Internati
o
nal
Journal of Ele
c
trical
and Computer
Engineering
(IJE
CE)
V
o
l.
6, N
o
. 4
,
A
ugu
st
2016
, pp
. 17
02
~
1
709
I
S
SN
: 208
8-8
7
0
8
,
D
O
I
:
10.115
91
/ij
ece.v6
i
4.7
560
1
702
Jo
urn
a
l
h
o
me
pa
ge
: h
ttp
://iaesjo
u
r
na
l.com/
o
n
lin
e/ind
e
x.ph
p
/
IJECE
Multiband OFDM for Cogn
itive Radio – A Way for
Cyclostationary Detection and
Interference Cancellation
J. Avila, K.
T
h
enmoz
h
i
Department o
f
Electronics
and C
o
mm
unication Engineer
ing, SASTRA Univer
sit
y
,
Than
javur
, T
a
m
il Nadu
, Ind
i
a
Article Info
A
B
STRAC
T
Article histo
r
y:
Received Feb 17, 2015
Rev
i
sed
May 17
, 20
16
Accepted
May 30, 2016
With the tremendous growth in
wireless
techno
log
y
ther
e has been a shortage
in the spec
trum
utili
zed for c
e
rt
a
i
n appli
cat
ions while som
e
spectrum
rem
a
ins
idle
. To over
c
o
m
e this proble
m
and for the effic
i
ent u
til
iz
ation of th
e
spectrum
cognit
i
v
e radio
is the suitabl
e solut
i
on. Multiband OF
DM can be
easil
y m
odeled a
s
cognitive rad
i
o
,
a techno
log
y
th
at is em
plo
y
ed for utiliz
ing
the ava
i
l
a
ble s
p
ectrum
in the m
o
s
t
effici
ent wa
y. S
i
n
ce s
e
ns
ing
of the free
s
p
ectrum
for detect
ing the arriv
a
l of the prim
ar
y
us
ers
is
the forem
o
s
t
job of
cognitiv
e, h
e
re
c
y
clost
a
tion
a
r
y
based spectrum
sensing is carr
i
ed out
. Its
perform
ance is
i
nves
tigat
ed
using universal software
defined rad
i
o peripher
a
l
(USRP) kit which is the
hardw
a
re t
e
st bed fo
r
the cogn
itiv
e ra
dio s
y
stem
.
Results are shown using Labview software. Further
to m
itigate the
interf
eren
ce b
e
t
w
een th
e prim
a
r
y
and cogn
itiv
e users a m
odif
i
ed in
trusion
elimination (AI
C
) algorithm had been
proposed which in turn
ensures the
coexis
t
e
nce
of b
o
th th
e us
ers
in t
h
e s
a
m
e
wir
e
les
s
environm
ent
.
Keyword:
Co
gn
itiv
e rad
i
o
Im
prove
d A
I
C
Mu
ltib
and
OFDM
Spectrum
sensing
USRP
Copyright ©
201
6 Institut
e
o
f
Ad
vanced
Engin
eer
ing and S
c
i
e
nce.
All rights re
se
rve
d
.
Co
rresp
ond
i
ng
Autho
r
:
Av
ila J,
Depa
rt
m
e
nt
of
El
ect
roni
cs
an
d C
o
m
m
uni
cati
on
En
gi
nee
r
i
n
g,
SASTR
A
Un
iversity,
Th
an
j
a
vu
r
,
Tamil N
a
d
u
-
6
134
01
, Ind
i
a.
Em
a
il: avila@ece.sastra.e
du
1.
INTRODUCTION
A p
r
im
ary p
r
oto
c
o
l
of UW
B
co
mm
u
n
i
cati
o
n
is th
e Mu
ltib
and
OFDM (MB-OFDM
) syste
m
th
at
sup
p
o
rt
s
hi
g
h
dat
a
rat
e
s w
h
i
l
e
cons
um
i
ng very
l
e
ss p
o
w
e
r [1]
.
It
i
s
prefe
rre
d beca
u
s
e i
t
addresse
s t
h
e
m
u
lt
i
p
at
h fadi
ng
p
r
o
b
l
e
m
t
h
ereby
m
a
ki
ng
i
t
a go
od c
o
m
m
uni
cat
i
on sy
st
em
i
n
an ur
ba
n/
i
n
d
o
o
r a
r
ea.
Hen
c
e
fu
rt
he
r di
sc
uss
i
on
fo
r i
m
pro
v
em
ent
of i
t
s
per
f
o
r
m
a
nce i
n
a
hi
g
h
c
o
n
g
e
st
i
on,
hi
gh
n
o
i
s
e en
vi
r
o
nm
ent
i
s
im
perat
i
v
e. T
h
e best
way
t
o
i
m
prove t
h
e
per
f
o
r
m
a
nce wi
t
h
out
cha
n
gi
n
g
t
h
e t
r
a
n
s
rece
pt
i
o
n
p
r
ot
ocol
i
s
by
t
h
e
use
of E
r
r
o
r
cont
rol
c
ode
s
[2]
.
Whi
l
e
hi
g
h
s
p
eed
dat
a
rat
e
s are
of t
h
e n
o
r
m
of t
h
e day
,
t
h
i
s
de
m
a
nd
inadverte
n
tly leads to t
h
e hi
gher e
r
ro
r
rat
e
s.
To t
ackl
e
i
t
co
ncat
enat
i
o
n
of
err
o
r c
o
nt
rol
c
ode
s i
s
pe
rf
or
m
e
d [3
]
fo
r t
h
ey
ha
ve a
hi
g
h
e
r
c
odi
ng
gai
n
t
h
an
a si
n
g
l
e
co
de.
It b
eco
m
e
s n
ecessary th
at th
e
m
u
ltib
an
d OFDM system [4
]
m
u
st c
o
ex
ist with
variou
s well
estab
lish
e
d pr
oto
c
o
l
s su
ch as
r
a
d
i
o
A
s
t
r
onomy o
f
Jap
a
n
,
W
i
Max
d
e
v
i
ces, I
E
EE 802
.11
d
e
v
i
ces an
d
so
o
n
.
Hen
c
e in
terferen
ce can
cellatio
n b
e
t
w
een the m
u
ltib
an
d
OFDM system
an
d o
t
h
e
r b
e
comes a fo
rem
o
st task
.
The de
velopment
of wireles
s
comm
unication
as
a via
b
le
sou
r
ce of inco
m
e
led
to
the adv
e
n
t
of licen
sed
sp
ectru
m
s
. The do
wn
si
d
e
o
f
th
is was th
e inefficien
t
u
s
e of allo
tted
sp
ectru
m
,
wh
ich
is
no
w b
e
i
n
g
ad
dressed
.
b
y
th
e
d
e
v
e
l
o
pmen
t o
f
C
o
gn
itiv
e rad
i
o
[5
].
Sin
ce in
th
e m
u
l
tib
an
d OFDM
syste
m
th
e Fast Fo
urier
Tran
sform
ope
ration is
pe
rform
e
d at the
receiver side
, i
t
can
be
easily
m
odelled as a
cognitive
ra
di
o .They a
d
dress the
pr
o
b
l
e
m
by
scanni
ng a
nd se
n
s
i
ng t
h
e
un
use
d
spect
r
u
m
i
n
the l
i
censed a
r
e
a
and i
n
t
r
o
duci
ng a sec
o
n
d
a
r
y
user
t
o
occ
u
py
i
t
[
6
]
The si
m
p
l
e
st way
of
sensi
n
g t
h
e
u
s
ed
an
d
u
nus
ed
spect
r
u
m
i
s
by
t
h
e
u
s
e o
f
cy
cl
ost
a
t
i
ona
ry
m
e
t
hod,
whi
c
h
com
p
ares t
h
e ener
gy
of t
h
e
si
gnal
t
o
a pre
d
efi
n
ed
val
u
e
and st
at
es w
h
e
t
her t
h
at
spect
r
u
m
i
s
occu
pi
ed
o
r
n
o
t
[
7
]
,
[
8
]
.
Thi
s
pa
per m
a
i
n
l
y
foc
u
ses
on t
h
e pe
rf
o
r
m
a
nce of t
h
i
s
cy
cl
o
s
t
a
t
i
onary
m
e
tho
d
of
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
Mu
ltib
and
OFDM f
o
r C
o
gn
itive Rad
i
o
- A
Wa
y f
o
r Cyclosta
tio
na
ry
Det
ectio
n
a
n
d
In
terferen
ce .... (J.
Avila
)
1
703
sensing
[9]. Si
nce the
UWB
syste
m
in
corporated
data
ove
r a wi
de ra
nge
of
frequ
encies, it is i
m
p
e
rativ
e th
at
th
e in
terferen
ce b
e
tween
t
h
e
p
r
im
ary u
s
er an
d th
e cog
n
itive u
s
er
were li
mited
to
a m
i
n
i
mu
m
.
The
wo
rd
La
bV
IE
W is an
abb
r
e
v
iation
fo
r La
bo
rat
o
ry
Vi
rt
ual
I
n
st
r
u
m
e
nt
at
i
on El
ect
ro
ni
c
Wo
r
kbe
nc
h. T
h
i
s
so
ft
wa
re
uses
G-l
a
ng
ua
ge (
G
ra
p
h
i
cal
User
I
n
t
e
rfac
e
l
a
ng
uag
e
) .
I
t
hel
p
s i
n
vi
r
t
ual
l
y
sim
u
l
a
t
i
ng any
of t
h
e i
n
st
r
u
m
e
nt
s f
unct
i
o
ni
n
g
wi
t
h
o
u
t
act
u
a
l
l
y
havi
ng t
o
use t
h
e
har
d
wa
re com
pon
ent
s
of t
h
e
circuit. T
h
e
VI’s
are t
h
e
program
s
which re
prese
n
t
th
e fu
n
c
tion
i
ng
of th
e in
stru
m
e
n
t
. Each
v
i
has t
w
o
w
i
nd
ow
s
n
a
m
e
d
th
e fr
on
t
p
a
nel an
d b
l
o
c
k pan
e
l. Th
e
f
r
on
t p
a
n
e
l is th
e
rep
r
esen
tatio
n of th
e in
st
ru
m
e
n
t
i.e.
how it will look after
practica
lly bei
ng im
plem
enting the idea. It has all
the controlling
knobs the instrum
e
nt
is expected to have a
nd the
in sim
p
le words it’s
the int
e
rface for the user. The
bloc
k pa
nel has the ful
l
fun
c
tion
s
o
f
the VI. Th
at is it
h
a
s t
h
e fun
c
tion
s
wh
ich
p
e
rform
th
e actio
n
of th
e in
stru
m
e
n
t
[10
]
.
Th
is work
concen
trates o
n
th
e in
terferen
ce
b
e
tween
th
e mu
ltib
an
d
OFDM actin
g
as cog
n
itiv
e
rad
i
o
an
d o
t
h
e
r wireless syste
m
s th
at co
ex
ist in th
e sam
e
env
i
ro
n
m
en
t. Bein
g
a cogn
itiv
e rad
i
o
it
p
e
rform
s
sp
ectru
m
sen
s
in
g (ie) th
e m
o
st v
ital job
o
f
an
y co
gn
itiv
e
rad
i
o
is t
o
sense th
e arri
v
a
l
o
f
licen
sed
u
s
ers an
d
av
ailab
ility o
f
th
e free sp
ectru
m
.
Cyclo
s
tat
i
o
n
a
ry b
a
sed
spectru
m
sen
s
in
g is d
i
scu
ssed
.
Th
e hard
ware t
e
st b
e
d
for t
h
e co
gn
itiv
e
rad
i
o
is setu
p wit
h
th
e
u
s
e of
Un
i
v
ersa
l
so
ft
ware d
e
fine p
e
riph
eral
k
i
t with
Labv
iew as th
e
pr
ocessi
ng
t
o
ol
.
Next
part
ex
pl
ai
ns i
n
det
a
i
l
abo
u
t
t
h
e p
r
op
o
s
ed m
e
t
hodol
o
g
y
.
The l
a
st
pa
rt
deals with the discussi
on
of
res
u
l
t
s
an
d
f
i
nal
l
y
t
h
i
s
rese
arch
co
nt
ri
b
u
t
i
on
ha
s t
h
e
co
n
c
l
udi
n
g
rem
a
rk
s o
n
t
h
e
pr
op
os
ed sy
st
em
.
2.
R
E
SEARC
H M
ETHOD
The pr
o
p
o
s
ed
wo
rk
c
onsi
s
t
s
of
t
w
o
st
age
s
.
C
y
cl
ost
a
t
i
onar
y
base
d s
p
ect
r
u
m
sensi
n
g
t
o
f
i
nd t
h
e e
x
i
s
t
e
n
ce o
f
pri
m
ary
user
Effectiv
e in
terferen
c
e can
cellatio
n
b
e
twee
n
t
h
e
p
r
im
ary u
s
er an
d cogn
itiv
e u
s
er
The bl
oc
k di
agram
of
t
h
e Lab
V
IE
W bas
e
d
cy
cl
ost
a
t
i
o
nary
m
e
t
hod i
s
sh
ow
n
i
n
Fi
gu
re 1. It
com
p
ri
ses o
f
F
F
T bl
ock
,
f
o
l
l
o
we
d
by
wi
nd
owi
n
g
,
co
rr
el
at
i
ng a
nd m
a
gn
i
t
ude s
qua
ri
n
g
.
The
out
put
of
t
h
es
e
bl
oc
ks y
i
el
ds t
h
e p
o
w
e
r s
p
ec
t
r
al
densi
t
y
. T
h
e o
u
t
p
ut
i
s
then com
p
ared
w
ith
th
e pred
efin
ed
th
resho
l
d. If th
e
receive
d signal crosses t
h
e threshold
the
pri
m
ary user is present and vice
versa
.
The
dec
i
sion is m
a
de
on the
hypotheses
H
0
= signal+
noise : Pr
im
ary user
prese
n
t
H
1
=
noi
se
:
p
r
i
m
ary
user
abse
nt
Fi
gu
re 1.
B
l
oc
k di
ag
ram
of
c
y
cl
ost
a
t
i
onary
m
e
t
hod (La
b
vi
ew)
The
proposed bloc
k dia
g
ra
m
for interference ca
n
cellatio
n
is sh
own in
Figu
re 2. Th
e BPSK
m
odulated dat
a
is grouped. The subca
rriers are gro
upe
d
in three ways
.
Adjacent
,
Int
e
rleave
d
and random
p
a
rtitio
n
s
are carried
o
u
t
in
t
h
is wo
rk
. In
ad
j
acen
t
p
a
rtitio
n
th
e sub
carriers wh
ich
are ad
j
acen
t
to
each
o
t
h
e
r
are gro
u
p
e
d
.
In in
terleav
ed
p
a
rtitio
n
altern
at
e sub
carri
ers are gro
u
p
e
d
.In
rando
m
p
a
rtiti
o
n
th
e sub
carriers are
cho
s
en
an
d
gr
ou
pe
d i
n
a ra
n
dom
m
a
nner
.
Out
of
1
2
8
s
u
b
-
t
o
nes i
t
i
s
ass
u
m
e
d t
h
at
t
h
re
e t
one
s nam
e
l
y
8
5
,
8
6
and
8
7
as i
n
t
e
rfe
ri
n
g
t
o
nes.
Si
nce t
h
e
r
e a
r
e 1
28
su
b-t
o
n
e
s t
h
e
num
ber
of
g
r
o
u
p
i
n
g
here i
s
f
o
u
r
w
i
t
h
3
2
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
I
J
ECE
Vo
l. 6
,
N
o
. 4
,
Au
gu
st 2
016
:
17
02
–
1
709
1
704
subcarriers
i
n
each group.
The output
of partitioning block
is
gi
ve
n as
input to AIC
bl
ock. The M
I
MO
AIC
alg
o
rith
m
is g
i
v
e
n b
y
t
h
e transmitted
sig
n
a
l
i
s
written
in
m
a
trix
form
as E=n
m
, wh
ere
T
i
n
n
n
n
)
1
(
........
).........
1
(
)
0
(
(
1
)
T
i
s
m
m
m
m
)
1
(
........
).........
1
(
)
0
(
(2)
E
E
E
E
i
si
i
si
E
,
1
11
.........
1
,
.
.
.
...
..........
(3)
))
(
2
exp(
)
,
(
1
0
s
k
i
N
m
j
l
k
E
N
m
ki
E
2
1
1
min
ar
g
v
E
c
c
c
(
4
)
C is th
e targ
et
AIC
v
ector. The in
terfe
re
nce
vector is
given by
0
,
)
2
(
)
1
(
:
1
1
y
y
f
y
v
r
fr
v
n
(5)
The m
i
nim
i
zat
i
on
p
r
obl
em
i
s
gi
ve
n
by
2
min
F
E
x
x
s
ubject to
2
:
H
G
x
(6)
2
1
1
min
arg
v
E
c
c
c
opt
Subject t
o
:
2
/
,.....
1
;
)
(
2
/
,.....
1
;
)
(
2
2
.
2
1
.
N
c
N
c
ET
j
ET
ET
i
ET
j
k
i
k
(
7
)
2
2
.
1
.
]
[
c
c
c
ET
null
ET
c
(8)
Lagra
n
ge
opt
i
m
i
zat
i
on al
go
ri
t
h
m
i
s
used t
o
opt
i
m
i
ze t
h
e pr
obl
em
whi
c
h
i
s
gi
ve
n a
s
)
)
(
(
)
)
(
(
)
,
,
(
2
2
.
2
/
1
2
2
/
1
1
.
2
1
1
j
ET
j
j
i
i
ET
i
j
i
j
N
N
i
c
c
f
c
c
v
E
ET
ET
(9)
v
E
c
E
E
H
opt
H
1
1
1
1
)
(
(
1
0)
v
E
E
E
c
H
H
opt
1
1
1
1
1
)
(
]
........
0
........
0
.........
[
1
N
diag
ET
(11)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
Mu
ltib
and
OFDM f
o
r C
o
gn
itive Rad
i
o
- A
Wa
y f
o
r Cyclosta
tio
na
ry
Det
ectio
n
a
n
d
In
terferen
ce .... (J.
Avila
)
1
705
Fig
u
r
e
2
.
Blo
ck
d
i
agr
a
m
o
f
pr
opo
sed MI
M
O
A
I
C
3.
R
E
SU
LTS AN
D ANA
LY
SIS
Si
ngl
e i
n
put
si
ngl
e o
u
t
p
ut
(S
ISO
)
bas
e
d A
c
t
i
v
e i
n
t
e
rfere
n
ce al
gori
t
h
m
was p
r
o
p
o
se
d i
n
[1
1]
whi
c
h
was
fu
rt
he
r e
n
hance
d
i
n
by
cy
cl
i
c
shi
f
t
i
n
g
an
d
ph
ase s
h
i
f
t
i
n
g
o
f
t
h
e
da
t
a
[1
2]
.
St
at
i
s
t
i
cal
rel
a
t
i
ons
base
d
Active inter
f
er
ence canceller
(AIC
) was
pr
op
ose
d
in [
1
3]
to im
pro
v
e th
e basic AIC al
go
rithm
.
The use
o
f
p
a
rtial tran
sm
i
t
seq
u
e
n
ce m
i
tig
atio
n
is carried
ou
t in
[14
]
.In
t
ro
du
ction o
f
weigh
ting facto
r
s on
sev
e
ral
subcarriers m
itigation is
carried out inv
[15]. Multi input m
u
lti
output (MIMO)
base
d active interferenc
e
can
cellatio
n
was propo
sed in [16
]
,
wh
ich
was fu
rther enhan
ced in
[1
7
]
.
Th
e m
a
in
fo
cus of th
is
p
a
p
e
r is to
p
r
op
o
s
e a m
o
d
i
fied in
tru
s
ion
elim
in
atio
n
tech
n
i
qu
e fo
r
MIMO
b
a
sed
m
u
l
tib
an
d OFDM co
gn
itiv
e rad
i
o
syste
m
.
The fi
rst
t
w
o r
e
sul
t
s
are pl
ot
t
e
d usi
n
g M
A
T
L
AB
t
ool
. T
h
e
perf
orm
a
nce m
e
t
r
i
c
i
s
subc
arri
er i
n
de
x
vers
us m
a
gni
t
ude
spect
r
u
m
.
The rem
a
i
n
i
n
g
resul
t
s
a
r
e pl
ot
t
e
d u
s
i
n
g La
bV
IE
W t
o
ol
.
Here t
h
e pe
rf
o
r
m
a
nce
metric is Prob
ab
ility o
f
d
e
tectio
n (
d
P
) vers
us
Si
gnal
t
o
Noise
Ratio
(S
NR).
Fig
u
re
3
sh
ows th
e ad
j
acen
t
p
a
rtitio
n
withou
t wei
g
h
ting
facto
r
s. Th
e 128
su
b
c
arriers are gro
u
p
e
d
i
n
t
o
4 g
r
ou
ps.
Ad
jace
nt
car
ri
ers are
g
r
o
u
p
e
d
. T
h
e
n
t
h
e
m
odi
fi
ed
A
I
C
al
g
o
ri
t
h
m
i
s
appl
i
e
d.
Fi
g
u
re
4 s
h
ows
t
h
e
o
u
t
p
u
t
o
f
adj
a
cen
t p
a
rtitio
n
with
wei
g
h
ting
fact
o
r
s. From
th
e fig
u
re it is clear th
at ad
j
acen
t
partitio
n
with
wei
g
ht
i
ng
fact
ors
d
o
m
i
nat
e
s. A dee
p
not
c
h
i
s
achi
e
ve
d.
There i
s
at
l
e
a
s
t
1dB
di
ffe
re
nce bet
w
een
b
o
t
h
t
h
e
cases.
Fi
gu
re 3
a
n
d 4
.
Ad
j
acen
t
p
a
rti
tio
n
with
an
d with
ou
t weigh
tin
g
factors
Fig
u
re 5
sh
ows th
e rand
o
m
p
a
rtitio
n
with
ou
t weig
h
tin
g
facto
r
s. Th
e 128
sub
carriers are g
r
ou
p
e
d
rando
m
l
y. Th
en
th
e m
o
d
i
fied AIC alg
o
rithm is
ap
p
lie
d
.
Fig
u
re 6
shows th
e rando
m p
a
rtitio
n
with
weig
h
ting
fact
or
s. The
1
28 s
u
bcar
ri
ers
are gr
o
upe
d ra
nd
om
l
y
. Th
en the weighting factors ar
e applied
.
Th
e wei
g
h
tin
g
factors are
bas
e
d on the m
o
dulation
sch
e
me u
s
ed
. Si
n
ce in
th
is
work
it
is BPSK m
o
d
u
latio
n
th
e
weigh
ting
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
I
J
ECE
Vo
l. 6
,
N
o
. 4
,
Au
gu
st 2
016
:
17
02
–
1
709
1
706
factors are
chosen as
+1 and
-1.
Afte
r
ap
p
l
yin
g
th
e
weigh
tin
g fact
o
r
s
AIC alg
o
rith
m
is u
s
ed
.
Fro
m
th
e fig
u
res
i
t
i
s
cl
ear t
h
at
a deep
not
c
h
i
s
achi
e
ve
d wi
t
h
t
h
e i
n
cl
usi
on
of
wei
ght
i
ng
fact
ors
.
Dee
p
n
o
t
c
h i
n
t
u
r
n
sy
m
bol
i
zes
th
at th
e i
n
terferen
c
e b
e
t
w
een
th
e prim
ary u
s
er an
d th
e cog
n
itiv
e u
s
ed
is
red
u
c
ed
.
Figure 5
a
n
d 6.
Random
partition
with a
n
d
wit
hout weighting factors
Figu
re 7
sh
o
w
s
the ha
rd
ware
s
e
tup
f
o
r the cy
clos
t
a
t
i
ona
ry
based spect
r
u
m
sensi
n
g.
Here t
w
o U
S
R
P
s
are
use
d
.
One
acts as
trans
m
itter and t
h
e
ot
her acts as
receiver.
USR
P
is a
tra
n
sceiver which c
o
nverts
a
pers
o
n
al
com
put
er i
n
t
o
wi
rel
e
ss pr
ot
ot
y
p
e
sy
st
em
. The b
a
si
c com
pone
n
t
s of t
h
e
USR
P
are R
F
am
pl
i
f
i
e
r,
Lo
cal o
s
cillato
r, ADC
and
DAC.
Th
e
arch
itectu
r
e
also
i
n
clu
d
e
s FPGA fo
r
pro
c
essing
t
h
e
d
a
ta.
U
S
RP 29
20
h
a
s
b
een u
tilized
in
t
h
is work
.
T
h
e
USRPs
were
connected to t
h
eir
host p
e
r
s
on
al com
p
u
t
er
s th
ro
ug
h 2.0
Ethernet cable
. The
IP a
d
dre
ss of th
e tra
n
s
m
itter was set
as 192.168.10.
1 a
n
d that of
receiver was
s
e
t as
192.
168.10.
2
.
At the
receive
r side
, the cy
clostati
ona
ry based s
p
ectrum
sensi
n
g m
odule de
vel
ope
d usi
ng
LabVIEW
was u
tilized
an
d d
e
tectio
n
of
PU
was
d
o
n
e
and
p
l
o
tted
.
Figu
re
7.
Ha
rd
ware
setu
p
Fi
gu
re 8
gi
ves
t
h
e com
p
ari
s
o
n
bet
w
ee
n va
ri
o
u
s FFT si
zes
(
1
2
8
,
2
56
,5
1
2
a
nd
10
2
4
) i
n
cy
cl
ost
a
t
i
ona
ry
m
e
t
hod.
Ham
m
i
ng wi
n
d
o
w
wi
t
h
wi
nd
o
w
s
i
ze of
10
i
s
us
ed. T
h
e
gra
p
h
sho
w
s t
h
at
wi
t
h
i
n
c
r
ease i
n
F
F
T si
ze
p
r
ob
ab
ility o
f
d
e
tectio
n
cou
l
d
b
e
ach
i
ev
ed
earlier. For th
e Pro
b
a
b
ility o
f
d
e
tectio
n (
d
P
)
of
0.
9 t
h
e S
N
R
f
o
r
th
e FFT size of
10
24
is -2
7dB w
h
er
eas the
SNR val
u
e is -22dB
for the
FFT size of
128. As t
h
e FFT size
increases
frequency res
o
luti
on increase
s
an
d it h
e
lp
s in
d
e
tectin
g
th
e prim
ary
use
d
i
n
hi
gh
noi
sy
e
n
vi
ro
n
m
ent
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
Mu
ltib
and
OFDM f
o
r C
o
gn
itive Rad
i
o
- A
Wa
y f
o
r Cyclosta
tio
na
ry
Det
ectio
n
a
n
d
In
terferen
ce .... (J.
Avila
)
1
707
Fi
gu
re
8.
va
ri
o
u
s F
F
T si
zes
Fi
gu
re
9
gi
ves
t
h
e com
p
ari
s
o
n
bet
w
ee
n
vari
ous
wi
nd
o
w
si
zes i
n
cy
cl
ost
a
t
i
onary
m
e
t
hod
. Ham
m
i
ng
wind
ow with
FFT size o
f
512
is u
s
ed. For th
e Pro
a
b
a
b
ility o
f
d
e
tection
(
d
P
) o
f
0
.
9
th
e SNR v
a
lu
e is -25d
B
fo
r the
wi
nd
o
w
size
of
1
0
whereas
for the
sam
e
d
P
th
e SNR v
a
lu
e is -27
d
B for t
h
e
windo
w size
o
f
40
.
The
gra
p
h shows that with increas
e in wi
ndow size the res
o
luti
on i
n
crea
ses a
nd
he
nce t
h
e
p
r
i
m
ary
users coul
d b
e
d
e
tected
i
n
low SNR cond
itio
n
.
Fi
gu
re 9.
v
a
r
i
ou
s w
i
nd
ow
sizes
Fi
gu
re 1
0
gi
v
e
s t
h
e com
p
ari
s
on
bet
w
ee
n vari
ous
wi
n
d
o
w
s i
n
cy
cl
ost
a
t
i
onary
m
e
t
hod. Ham
m
i
n
g
wi
n
d
o
w
, B
l
a
c
k
m
a
n wi
nd
ow
and Kai
s
e
r
W
i
n
d
o
w
are c
o
m
p
ared. T
h
e
gra
ph s
h
o
w
s
Kai
s
er
W
i
nd
o
w
of
fe
rs
b
e
tter
d
P
wh
en
co
m
p
ared
with
o
t
h
e
r wi
n
dows. For th
e Prob
ab
ility o
f
d
e
tectio
n
(
d
P
) of
0
.
9
th
e SNR
v
a
lu
e is
-2
5
d
B
i
n
ca
se
of
Kai
s
e
r
wi
nd
ow
w
h
ere
a
s i
t
i
s
-2
3
d
B
f
o
r
B
l
ackm
a
n wi
nd
ow
. It
i
s
g
o
v
er
ned
by
t
h
e
par
a
m
e
t
e
r
b
e
ta wh
ich h
e
l
p
s in con
t
ro
llin
g th
e
wid
t
h
of th
e m
a
in
lo
be. Greater t
h
e
main
lo
b
e
and lesser th
e side lo
be
depicts that t
h
e spectral leakage is less
. Hence
pr
ope
r
choice of the wi
ndow e
nha
nces t
h
e s
y
ste
m
per
f
o
r
m
a
nce.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
I
J
ECE
Vo
l. 6
,
N
o
. 4
,
Au
gu
st 2
016
:
17
02
–
1
709
1
708
Fi
gu
re 1
0
.
v
ari
ous
wi
nd
o
w
s
Fig
u
re 11
g
i
ves
th
e
co
m
p
ariso
n
b
e
t
w
een v
a
ri
o
u
s
v
a
lu
es
of
Pro
b
a
b
ility
o
f
false alarm (
f
P
) in
cyclostationary
m
e
thod.
For
the Probability of
detection(
d
P
) of
0.
7 t
h
e S
N
R
val
u
e i
s
-
2
8dB
fo
r t
h
e
f
P
of 0.
1
wh
ereas the SNR v
a
l
u
e is -2
6d
B for th
e
f
P
v
a
lu
e
o
f
0
.
000
1. Figur
e sho
w
s lesser
t
h
e
f
P
val
u
e e
r
r
o
r
d
ecreases an
d
p
r
ob
ab
ility o
f
d
e
tectio
n is acco
m
p
lish
e
d
in
h
i
gh
n
o
i
sy environ
m
en
t.
Fig
u
re
11
.v
ari
o
u
s
pro
b
a
b
ility
of false alarm
4.
CO
NCL
USI
O
N
Co
gn
itiv
e
b
e
in
g a so
lu
tion
for efficien
t
utilizat
io
n
of t
h
e sp
arse sp
ectru
m
, th
e p
e
rform
a
n
ce o
f
cy
cl
ost
a
t
i
onary
base
d spect
ru
m
sensi
ng m
e
tho
d
i
s
an
al
yzed. T
o
m
i
tigate the interfe
re
nce betwee
n pri
m
ary
user
an
d sec
o
n
d
ary
user
a m
odi
fi
ed i
n
t
e
rfe
r
e
nce
red
u
ct
i
o
n
al
go
ri
t
h
m
has bee
n
pr
o
pose
d
a
n
d
i
t
i
s
seen t
h
a
t
th
ere is an
im
p
r
ov
em
en
t in
notch
d
e
p
t
h
with
th
e in
ser
tio
n of weigh
ting
faco
trs. All th
ese
tech
n
i
qu
es
enhan
c
e
th
e p
e
rform
a
n
ce o
f
th
e m
u
ltib
an
d
OFDM syste
m
an
d
m
a
k
e
s it
m
o
re ap
p
r
o
p
riate an
d
a
perfect so
l
u
tio
n
wh
en
the noise is
so
severe
that t
h
e
juice
of
h
i
gh
data
r
a
tes
can
b
e
enj
o
yed
.
ACKNOWLE
DGE
M
ENTS
The Authors
would like to expre
ss th
eir sin
cere th
anks to
SASTRA Un
i
v
ersity, fo
r th
e
g
r
an
t
receive
d
unde
r R&M
fund(R
&M / 0027 / SEEE
–
010
/
2012 – 13) to carry ou
t this
rese
arch work.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
Mu
ltib
and
OFDM f
o
r C
o
gn
itive Rad
i
o
- A
Wa
y f
o
r Cyclosta
tio
na
ry
Det
ectio
n
a
n
d
In
terferen
ce .... (J.
Avila
)
1
709
REFERE
NC
ES
[1]
J. Balakr
ishnan,
et al.
, “A multiband OFDM sy
stem
for UWB communication,”
in
IEEE conference on ultrawid
e
band systems an
d technology
, pp
. 354-358
, 2003
. doi: 10.1109
/UWBST.2003.1267863.
[2]
P. Praveenkumar
, et al
.,
“Regulated OFDM-role of
ECC and AN
N-A review,
”
Jo
urnal of Applied
sciences,
vo
l. 1
2
,
pp. 301-314
, 20
12. doi: 10
.3923
/jas.2012
.301.31
4.
[3]
J. Avila,
et al
., “Anti Interference for anti co
rroupt for 3G
spectrum-A mul
ti carrier appro
ach,” in
IEEE
International co
nference on
com
puter communication and
Informatics
, pp. 1-4, 20
12.
[4]
A.
Batra,
et
al
.,
“Multiband OFDM: A cognitiv
e radio for UW
B,” in
Proceedings of International symposium on
circuits
and systems
, pp. 4093-4
097, 2006
.
[5]
S. Ha
ykin,
“
C
ognitive
rad
i
o: B
r
ain
em
powered wirel
e
ss com
m
unica
tions,”
IEEE Journal on
selected
areas in
communications
, vol. 23, pp. 201
-220, 2006
.
[6]
M. Hejaz
i
and
B. Abolhassani,
“
E
ne
rg
y
d
e
te
ct
ion based spect
rum
sensing in cognitiv
e radi
o networks over
spatially
-corr
e
lated channels,” in
IEEE Symposium on
Industrial Ele
c
tronics
&
A
pplicat
ions (
I
SI
EA
), pp.738-743
,
2010. doi:10.110
9/ISIEA.2010.5
679368.
[7]
G. A. F. M. Khalaf, “
A
n optim
al
Sinsing
algorith
m
for Multibnad Cognitive R
a
di
o Network,”
International Journa
l
of Information
&
Network Secu
rity
, vol. 2
,
pp
. 6
0
-67, 2013
.
[8]
W.
S.
Sha
n
,
et al
.,
“
P
rim
a
r
y
U
s
er Em
ulat
ion
Attacks Ana
l
y
s
i
s
for Cognitiv
e
Radio Networks
Com
m
unicatio
n,”
TELKOMNIKA Indonesion Journ
a
l
of Electrical
Engineering
, vol. 11
, pp
. 3905-3
914, 2013
.
[9]
V. Prithivira
j
,
et al.
, “C
y
c
lost
at
ionar
y
Anal
y
s
is
Method of Spectrum
Sensing for Cognitive
Radio,” in
IEEE
International
co
nference on wireless
communication, vehicular techno
logy,
In
fo
rmation theory and Aerospac
e
&
Electronic system techno
logy
, p
p
. 1-5
,
2011
. doi
: 10.1109
/W
IRELESSVITAE.20
11.5940821.
[10]
J. Avila and K. Thenm
o
zhi, “Cognitive (L
abVIE
W
) on
FP
GA–a
novel approach for d
y
nam
i
c spectrum
sensing,”
International Jo
urnal of
electron
ic letters,
vo
l. 3,
pp. 13-23
, 2014
. doi:10
.
1080/21
681724.901422
[11]
H. Yam
a
guchi,
“
A
ctive Inter
f
er
ence C
a
nc
ell
a
tio
n techn
i
que for
MB-OFDM Cognitive r
a
dio
,
”
i
n
Pr
oceed
ings
of
34th European
Microwave con
f
erence
, pp
. 1105
-1108, 2004
.
[12]
P.
T
a
ra
sa
k,
et al
.
,
“
F
urther Enhancem
ent for Activ
e Interfer
e
nce Cancellation on MB-O
FDM
U
W
B
Tra
n
smission,
”
in
IEEE 68
th
con
f
erence on
vehicular technology
,
pp. 1-5
,
2008
. d
o
i:
10
.1109/VETECF.2008.244.
[13]
S.
G.
Huang and C.
H.
Hwang,
“Imp
rovement of Active Inter
f
erence Ca
n
cellation: Avoidance Techniqu
e for
OFDM Cognitive Rad
i
o,
”
IEEE Transactions
on Wireless
Communication
, vo
l. 8,
pp. 5928-5937
,
2009.
[14]
P. Tar
a
sak,
et
al
.
,
“
P
artia
l Tr
ans
m
it S
e
quenc
e-Activ
e Int
e
rfer
e
nc
e Can
cel
lat
i
on f
o
r UW
B OF
DM Trans
m
is
s
i
on,”
i
n
IEEE 20
th
Int.
Symposium on personal Indoor
and Mobile Radio Communications
, pp
. 943-
947, 2009. doi:
10.1109/PIMRC.2009.5450339
.
[15]
A. Minakawa,
et
al
., “A stud
y
on interfer
e
nce
suppression usin
g we
ighting factors for MB-OFDM sy
stems,” in
IEEE Internatio
nal Conf
erence
on Ultrawide ba
nd
, pp
. 88-92
, 2
009. doi: 10
.110
9/ICUWB.2009.5288702.
[16]
F. Sarabchi and C. Nerguizi
an, “
I
nterfer
e
nc
e can
cel
lat
i
on Techn
i
que for MIMO
MB-OFDM
UWB Cognitive Rad
i
o
Sy
s
t
e
m
,
”
i
n
6
t
h Internationa
l Conferen
ce
Wireless and
Mobile Communications
, pp.
472-477, 2010.
doi: 10
.1109/IC
WMC.2010.53
.
[17]
J. Avila
,
et
al
., “ANN assisted-Augumentatio
n of AIC
for
MIMO Multiband OFDM Sy
stem,”
in
IE
EE
International
co
nference on
Advances
in
Eng
i
neering, S
c
ience a
nd Management
, pp. 228-232, 20
12.
BIOGRAP
HI
ES
OF AUTH
ORS
J. Avila
r
ece
iv
ed her B
.
E (E
CE) from
the
V.M
.
K.V coll
e
g
e of Engin
eer
ing and M
.
E
(Com
m
unication
Engineer
ing) fr
om
Vellore Institu
te Of Technol
og
y
.
Currentl
y
s
h
e is working
as Assistant Prof
essor III in the Department of
ECE in SASTRA
University
, Th
an
javur. She has
a te
aching
expe
rienc
e
of 13
ye
ars
and s
h
e has
publis
hed 19 R
e
s
earch
art
i
cl
es
in Nation
a
l &
Interna
tiona
l jo
urnals. She is
current
l
y
wo
rking towards her
Ph.D. Degree in SASTRA
Universit
y
.
Her r
e
search
ar
ea
inc
l
udes W
i
rel
e
ss co
m
m
unication an
d Cognitiv
e r
a
di
o
K. Thenmoz
h
i
receiv
e
d her
B.E (ECE) and
M
.
E (Com
m
u
nica
tion s
y
s
t
em
) degrees
from
Regional
Eng
i
n
eering
col
l
eg
e
(NIT) Tiru
chir
a
ppalli
and
P
h
.
D
. from
S
A
S
T
RA Univers
i
t
y
,
Thanj
a
vur.
Curr
entl
y s
h
e is
wo
rking as
an As
s
o
ciat
e De
an in
the
Depar
t
m
e
nt of
ECE
in
S
A
S
T
RA Univers
i
t
y
,
Thanj
a
vur.
S
h
e has
a t
e
a
c
hi
ng exper
i
en
ce of
23
ye
ars
.
Her
c
u
rrent r
e
s
ear
ch
area
includ
es Wireless communication
,
Steganogr
ap
h
y
and Information Th
eor
y
an
d Coding. Sh
e
has supervised more than 100 UG projects,
10 Master Stud
ents and Supervising 5 Ph.D.
Scholars. So far
she has published 50 Research
artic
les in Na
tional
& Intern
a
tional
journals
conferen
ces. She receiv
ed EDI
award from the br
oadcast Engin
e
ering Society
for the
y
e
ar 2007
.
Evaluation Warning : The document was created with Spire.PDF for Python.