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.
5
,
Octo
be
r
2020
,
pp.
4818
~
4823
IS
S
N: 20
88
-
8708
,
DOI: 10
.11
591/
ijece
.
v10
i
5
.
pp
481
8
-
482
3
4818
Journ
al h
om
e
page
:
http:
//
ij
ece.i
aesc
or
e.c
om/i
nd
ex
.ph
p/IJ
ECE
Noise u
n
certainty
e
ff
ect o
n
m
ulti
-
channel
c
ogniti
ve
ra
di
o n
etworks
Amira
Os
am
a
1
, H
eb
a A.
Ta
g
El
-
Dien
2
, Ah
mad
A.
Az
iz
El
-
Bann
a
3
, A
dly S.
T
ag El
-
Dien
4
1
Depa
rt
m
ent
of
El
e
ct
ri
ca
l
and
C
om
pute
r
Engi
n
e
eri
ng,
High
Inst
i
tut
es
for
Eng
ineeri
ng
and Te
chn
olog
y
Al
-
obour,
Eg
y
p
t
2,3,4
Facul
t
y
of En
gine
er
ing
a
t
Sho
ubra
,
B
enha Uni
ver
sit
y
,
Eg
y
pt
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
Oct
28
, 201
9
Re
vised
Ma
r
11
,
2020
Accepte
d
Ma
r
30
, 202
0
Achie
ving
high
throughput
is
t
he
m
ost
import
ant
go
al
of
cog
nit
ive
rad
io
net
works
.
Th
e
m
ai
n
proc
ess
in
cognitive
rad
io
is
spec
trum
se
nsing
that
ta
rge
ts
getting
vac
an
t
ch
anne
ls
.
Th
ere
are
m
an
y
sensing
m
et
hods
li
k
e
m
at
che
d
filter
,
fea
ture
de
tect
ion,
interfe
r
e
nc
e
te
m
per
at
ur
e
and
ene
r
g
y
det
e
ct
ion
which
is
emplo
y
ed
i
n
the
proposed
sy
st
em;
howeve
r,
en
e
r
g
y
det
e
ct
ion
suffers
from
noise
unc
ert
a
inty
.
In
th
is
pape
r
a
stud
y
of
throughpu
t
under
noise
flu
ctuati
on
eff
e
ct
is
i
ntroduc
ed
.
Th
e
work
in
thi
s
pap
er
proposes
m
ult
i
-
cha
nn
el
s
ystem;
the
ov
eral
l
m
ult
i
-
ch
annel
t
hroughput
is
stu
die
d
unde
r
noise
fluctuatio
n
eff
e
ct
.
In
ad
dit
ion,
the
profi
ci
en
c
y
of
th
e
n
et
work
has
bee
n
ex
amined
under
diffe
r
ent
num
ber
of
cha
nnel
s
and
sensin
g
ti
m
e
with
noise
unc
ertain
t
y
.
Ke
yw
or
d
s
:
Cognit
ive
r
adi
o
M
ulti
-
channel
No
ise
uncertai
nty
Thro
ughput
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
:
Am
ira Os
am
a,
Dep
a
rt
m
ent o
f El
ect
rical
an
d
Com
pu
te
r
E
ng
i
neer
i
ng,
High
In
sti
tutes
for
En
gin
ee
rin
g
a
nd Tec
hnol
og
y
Al
-
obour
,
21 Cai
ro
-
Be
lbeis
Deser
t R
d,
Kali
o
beya
, E
gy
p
t
.
Em
a
il
:
a
m
iraosa
m
a11
1@
gm
ail.co
m
,
a
m
ira.o
sam
a@o
i.edu
.
e
g
1.
INTROD
U
CTION
Cognit
ive
ra
dio
(CR)
us
in
g
s
pectr
um
sensing
te
ch
nique
ca
n
be
us
e
d
to
s
olv
e
the
iss
ues
of
s
pectr
um
unde
ru
ti
li
zat
ion
[1
-
3]
.
I
n
c
ogniti
ve
ra
dio
f
ie
ld
the
sec
on
dar
y
us
e
r
(
SU)
detect
the
prim
ary
us
er
ba
nd
s
a
nd
dep
e
nds
on
th
is
decisi
on
it
se
le
ct
s
the
sp
ect
r
um
fo
r
its
com
m
un
ic
at
ion
[4
-
6]
.Th
e
CR
syst
e
m
cou
ld
be
s
ing
le
channel
or
Mu
lt
i
-
channel,
c
hoos
i
ng
the
sys
tem
accord
i
ng
to
nee
ds
,
bec
ause
eac
h
syst
e
m
has
ben
e
fits
a
nd
dr
a
w
back
s
.
I
n
CR
syst
e
m
,
t
he
sec
onda
ry
us
er
ha
ve
to
detect
the
vac
ant
cha
nnel
[
7
-
10]
.
Wh
e
n
a
vacant
channel
is
dete
ct
ed
,
the
sec
ondar
y
syst
em
will
acce
ss
that
channel.
H
ow
e
ver
,
s
pectr
um
sensing
a
s
a
n
urge
nt
issue
in
CR
,
dem
and
s
the
s
econda
ry
us
e
r
to
powe
rfull
y
and
su
cces
sf
ul
ly
sense
the
existe
nce
of
pri
m
ar
y
sign
al
[
11
-
13]
.
Sp
ect
r
um
manag
em
ent
is
the
cor
e
f
unct
ion
in
CR
;
con
ta
in
s
var
i
ous
processes
s
uch
a
s
sp
ect
r
um
sensing
(SS),
sp
ect
r
um
decisi
on
an
d
s
pectru
m
ha
ndoff.
SS
is
one
of
t
he
m
os
t
i
m
po
rtant
ad
va
ntage
s
of
CR
sens
ors
netw
ork
f
ro
m
old
wi
reless
se
ns
or
net
wor
ks
(
WSN
s
).
It
ov
ertakes
fe
w
op
ti
cal
s
l
ike
low
SN
R
f
or
pr
im
ary use
rs,
ti
m
e d
isper
sion, c
hannel f
adin
g,
a
nd
no
is
e uncertai
nty
[
14
-
16]
.
Ma
xim
u
m
thro
ug
hput
of
the
netw
ork
has
been
stu
died
f
or
seve
ral
num
ber
s
of
opti
m
al
CR
us
ers
.
Ah
ea
d
of
the
issue
of
thre
shold
m
is
m
at
ch
of
e
nergy
dete
ct
or
s
with
noise
power
un
ce
r
ta
inty
,
a
coo
pe
rati
ve
sp
ect
r
um
sensing
m
et
ho
d
wi
th
dy
nam
ic
dual
thres
hold
is
ex
pr
es
sed
in
[
17
].
In
[
18
]
the
op
ti
m
al
sensing
te
chn
iq
ue
has
bee
n
offer
e
d
m
axi
m
iz
ing
channel
th
r
oughput.
T
he
offer
e
d
op
ti
m
al
coope
rati
ve
s
pe
ct
ru
m
sensing
(CS
S)
s
et
ti
ng
s
f
or
wide
-
ba
nd
se
ns
in
g
cha
nnel
s
is
inv
e
sti
gated
a
nd
determ
ined
s
pecifica
ll
y
with
fe
w
si
m
ple
howev
e
r
de
pe
nd
a
ble
te
chn
i
qu
e
s
.
In
[
19
]
a
ne
w
s
pectr
um
sensing
a
da
ptive
al
gorit
hm
con
side
rin
g
no
ise
un
ce
rtai
nty
ha
s
been
pro
pos
ed.
I
n
[
20
]
th
e
researc
hers
stu
die
d
noise
un
ce
rtai
nty
effe
ct
and
fa
ding
on
the
detect
ion
pe
rfor
m
ance.
In
[
21
]
resea
rch
e
rs
offer
e
d
an
e
vid
e
nce
-
t
heory
based
f
us
io
n
r
ule
for
co
op
e
r
at
ive
energy
detect
ion
in
exista
nce
of noise
powe
r
uncertai
nty.
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
Noise u
ncert
ain
ty
eff
ect
on m
ulti
-
ch
annel c
ogniti
ve r
ad
i
o n
et
works
(
Amir
a Osam
a
)
4819
Au
t
hors
i
n
[
22
]
introdu
ce
a
m
od
ifie
d
tw
o
-
sta
ge
detect
ion
te
chn
i
que
that
relay
on
e
ne
r
gy
detect
io
n
unde
r
noise
uncertai
nty.
Re
searche
rs
i
n
[
23
]
e
xam
ined
the
act
of
s
pe
ct
ru
m
sensing,
t
hey
pro
pos
ed
that
the
thr
oughput
reaches
m
axim
u
m
at
op
tim
al
sensing
tim
e.
To
certi
fy
how
processes
of
pr
im
ary
us
er
is
no
t
disturbe
d,
it
has
to
keep
t
he
sp
e
ct
ru
m
-
sensing
ca
pab
il
it
y
to
get
vacan
t
cha
nn
el
wh
e
n
S
U
nee
ds
it
.
Ther
e
f
or
e
,
the
prof
ic
ie
ncy
of
sensing
the
exi
ste
nce
of
pr
im
ary
sign
al
s
re
m
ai
ns
i
m
po
rta
nt.
S
pectr
um
s
ensin
g
con
ta
in
s
seve
r
al
m
e
tho
ds
(e.
g.
Ma
tc
he
d
filt
er
,
ene
rg
y
dete
ct
ion
,
feat
ur
e
detect
ion,
inter
fer
e
nce
tem
perat
ur
e)
.
In
[
24
]
a
uthor
s
stu
died
a
nd
analy
zed
the
prof
ic
ie
ncy
of
energy
detect
or
sp
ect
r
um
sensing
m
et
ho
d
with
par
am
et
ers
affe
ct
ing
it
s
pe
rfor
m
ance
on
N
akag
am
i
-
m
fadi
ng
c
ha
nn
el
s
unde
r
no
ise
unc
ertai
nt
y
and
w
it
ho
ut.
Au
t
hors
in
[
25]
pr
op
os
e
d
a
di
ff
ere
nt
syst
em
of
t
wo
sta
ges
sp
ect
r
um
sensing
,
A
dap
ti
ve
two
-
sta
ge
s
pec
trum
sensing
(
ATSS
)
,
with
noise
un
ce
rtai
nty
effe
ct
ATS
S
is
a
m
od
ific
at
ion
of
a
pr
e
dicta
ble
two
sta
ge
sp
e
ct
ru
m
sensing
wh
e
n
the
decisi
on
th
res
hold
of
eac
h
sta
ge
is
a
da
pted
on
t
he
di
sta
nce,
e
xp
ect
ed
noise
var
ia
nce
an
d
con
cl
ud
e
d
nois
e
uncertai
nty
r
ang
e
.
No
ise
un
certai
nty
aff
ect
the
pe
rfo
rm
ance
of
m
ulti
-
chan
nel
is
st
ud
ie
d
with
diff
e
re
nt num
ber
of
c
ha
nn
el
a
nd d
eci
des w
hich best
nu
m
ber ti
g
et
h
i
gh th
r
oughput.
2.
RESEA
R
CH MET
HO
D
This
pa
pe
r
wil
l
introdu
ce
m
ulti
-
channel
syst
e
m
wh
ic
h
has
severa
l
cha
nnel
s
that
stud
ie
s
how
noise
un
ce
rtai
nity
af
fects
the
syst
em
per
f
or
m
ance
to
get
the
be
st
scenari
o.
O
ur
pro
po
se
d
sy
stem
her
e
co
nsi
sts
of
m
ul
i
-
channel
(
as
each
S
U
st
ands
for
one
c
hannel
in
pro
pose
d
syst
e
m
)
S
pectr
um
sensing
pe
rfo
rm
ance
has
been
offe
red
to
get
best
num
ber
of
CR
us
er
s
[
23
]
,
sen
sin
g
ti
m
e,
Throug
hput
R
unde
r
noi
se
fluctuati
on
eff
ect
.
The recei
ve
d
si
gn
al
f
or
e
ve
ry
CR
is sam
pled
at
sam
pling
f
re
qu
e
ncy
fs.
As
sho
wn
in
F
ig
ure
1,
eve
r
y
cog
niti
ve
f
r
a
m
e
con
sist
s
of
s
pectr
um
s
ensin
g
tim
e
(t
)
an
d
data
transm
issi
on
ti
m
e
(T
-
t)
,
wh
e
r
e
T
is
t
he
t
otal
fram
e
tim
e.
Con
si
der
that
t
he
distri
bu
ti
on
f
un
ct
io
n
f
or
noi
se
can
be
s
umm
arized
in
a
n
i
nter
val
[
(
1
+
)
−
1
2
,
(
1
+
)
2
]
,
w
he
re
2
noise
var
ia
nce
for
ℎ
c
hannel
a
nd
is
a
par
am
e
te
r
that
qu
antiz
ie
s
the
le
vel
of
the
un
ce
rtai
nity
.A
ss
um
ing
K
is
the
nu
m
ber
of
sam
ples
existi
ng
durin
g
t
.T
herefo
re
the
nu
m
ber
of
sam
ples
,
K=t.
fs
[
2
3
,
2
5
].
The
pro
ba
bi
li
t
y
of
detect
i
on
a
nd
pr
ob
a
bi
li
t
y
of
false al
arm
f
or
m
ul
ti
-
channel
syst
e
m
u
nd
e
r n
oise
un
ce
rtai
nt
y effect ca
n be
wr
it
te
n
as
=
a
rgmax
2
∈
[
(
1
+
)
−
1
2
,
(
1
+
)
2
]
1
2
(
1
√
2
(
2
−
1
)
√
.
)
(1)
=
a
rgm
a
x
2
∈
[
(
1
+
)
−
1
2
,
(
1
+
)
2
]
1
2
(
1
√
2
(
2
−
Ɣ
−
1
)
√
.
2
Ɣ
+
1
)
(2)
So
lvi
ng (1
)
, (2
)
under n
oise
we wil
l get
=
1
2
(
1
√
2
(
(
1
+
)
2
−
1
)
√
)
(3)
=
1
2
(
1
√
2
(
(
1
+
)
−
1
2
−
−
1
)
√
2
+
1
)
(4)
=
1
2
(
1
√
2
(
2
−
1
)
√
)
(5)
=
1
2
(
1
√
2
(
2
−
Ɣ
−
1
)
√
2
Ɣ
+
1
)
(6)
w
he
re
=
1
+
,
Ɣ
re
prese
nts
th
e
S
NR
a
t
the
CR
recei
ver
f
or
ℎ
cha
nn
el
an
d
re
pres
ents
t
he
decisi
on
thres
ho
l
d.
The
thr
oughput f
or
the total
f
al
se
a
la
rm
p
ro
ba
bili
ty
in
the a
bs
e
nc
e of P
U
is
[
20
,
23]
.
0
=
−
(
1
−
)
(
)
(7)
w
he
re
represe
nts
the
thr
oughput
in
the
a
bse
nce
of
P
U.
The
t
hroug
hput
for
the
total
m
issed
detect
io
n
pro
bab
il
it
y i
s
1
=
−
1
(
1
−
)
(
1
)
(8)
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
.
5
,
Oct
ob
e
r 2
020
:
48
18
-
48
23
4820
w
he
re
1
represe
nt
s the thro
ughp
ut in
t
he
e
xistence
of P
U.
Fro
m
[
20
,
23]
w
e
get
=
(
1
−
)
(9)
=
(
1
−
)
(10)
The
t
otal t
hro
ughp
ut R
of
t
he C
R netw
ork
ca
n be e
xpresse
d from
p
rev
i
ou
s
disscussi
on as
=
−
(
0
(
1
−
)
(
0
)
+
1
(
1
−
)
(
1
)
)
(11)
As
R
is
a
f
un
ct
ion
of
an
d
,it
will
be
aff
ect
e
d
by
c
hangin
g
no
ise
fluctuati
on
s
.
S
o
the
t
hroug
hput
will
be
decr
ea
sed
w
hen
pro
ba
bili
ty
of
false
al
ar
m
increased
or
pro
bab
il
it
y
of
detect
ion
decr
ease
d.
St
ud
yi
ng
n
that
represe
nt
s
the
nu
m
ber
of
CR
s
in
sp
e
c
trum
sensing
.
To
com
par
e
th
is
pr
o
pose
d
syst
e
m
and
[17]
unde
r
no
ise
f
luct
uation eq
uatio
n of
the syst
em
w
il
l
b
e
=
1
2
(
−
1
(
2
)
√
2
Ɣ
+
1
+
2
Ɣ
)
(12)
Fig
ure
1
.
Co
gnit
ive f
ram
e struc
ture
3.
RESU
LT
S
A
ND AN
ALYSIS
Now
we
a
re
goin
g
to
i
nvest
igate
the
noise
fluctuati
on
on
the
ED
ROC.
As
ass
um
ed
ab
ov
e
the
noise
var
ia
nce
with
un
ce
rtai
nty
ch
ang
e
d
i
n
the
i
nt
erv
al
2
ϵ[
2
/
,
2
]
w
here
β
>1
a
nd
c
ha
ng
e
d
from
1.
259
to
1.5
85
[
22]
.
BPSK
m
odulati
on
has
bee
n
use
d
by
pri
m
ary
us
e
r
to
tran
sm
it
it
s
data
with
3
M
Hz
band
wid
th
.
The
m
axi
m
u
m
t
i
m
e
fo
r
wh
ic
h
the
seconda
ry
us
er
unin
form
ed
of
the
pr
im
ary
act
ion
is
selected
a
s
Fs.T=
3000[
19]
.
T
he
fr
am
e
tim
e
of
detect
ion
cy
cl
e
is
100m
s
and
ta
rg
et
detect
io
n
pro
ba
bili
ty
i
s
0.7
.
We
c
hoos
e
(
0
)
=0.8,
(
0
)
=0.2,
=6.6582
a
nd
=6.6
137.
Fig
ure
2
sh
ows
t
he
rel
at
ion
betwe
e
n
thr
oughput
a
nd
S
NR
at
dif
f
eren
t
value
s
f
or
noise
fluctu
at
ion
s.
W
e
ob
serv
e
that
w
he
n
no
ise
inc
rea
sing
by
30%
an
d
SN
R
=
-
12
dB,
the
th
rou
ghput
R
te
nds
to
zer
o,
w
hich
m
eans
syst
em
a
t
β
=1.3
sti
ll
work
i
ng
unti
l
SNR
reaches
-
12dB.
Figure
3
ex
pla
ins
the
th
r
oughput
relat
io
n
with
num
ber
of
CR
s
op
e
rat
ing
at
t=
2
m
s.
stud
yi
ng
at
aver
a
ge
nu
m
ber
of
CR
s
=5.At
β=1,
the
t
hroughp
ut
R=
3.5.A
t
β=1.
05,
t
he
thr
ough
pu
t
R=
2.5,
w
hich
m
ea
ns
by
increasin
g
β
by
5%
throu
ghpu
t
dec
reasin
g
by
29
%.
At
β=1.1,
the
th
r
oughput
R=
1.8,
w
hich
m
ea
ns
by
increasin
g
by
10
%
th
rou
ghpu
t
dec
reasin
g
by
50
%.
At
β=1.3,
the
thr
ough
pu
t
R=
0.5,
w
hich
m
eans
by
increasin
g by
30% t
hroug
hput
d
ec
reasin
g by
86%.
Figure
4
s
how
s
the
relat
ion
betwee
n
thr
ou
ghput
and
se
nsi
ng
ti
m
e
with
diff
e
ren
t
num
ber
of
CR
s.
(a)
at
n=15,
we
obse
rv
e
t
ha
t
wh
e
n
no
ise
increasi
ng
by
30%,
the
t
hroug
hput
te
nds
to
zer
o.
(b)
a
t
n=10,
we
obser
ve
w
he
n
increasi
ng
no
ise
by
30%
,
the
thr
oughput
decr
easi
ng
by
70
%.
(c)
at
n=
5,
we
ob
se
r
ve
wh
e
n
increasin
g
no
is
e
by
30%,
the
t
hro
ughput
dec
reasin
g
by
60
%.From
that
we
can
say
at
n=10
,t
his
is
the
bes
t
nu
m
ber
for
CR
s in
t
his syste
m
und
e
r n
oise ef
fect.
Figure
5
s
how
s
a
com
par
iso
n
with
[
17]
under
no
ise
fl
uct
uation
β=
1.3
we
obta
ine
d
th
at
at
t=
3
m
s,
P_
f
in
re
fe
ren
c
e
syst
e
m
reache
s
to
0.5
5
wh
i
ch
m
eans
it
aff
ect
ed
from
no
ise
with
30%
bu
t
in
our
pro
pos
e
d
syst
e
m
that
re
aches
0.43
w
hi
ch
m
eans
it
a
ff
ect
ed
from
no
ise
with
14%
.
This
m
eans
that
our
pro
pos
ed
syst
e
m
is b
et
te
r
tha
n refe
renc
e syst
e
m
.
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
Noise u
ncert
ain
ty
eff
ect
on m
ulti
-
ch
annel c
ogniti
ve r
ad
i
o n
et
works
(
Amir
a Osam
a
)
4821
Figure
2. Th
r
ough
pu
t
vs
SN
R
Fi
gure
3. Th
r
ough
pu
t
vs
num
ber
of CR
s
(
a)
(
b)
(
c)
Fig
ure
4. The
re
la
ti
on
bet
wee
n
th
r
oughput R
and se
ns
in
g
ti
m
e t
,
(a)
N
o.
of CR
s
=15
,
(b) No
. of
CR
s=1
0
,
(c) N
o.
of CR
s=5
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
.
5
,
Oct
ob
e
r 2
020
:
48
18
-
48
23
4822
Fig
ure
5. Se
ns
ing t
i
m
e v
s
P
f
4.
CONCL
US
I
O
N
The
no
ise
unc
ertai
nty
eff
ect
on
the
perf
orm
ance
of
m
ult
i
-
cha
nn
el
is
st
ud
ie
d
a
nd
we
ob
s
er
ve
that
,
in
m
ulti
-
chann
el
,
fo
r
ou
r
syst
e
m
it
is
go
od
to
wo
r
k
with
10
c
hannels
to
achieve
reas
onable
thr
ough
pu
t
in
relat
ion
with
S
NR
or
with
se
ns
in
g
ti
m
e
by
increasin
g
nois
e
up
t
o
30%.F
ro
m
that
we
c
an
say
,the
pro
po
s
ed
syst
e
m
can r
ea
ch reaso
na
ble t
hro
ughput i
n
si
ng
le
or m
ulti
-
channel
unde
r n
oise that i
ncr
ea
se of
30% ef
fe
ct
.
ACKN
OWLE
DGE
MENTS
I
w
ould
li
ke
to
e
xpress
m
y
strong
tha
nk
s
to
m
y
su
pervisor
s
f
or
their
gu
i
dan
c
e.
I
w
ou
l
d
li
ke
t
o
express
m
y Lov
e a
nd tha
nk to
m
y fam
il
y
m
e
m
ber
s.
REFERE
NCE
S
[1]
Y
.
Gao,
et
al
.,
“
Eff
ec
ti
v
e
Capaci
t
y
of
Cognit
iv
e
Radi
o
S
y
stem
s
,”
2016
IEE
E
13
th
Inte
rnation
al
Confe
renc
e
o
n
Signal
Prec
essin
g
(
ICSP
)
,
pp.
17
57
-
1761
,
2016
.
[2]
D
.
W
.
Yue,
e
t
a
l
.,
“
Log
-
ave
r
age
-
SN
R
rat
io
and
coope
ra
ti
ve
sp
ectrum
sensing
,”
I
EE
E
Journal
o
f
Comm
unic
ati
ons
a
nd
Net
works
,
v
ol.
18
,
no
.
3
,
pp
.
311
-
319
,
2016
.
[3]
A
.
Os
ama
,
et
al
.,
“
Spect
rum
Sens
ing
in
Single
Ch
anne
l
and
Mult
i
-
C
hanne
l
Cogni
tive
Radi
o
Netwo
rks
,”
Indone
sia
n
Journal
of
Elec
t
rical
Engi
ne
erin
g
and
Computer
Sci
en
ce
(
IJE
ECS
)
,
vol.
16
,
no
.
2
,
pp.
812
-
817
,
20
1
9
.
[4]
D
hiv
y
a
,
et
al
,
“
Inge
nious
Method
for
Conduci
ve
Handoff
Applia
nce
in
Cognit
iv
e
Radi
o
Networks
,”
Inte
rnationa
l
Journal
of
Elec
t
rical
and
Computer
Eng
ine
ering
(
IJE
CE
)
,
v
ol.
8,
n
o.
6
,
pp
.
5195
-
5202
,
201
8
.
[5]
S
.
H.
Alnab
el
si
,
et
a
l
.,
“
D
y
n
amic
resourc
e
al
lo
c
at
ion
fo
r
opport
unisti
csoftwa
r
e
-
def
ine
d
IoT
n
etw
orks:
s
toc
hastic
opti
m
iz
ationfra
m
ework
,”
Inte
rn
ati
onal
Journal
of
El
e
ct
rica
l
an
d
Computer
Engi
nee
ring
(
IJE
C
E
)
,
v
ol.
10,
n
o.
4
,
pp.
3854
-
3861
,
20
20
.
[6]
R.
Abdelr
assoul,
et
al
,
“
Com
par
a
ti
ve
stud
y
of
spe
ct
rum
sensing
for
cogni
ti
v
e
rad
io
sy
st
em
using
ene
rg
y
d
e
t
ec
t
ion
over
diff
ere
n
t ch
anne
ls,
”
World
S
ymposium on
Co
mputer
Applicati
ons
and
Re
s
earc
h
,
IE
EE
,
C
ai
ro.
2016.
[7]
M
.
Alja
rah
,
e
t
al
.,
“
Coopera
t
ive
hie
rar
chi
c
al
base
d
edge
-
computin
g
appr
oac
h
for
resourc
es
al
lo
ca
t
i
on
of
distri
bu
te
d
m
obil
e
and
Io
T
appl
i
ca
t
ions
,”
I
nte
rnational
Jou
rnal
of
Elec
trical
and
Compute
r
Engi
nee
ring
(
IJE
CE
)
,
v
ol
.
10,
n
o.
1
,
pp
.
296
-
3
07
,
2020
.
[8]
H
.
Al
-
Mahdi
an
d
Y
.
Fouad
,
“
D
esign
and
anal
ysis
of
routi
ng
protoc
ol
for
cog
nit
ive
r
adi
o
ad
hoc
net
works
in
het
ero
g
ene
ous
e
nvironment
,
”
Int
ernati
onal
Journal
of
El
e
ct
ri
cal
a
nd
Computer
En
gine
ering
(
IJE
C
E
)
,
v
ol.
9
,
n
o.
1
,
pp.
341
-
351
,
20
1
9
.
[9]
Q.
Zou,
et
al
.
,
“
Coopera
t
ive
s
en
sing
via
seque
nt
ia
l
detec
ti
on
,”
IEE
E
Tr
ansacti
o
ns
on
Signal
Pr
oce
ss
ing
,
vol.
58
,
no.
12
,
pp
.
6266
-
6283
,
2010
.
[10]
M
.
Subhedar
an
d
G
.
Bira
jda
r
,
“
Spect
rum
sensing
Te
chni
q
u
es
in
Cognit
iv
e
Radi
o
Netwo
rks
:
A
survey
,”
Inte
rnational
Jo
urnal
of
N
ex
t
-
G
e
nerati
on
N
et
wor
ks
(
IJNGN)
,
vol.
3
,
no.
2
,
pp.
527
7
-
5288
,
201
0
.
[11]
S
.
Chowdhur
y
,
et
al
.,
“
A
Throughput
-
eff
i
ci
en
t
Coopera
ti
v
e
Se
nsing
and
Allocati
on
Mode
l
for
Cognit
ive
Rad
i
o
Networks
,”
201
5
IEE
E
Int
ernational
Confe
ren
ce
on
Adv
an
ce
d
N
et
works
and
Tel
ec
omm
unic
at
ion
s
Syste
ms
(
ANTS
)
,
pp.
1
-
3
,
2015
.
[12]
A
.
Bhowm
ic
k,
et
al
.,
“
A
Hy
b
ri
d
Coopera
ti
v
e
Spect
rum
Sensing
for
Cognit
ive
Radi
o
Network
s
in
Presenc
e
o
f
Fading
,”
2015
T
went
y
Fi
rs
t
Na
tional Conferenc
e
on
Comm
unic
at
ions
(
NCC
),
Mum
bai
,
pp
.
1
-
6
,
2
015
.
[13]
Y.
Chu
et
al
.
,
“
Hard
Dec
ision
F
usion
Based
Coopera
t
ive
Spec
tr
um
Sensing
over
Na
kaga
m
i
-
m
Fading
Chann
el
s
,”
IEE
E
,
vo
l. 12, n
o.
31
,
pp
.
1
-
4.
2
012.
[14]
H
.
Li
et
al
.
,
“
Util
ity
-
B
ase
d
Coopera
t
ive
Spec
trum
Sensing
S
che
dul
ing
in
Cognit
ive
Rad
io
Networks
,”
IEEE
,
pp.
1
-
12
,
2015
.
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
Noise u
ncert
ain
ty
eff
ect
on m
ulti
-
ch
annel c
ogniti
ve r
ad
i
o n
et
works
(
Amir
a Osam
a
)
4823
[15]
S
.
Zha
ng
,
et
a
l
.
,
“
Cross
-
lay
er
Ret
h
ink
on
Sensing
-
throughp
ut
Tra
deof
f
for
Multi
-
c
hannel
Cognit
ive
Rad
io
Networks
,
”
I
EEE
Tr
ansacti
ons
o
n
Wireless
Comm
unic
ati
ons
,
v
ol.
15
,
no
.
10
,
pp
.
6883
-
68
9
7
,
20
16.
[16]
P
.
Vara
d
e,
et al
.,
“
Throughput
m
axi
m
iz
a
ti
on
of
c
ognit
ive
rad
io
m
ult
i
r
ela
y
n
et
wor
k
with inte
rf
ere
n
ce
m
ana
g
ement
,”
Inte
rnational
Jo
urnal
of El
e
ct
ri
c
al
and
Comput
er
Engi
n
ee
ring
(
IJE
CE
)
,
v
ol
.
8
,
n
o.
4,
pp.
2230
-
223
8,
2018
.
[17]
R
.
W
an,
e
t
al
.
,
“
Dy
n
amic
dual
thre
shold
coop
era
t
ive
spe
ct
ru
m
sensing
for
cogni
ti
ve
rad
i
o
u
nder
noise
pow
er
unce
rt
ai
nt
y
,”
Hum
an
-
ce
ntic
Com
puti
ng
and
Infor
mation
Sc
ie
nc
e
,
vol.
9
,
no
.
1
,
pp
.
1
-
21
,
2019
.
[18]
J
.
Shen,
et
a
l
.,
“
Maximum
Chan
nel
Throughpu
t
via
Cooper
at
iv
e
Spect
rum
Sens
ing
in
Cognit
iv
e
R
adi
o
Networks
,”
IEE
E
Tr
ansacti
o
ns
o
n
Wireless
C
omm
unic
ati
ons
,
vol.
8
,
no
.
10
,
pp
.
5166
-
5175
,
20
09
.
[19]
D
.
Raman
and
N.
P.
Singh
,
“
An
Algorit
hm
for
Spect
rum
Sensi
ng
in
Cogni
ti
v
e
Radi
o
und
er
No
ise
Unce
r
ta
in
t
y
,”
Inte
rnational
Jo
urnal
of Fut
ure
Gene
ration
Com
municat
ion
and
Net
working
,
vol
.
7
,
no
.
3
,
pp.
61
-
68
,
2014
.
[20]
Z
.
Quan,
et
al
.,
“
Optimal
Multi
band
Joint
Dete
c
t
ion
for
Spect
rum
Sens
ing
in
C
ognit
ive
Rad
io
Networks
,”
IEEE
Tr
ansacti
ons
o
n
Signal
Proce
ss
in
g
,
vol
.
57
,
no
.
3
,
pp.
1128
-
1140
,
2009
.
[21]
P
.
B.
Gohain,
et
al
.,
“
Evi
de
nce
The
or
y
ba
sed
Coopera
ti
v
e
Ene
rg
y
D
et
e
ct
ion
under
Noise
Unce
rtai
n
t
y
,”
GL
OBECOM
2017
-
2017
IEEE
G
lobal
Comm
unications
Conf
ere
n
ce
,
pp.
1
-
7
,
2017
.
[22]
H
.
A.
T
.
El
-
Di
en,
et
al
.
,
“
Noi
se
Unce
rta
in
t
y
Eff
ect
on
a
Modifie
d
Two
-
Stag
e
Spect
rum
Sensing
Te
chni
qu
e
,”
Indone
sian J
our
nal
of
Elec
tric
al
Engi
ne
ering
and
Computer
Sc
ie
n
ce
(
IJEECS
)
,
vol
.
1
,
no
.
2
,
pp
.
34
1
-
348,
2016
.
[23]
A
.
Bhowm
ic
k,
e
t
al
.
,
“
Throughp
ut
Optimization
with
Cooper
at
iv
e
Spectrum
Sen
sing
in
Cogn
it
iv
e
Radi
o
Network
,”
IEE
E
Inte
rnat
io
nal
Ad
vanc
e
Co
mputing
Conf
ere
nce
(
IACC)
,
Gur
gao
n
,
pp
.
329
-
33
2
,
2014
.
[24]
A
.
E
slami
and
S
.
K
ara
m
za
d
eh,
“
Perform
anc
e
Anal
y
sis
of
En
e
rg
y
B
ase
d
Spe
c
trum
Sensing
over
Naka
g
ami
-
m
Fading
Channel
s
with
Noise
u
nce
rt
ai
nt
y
,”
Inter
nati
onal
Journ
al
o
f
El
e
ct
ronic
s,
Me
chan
ic
al
a
nd
Me
chat
roni
c
s
Engi
ne
ering
,
vol
.
6
,
no
.
1
,
pp
.
11
01
-
1106
,
2016
.
[25]
W
.
Lee,
e
t
al
.,
“
Adapti
ve
Two
-
stage
Spect
rum
Sensing
under
Noise
Unce
rta
i
n
t
y
in
Cogni
ti
ve
Radi
o
Networks
,
”
ECTI Trans
act
ions
o
n
Elec
tric
al
Eng
ineering
,
Ele
ct
ronics,
a
nd
Co
mm
unic
ati
ons
,
v
ol.
14
,
no
.
1
,
pp
.
21
-
35,
2016
.
BIOGR
AP
H
I
ES
OF
A
UTH
ORS
Am
ira
Os
ama,
Te
a
chi
ng
As
sistant
in
Elec
tr
ical
and
Comm
unic
at
ion
Eng
ine
er
in
g,
Univer
sita
d
High
Instit
ute
s
for
Engi
nee
r
ing
and
Te
chnol
og
y
Al
-
obour.
She
rec
ei
v
ed
her
B.
En
g.
,
from
Benha
unive
rsit
y
(
Eg
y
p
t)
in
2013.
She
has
bee
n
a
Teac
hing
As
sis
ta
nt
High
Instit
ute
s
f
or
Engi
ne
eri
ng
and
Technol
og
y
Al
-
obour
since
2014.
Her
res
ea
rch
int
er
ests
inc
lud
e
the
f
ie
l
d
of
net
work
comm
unca
ti
on,
Mobile
s
y
s
te
m
s
,
W
SN
and
IOT
.
Heba
A.T
ag
E
l
Dien
rec
ie
v
ed
B.
Sc.
,
M.Sc.
and
Ph.D.
degr
ee
s
from
Shou
bra
Facul
t
y
of
Engi
ne
eri
ng,
B
e
nha
Univer
sit
y
,
Eg
y
p
t,
in
2007
,
2013,
and
2017
respe
ct
iv
ely
.
S
he
is
cur
ren
t
l
y
assistant
Profess
or
in
the
e
lectr
i
c
al
engi
n
ee
r
ing
d
epa
rtment
in
Shoubra
Facult
y
o
f
Engi
ne
eri
ng.
She
is a legal i
ns
truc
tor
in
Cisco ac
ad
em
y
in
Sho
ubra
fa
cult
y
of
E
ngine
er
ing.
Ahm
ad
A.
Aziz
El
-
Bann
a
re
ce
i
ved
a
m
aste
r’s
degr
ee
f
rom
Benha
Univer
sit
y
,
Eg
y
p
t,
in
2011,
and
a
Ph.D.
degr
ee
from
the
Eg
ypt
-
Japa
n
Univer
s
ity
of
Scie
n
ce
an
d
Te
chnol
og
y
,
E
g
y
pt
,
in
2014
.
He
serve
d
as
a
Visiti
ng
Resea
r
c
her
with
Os
aka
Un
ive
rsit
y
,
Jap
a
n,
from
Septe
m
ber
2013
to
June
2014.
He
a
lso
complet
ed
a
nine
-
m
onth
sch
ola
rship
in
embedde
d
s
y
s
te
m
s
at
Inform
ation
Te
chno
log
y
Inst
it
ute,
Eg
y
p
t,
in
2008.
Since
Ju
ne
2018,
he
h
a
s
bee
n
Pos
tdoc
t
ora
l
Rese
arc
h
Fell
ow
with
the
Sm
art
Sensing
and
Mobi
le
Com
puti
ng
La
bor
at
or
y
,
Shen
zhe
n
Uni
ver
sit
y
,
Chin
a.
He
al
so
holds
th
e
positi
on
of
an
As
sistant
Profess
or
with
the
Ele
ct
ri
ca
l
Engi
ne
ering
Depa
rtment,
Facul
t
y
of
Enginee
ring
at
Shoubra,
B
enha
Uni
ver
sit
y
,
Eg
y
pt
.
His
rese
arc
h
intere
sts
in
cl
ud
e
wire
le
ss
commun
ic
a
ti
ons,
embedde
d
s
y
st
ems
,
coope
rative
n
et
working,
MI
MO
,
spac
e
-
ti
m
e
codi
ng,
IoT
,
W
SN
,
under
wa
te
r
c
om
m
unic
at
ion, and m
ac
hine l
ea
r
ning.
Adl
y
S.
Ta
g
E
ldi
en
re
ce
ivrd
B.
Sc.
,
M
.
Sc.
and
Ph.D.
degr
ee
s
Benha
Univer
sit
y
,
Eg
y
pt
,
in
1984,
1989
and
1993
respe
ctively
.
H
e
is
cur
ren
tly
an
associate
profe
ss
or
in
the
depa
rtment
of
el
e
ct
ri
ca
l
engi
n
e
eri
ng
-
Benh
a
Univer
sit
y
.
H
e
wa
s
the
X
-
hea
d
of
Benha
Univer
si
t
y
n
et
work
and
informati
on
ce
n
t
er,
his r
ese
ar
ch in
te
rests
inc
lud
e
r
oboti
cs,
n
et
work
and
m
obil
e
com
m
unic
at
ion
.
Evaluation Warning : The document was created with Spire.PDF for Python.