Figure 6. Total co
st inclu
d
i
ng the entry co
st and the transmi
ssion
cost for both P
S
and ES
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
A Novel Ap
proach to Optim
i
ze Cogniti
ve Radi
o Net
w
ork
Utilizatio
n
…
(H Ve
nkatesh Kum
a
r)
1240
The Fi
gure 7
de
scribe
s
ab
out the total
co
st of
tran
smissi
on i
n
bit
s
p
e
r
se
co
nd
for b
o
th
with ca
scadin
g
a
s
well
a
s
without ca
scading
te
chniq
ue that i
n
cl
u
des
the starti
ng
tra
n
smissi
on
co
st
an
d t
h
e
t
o
t
a
l t
r
an
smi
s
sion
c
o
st
[1
2]. He
re from t
he g
r
ap
h it i
s
cle
a
r th
at th
e tran
smi
s
sio
n
is
co
stlier i
n
th
e existing
system i.e. the syst
e
m
wi
thout the
ca
scadin
g
tech
nique th
an t
he
prop
osed sy
stem i.e. the one with
the cascadi
ng techniqu
e involved.
Figure 7. Through
put vs. simulation tim
e
The final gra
ph de
scribe
s about the compa
r
ison of
the average
through
put
of the
prop
osed sy
stem and the existing syst
e
m
[13]. Here
also the p
r
op
ose
d
system
with ca
scadin
g
techni
que gi
ves better result in ca
se of th
rough
put than the existing system without the
ca
scadin
g
techniqu
e.
6. Conclusio
n
The interfe
r
e
n
ce
s faced
by the chan
nel us
ers in
a cognitive
network a
r
e
tried to
minimize in this p
ape
r. The problem i
s
mainly
suffered by
the unlicen
sed
u
s
er
to stop
t
hei
r
transmissio
n
and
han
dle th
e chan
nel to
l
i
cen
s
e
d
u
s
e
r
.
Since the
licensed
users
are
re
giste
r
e
d
to
the chan
nel
s so they always get the first prio
rity in using the ch
annel
s no m
a
tter even if
the
unlicen
sed
u
s
er’
s
tran
smi
ssi
on
nee
ds
to be i
n
terru
p
ted. Thi
s
re
sults in
delay
in tran
smissi
on.
Our
pap
er
propo
sed
a
ca
scadi
ng n
ode
algorith
m
wh
ere th
e d
a
ta i
s
store
d
a
nd
redu
ce
s th
e l
o
st
and del
ay. It is better tha
n
the other
stan
dard
system
s.
Referen
ces
[1]
F
Ak
y
i
l
d
iz, W
Y
Lee, a
nd K
R
Ch
o
w
d
hur
y.
“CRAHNs: C
ogn
itive ra
di
o
ad h
o
c n
e
t
w
or
ks”.
Ad Hoc
Netw
orks
. 200
9; 7(5): 810
–83
6.
[2]
AW
Min, KH Kim, JP Singh
, and KG Shi
n
. “O
pportunisti
c
Spectrum A
ccess for Mob
ile C
o
g
n
itiv
e
Radi
os”. in
IEEE INFOCOM
. 2011: 29
93-
300
1.
[3]
Peng J
un, Ji
a
ng Mi
ng
ya
ng,
Jian
g F
u
, Li
u
W
e
iron
g. "Acti
v
e coo
per
ation
-
a
w
are
spectr
u
m
resourc
e
alloc
a
tio
n
in c
ogn
itive ra
dio
net
w
o
rk". C
ont
ro
l Co
nferenc
e
(CCC), 201
3 32n
d Chi
nes
e. 2013: 6
409
,
641
4.
[4]
Guope
ng Z
h
a
ng, Kun Yan
g
, Qingson
g Hu, et.a. Bargai
nin
g
Game
T
heoretic F
r
ame
w
ork fo
r
Stimulati
ng Co
oper
ation
in W
i
reless C
oop
era
t
ive Multicast N
e
t
w
o
r
ks. In:
IEEE Co
mmun
ic
ation l
e
tters
.
201
2; 16(2): 20
8-20
11.
[5]
Shan-S
h
a
n
W
et.al. Primary User Emu
l
ation
Attacks
Anal
ysis for
Cogn
itive R
adi
o Net
w
orks
C
o
mmu
ni
ca
tion
.
T
E
LKOMNIKA Ind
ones
ia
n
Jour
nal
of
El
ectrical
Eng
i
n
eeri
n
g
. 201
3; 11(7): 39
05-
391
4.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 13, No
. 4, Decem
b
e
r
2015 : 123
3 – 1241
1241
[6]
X F
an, J Ca
o, and W
W
u
. “Desig
n an
d Perf
orm
anc
e Eval
u
a
tion of Overh
eari
ng-a
i
de
d D
a
ta Cach
in
g
in Wireless A
d
Hoc Net
w
orks
”.
IEEE
Transactions on P
a
rallel and Distributed System
s
. 2
013; 24(
3):
450
–4
63.
[7]
Shans
ha
n W
a
ng, Ju
nsh
an Z
han
g, L
ang
T
ong. "
Delay Analysis f
o
r Cognit
ive
Ra
dio
N
e
tw
orks w
i
th
Ran
d
o
m
Acces
s
: A F
l
uid Queue View
". INFOCOM, 2010 Procee
din
g
s IEEE. 2010; 1(9).
[8]
Sun Yon
g
*, Qi
an Jia
n
sh
eng.
“Cogn
itive Ra
dio
C
han
ne
l Selecti
on Strate
g
y
B
a
sed
on
Exp
e
ri
enc
e
W
e
ighte
d
Attra
c
tion L
ear
nin
g
”
.
T
E
LKOMNIKA Indo
nesi
an
Journ
a
l
of Ele
c
trical En
gi
nee
ring
. 20
14
;
12(1): 14
9 ~
15
6
[9]
Z
hang
D,
et al. A
n
Impr
oved
Co
gn
itiv
e R
adi
o S
p
e
c
trum Sens
in
g Al
gorithm.
TEL
K
OMNIKA
Indon
esi
an Jou
r
nal of Electric
al Eng
i
ne
eri
n
g
.
2013; 1
1
(2): 5
83-5
90.
[10]
Xi
uku
i
Li; (R
ez
a) Z
e
kavat, Se
ye
d A. "C
on
ne
ctiv
it
y
a
n
a
l
ysis
of grou
p-b
a
se
d
cogn
itive r
adi
o
net
w
o
rks
"
.
201
2
IEEE I
n
ternational Sy
m
p
osium on
a World of
Wireless,
Mob
ile and M
u
ltim
edia Networ
k
s
(WoWMoM)
. 2012; 1(6).
[11]
O Olabi
yi, A
Annam
ala
i
.
Efficient P
e
rfor
mance Ev
alu
a
tio
n
of
Co
op
erati
v
e No
n-reg
e
n
e
r
ative R
e
lay
Netw
orks
. In: IEEE Co
nsume
r
Commu
nicati
ons
and
Net
w
orkin
g
C
onfere
n
ce (C
CNC),
Las V
ega
s
,
U.S.A., 2012; 797-8
01.
[12]
Gang
Xi
on
g,
K
i
shor
e
S, Ye
ne
r A. "Cost co
n
s
traine
d sp
ectrum
sens
in
g i
n
cogn
itive r
adi
o
net
w
o
rks
"
.
201
0 44th An
n
ual C
onfere
n
ce
on Informatio
n
Sciences a
nd
S
y
stems (CISS
)
. 2010; 1(6).
[13]
Ying-C
h
a
ng L
i
ang, Yo
ngh
on
g Z
eng, Peh E
C
Y,
Anh T
uan Hoan
g. "Sensi
ng-T
h
rough
put
T
r
adeoff for
Cog
n
itive R
adi
o Net
w
orks".
IEEE Transactions on Wi
reles
s
Communic
a
tions
. 200
8; 7(4)
: 1326, 13
37.
Evaluation Warning : The document was created with Spire.PDF for Python.