Indonesian Journal of
Electrical
Engineer
ing and
Computer Science
V
o
l. 10
, No
. 3, Jun
e
20
18
, pp
. 10
45
~
1
052
ISSN: 2502-4752,
DOI: 10.115
91/ijeecs
.v10.i
3.pp1045-1052
1
045
Jo
urn
a
l
h
o
me
pa
ge
: http://iaescore.c
om/jo
urnals/index.php/ijeecs
PAPR Reduction Techniques in
Generali
zed Invers
e Di
scret
e
Fourier Transform Non-orth
ogonal Frequency Division
Multiplexing System
Ira
m
Ma
isa
r
ah
Mo
khta
r
1
, Nor
ulhusn
a
Ahmad
2
, Ha
zila
h Ma
d Ka
idi
3
,
M
o
hd Az
r
i
M
o
hd I
z
har
4
M
o
hd,
Norliz
a Mohamed
5
1
Facult
y
of UT
M Razak
School
of Eng
i
ne
ering
& Advanc
ed
Te
chnolog
y
,
Unive
r
siti Teknolog
i Mala
y
s
ia
,
Ma
la
ysia
2,3,4,5
Electrical D
e
partment, UTM
Razak Schoo
l o
f
Engin
eer
ing
&
Advanced
Techn
o
log
y
, Univ
ersiti Tekno
logi
Mala
y
s
ia
, Kua
l
a
Lum
pur, Mal
a
y
s
ia.
Article Info
A
BSTRAC
T
Article histo
r
y:
Received Ja
n
7, 2018
Rev
ised
Mar
20
, 20
18
Accepte
d
Mar 31, 201
8
A promising s
y
stem of Generalized Inverse Discr
e
te Fourier
Tran
sform Non-
Orthogonal Freq
uency
Division
Multiplex
i
ng (GIDFT n-OFDM)
s
y
stem
can
fulfil th
e requ
irement of supporting higher
data r
a
te
in Fifth Gen
e
ration (5G)
techno
log
y
. However, this
s
y
s
t
em
experienc
e
High P
eak to Average P
o
we
r
Ratio (PAPR) due to m
a
ssive num
b
er of subcarriers signal is tra
n
sm
itted. In
this paper,
three ty
p
e
s of usual PAPR
reduction techniqu
es were applied in
GIDFT n-OFDM sy
stem
whic
h are Clipping
,
Partial tr
ansm
it Transform
(P
TS
) and S
e
lec
tive M
a
pping (S
LM
). The s
y
s
t
e
m
perform
ance is
com
p
ared
and ev
aluated
using Complementar
y
Cumulative Distribu
tion
Function
(CCDF) plot. Si
mulation results
show
that SLM technique give significan
t
reduction of
PAPR 9 dB of
the o
riginal perfo
rmance.
K
eyw
ords
:
5G
C
l
i
ppi
n
g
GI
DFT
n
-O
F
D
M
PTS
SLM
Copyright ©
201
8 Institut
e
o
f
Ad
vanced
Engin
eer
ing and S
c
i
e
nce.
All rights re
se
rve
d
.
Co
rresp
ond
ing
Autho
r
:
Iram
Maisarah
Bin
ti Mo
kh
tar,
UTM
R
aza
k S
c
ho
ol
of
En
gi
n
eeri
n
g &
A
d
va
nced
Tec
h
n
o
l
o
gy
,
UTM
Kuala L
u
m
pur,
Level
7, R
aza
k
To
wer
Jal
a
n
S
u
l
t
a
n
Yahy
a
Pe
tra 54100
Kuala Lum
pur, Mal
aysia.
Em
a
il: ira
mma
isarah@g
m
a
il.
co
m
1.
INTRODUCTION
The i
d
ea
of
5G
i
s
t
o
ful
f
i
l
t
h
e dem
a
nds o
f
m
o
re
net
w
or
k ca
paci
t
y
by
20
20
. 5G
nee
d
s t
o
have
hi
g
h
e
r
carrier
fre
quencies with bigge
r
ba
ndwidt
hs t
o
support
hi
gh
data rates [1].
There
f
ore, m
u
lticarrier system
s
are
capabl
e
t
e
c
h
no
l
ogi
es t
h
at
ha
v
e
bee
n
di
sco
v
e
r
ed
gi
ve
hi
g
h
er
ba
nd
wi
dt
h e
f
f
i
ci
ency
i
n
P
h
y
s
i
cal
(PH
Y
)
l
a
y
e
r[2]
.
Ort
h
ogonal F
r
eque
ncy Di
vis
i
on M
u
ltiplexing
(OFDM
)
is
the m
o
st recognize
d m
u
lticarrier system
that has
becom
e
the leading technol
ogy for
hi
g
h
sp
eed wi
rel
e
ss c
o
m
m
uni
cat
i
on[
1]
,
[3
].
Th
e
fu
nd
am
ental idea of
OF
DM is to split the bit stream
s
into
m
a
ny
of s
u
bstream
s. Each
of the
se s
ubst
r
eam
s will
m
odulate a different
carrier
with low data rate. T
h
is process can
reduce a lo
t o
f
Inter
sy
m
bol In
terfere
nce (
I
S
I
) with a d
r
a
w
b
ack o
f
hig
h
inte
rcar
rier inter
f
ere
n
ce
(IC
I)
[4]
,
[
6
]
.
The
r
ef
o
r
e, i
t
i
s
neede
d
a
p
r
ope
r
desi
g
n
of
cha
nnel
c
o
di
n
g
a
n
d
eq
u
a
lization
to ov
erco
m
e
th
e p
r
ob
lem
.
In
o
r
d
e
r t
o
ach
i
ev
e
5
G
stand
a
rd
, exp
l
o
itatio
n
o
f
o
r
t
h
ogon
ality tech
n
i
que in
m
u
lt
icarrier can
b
e
a
pr
om
i
s
i
ng t
ech
nol
ogy
[
1
]
-
[
2
]
,
[
4
]
. T
h
e a
d
va
nt
ages
o
f
n
o
n
-
ort
h
o
g
onal
sy
st
em
s i
s
bet
t
e
r
ban
d
w
i
d
t
h
ef
fi
ci
ency
[4]
, [
6
]
as t
h
e IC
I can
ful
f
i
l
t
h
e un
use
d
subca
rri
e
r
sp
aci
ng t
o
ben
e
fi
t
t
h
e req
u
i
r
e
d
ba
n
d
wi
dt
h[
2
]
. The
m
odulation ca
n be categori
sed as filtered [7] and
Di
screte Fourier Trans
f
orm
(
D
FT
) base
d.
One
of
m
odulation that using s
q
uare
root
of Nyquist filter as c
arrier was
first
introduce
d by
Saltzberg
[8]. Filte
r
B
a
nk M
u
l
t
i
carri
er (
F
B
M
C
)
m
odul
at
i
on t
e
chni
que
s we
re
i
n
t
r
o
d
u
ced
e
v
en
be
fo
re t
h
e i
n
t
r
o
d
u
ct
i
o
n
of t
h
e
OFDM [9].
It is an alternative
and ca
n be a
replacem
en
t to
OFDM in orde
r to fulf
il trans
m
itting data pa
cket
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
502
-47
52
I
ndo
n
e
sian
J Elec Eng
& Com
p
Sci, V
o
l. 10
,
No
.
3
,
Jun
e
2
018
:
10
45
–
1
052
1
046
at
a ve
ry
hi
gh
rat
e
[
8
]
. T
h
i
s
t
echni
que
wa
s
devel
ope
d t
o
p
reve
n
t
bot
h I
S
I
an
d
IC
I,
suc
h
as di
s
p
ersi
on
i
n
bot
h
ti
m
e
an
d
freq
u
en
cy. Th
e fu
nd
am
en
tal o
f
FBMC is a set
o
f
p
a
rallel d
a
ta sym
b
o
l
s were tran
sm
itted
th
roug
h
a
ban
k
of
m
odul
at
ed
fi
lters. The u
s
e
of th
is
fi
lter is m
o
re efficien
t in
sp
ectru
m
u
tilizat
io
n
wh
ere it m
i
n
i
mized
in
terferen
ce acro
ss th
e su
b
c
arriers. An
o
t
her filter b
a
sed tech
n
i
q
u
e is, Gen
e
ralized
Freq
u
e
n
cy Div
i
sion
Mu
ltip
lex
i
ng
(GFDM) th
at
was stud
ied
by Fettweis et
al
. is b
ased
on
flex
i
b
le pu
lse sh
ap
i
n
g
[11
]
. Th
is
tech
n
iqu
e m
ig
h
t b
e
an
en
cou
r
ag
ing
techno
log
y
for
p
o
s
i
tio
n
ing
in
co
gn
itiv
e rad
io
network
s
and
fo
r
5
G
wi
rel
e
ss com
m
uni
cat
i
on [1
1]
, [7]
. Th
e t
echni
que al
l
o
ws t
h
e fl
exi
b
l
e
pul
se sha
p
i
n
g of t
h
e s
p
ect
rum
fo
r
effi
ci
ent
m
u
l
t
i
user
sche
d
u
l
i
n
g,
whe
r
e eac
h
si
ngl
e s
u
bcar
r
i
er can
be m
odul
at
ed
i
n
di
vi
dual
l
y
.
Ho
we
v
e
r, t
h
e
per
f
o
r
m
a
nce anal
y
s
i
s
of
G
F
DM
i
s
l
o
nge
r
6.
25
t
i
m
e
s
com
p
ares to OFDM due t
o
la
rge
r
FFT size
d
used.
GF
DM
o
ffe
r l
o
wer
PA
PR
t
h
a
n
OF
DM
,
but
t
h
e
per
fo
rm
a
nce m
a
y
be re
d
u
c
e
d i
f
P
A
PR
t
e
c
hni
que
i
s
a
ppl
i
e
d.
As filter bank base
d have
be
com
e
m
o
re popula
r
th
is days, DFT base
d m
o
du
lation als
o
now ra
pidly
ev
o
l
v
i
ng
. Sp
ectrally
Efficien
t
Frequ
e
n
c
y Div
i
sion
M
u
ltip
lex
i
ng
(SEFDM) was
an
in
terestin
g
tech
n
i
q
u
e
in
v
e
stig
ated
b
y
M. Rod
r
i
g
u
e
s
an
d I. Darwazeh
wh
ere
it sav
e
th
e spectral b
y
m
u
ltip
lex
e
d
fi
rst th
e
sub
c
arriers
to
cr
eate n
e
w
sy
m
b
o
l
an
d
r
e
located
it at n
ear
er
sp
aci
n
g
[
3
],
[
6
]. Th
e m
a
in
i
d
ea of
th
is techn
i
qu
e is to
ab
an
don
t
h
e ort
h
o
g
onal
i
t
y
and sy
nchr
oni
sat
i
o
n o
f
t
h
e us
ual
pract
i
ce of OF
DM
[3]
. B
e
si
des SEFDM
, O
v
er
l
a
ppe
d
Freq
u
e
n
c
y Di
v
i
sion
Mu
ltip
lex
i
ng
(OvFDM) wh
ich
was in
trod
u
c
ed
by J.
W
a
ng
et al h
a
s sho
w
n
th
at th
e
ban
d
w
i
d
t
h
can
be l
a
r
g
e
r
t
h
a
n
t
h
e f
r
e
que
ncy
co
here
nce
ba
nd
wi
dt
h al
t
h
ou
gh
t
h
e s
u
bcar
ri
ers
were
o
v
erl
a
ppe
d
t
oget
h
e
r
[
1
2]
.
Next
pr
om
i
s
i
n
g m
u
l
t
i
carri
er t
echni
que t
h
at
was st
u
d
i
e
d
by
N.
Ahm
a
d et
al
[1
3]
i
s
GI
D
F
T n
-
OF
DM
w
h
i
c
h i
s
m
o
re ba
nd
wi
dt
h ef
fi
ci
ent com
p
are to OFDM syste
m
. This ap
pro
ach
ed
allo
w m
o
re d
a
ta to
be
tran
sm
it
ted
wi
th
fewer
sam
p
les wh
ich
can u
tilize th
e b
a
n
d
wid
t
h
m
o
re efficien
tly. Th
e
GIDFT
n-OFDM
sy
st
em
used sel
f-canc
e
l
l
a
t
i
on m
i
nim
u
m
m
ean squa
re
d
erro
r (SC
-
M
M
SE) t
u
r
b
o equal
i
zer t
o
de
t
ect
t
h
e
tran
sm
it
ted
sign
al wh
ich sh
ows
b
e
tter erro
r
rate
pe
rform
a
nce eve
n
at
higher IC
I
occurre
nce.
Ev
en
tho
ugh
no
n-o
r
thog
on
al
m
u
l
ticar
r
ier
s giv
e
s pr
o
m
isin
g r
e
su
lt in
h
igher
d
a
ta tr
an
sm
i
ssio
n
, h
i
g
h
Peak
to
Av
erag
e Power Ratio
(PAPR) still
an
issu
e th
at
need
to
b
e
con
s
i
d
ered
. An
an
al
ysis co
n
d
u
c
ted b
y
S.
S. K. C
. B
u
l
u
s
u
et
al
[14]
ha
d
i
n
t
r
o
duce
d
a P
A
PR
re
duct
i
on
t
echni
q
u
e w
h
i
c
h i
s
Di
spe
r
si
v
e
Sel
ect
i
v
e M
a
ppi
ng
(DS
L
M
)
t
o
de
crease t
h
e P
A
PR
val
u
e i
n
F
B
M
C
. It
was
fo
u
nd t
h
at
, t
h
e
m
e
t
hod
pr
o
d
u
ced a
ve
ry
n
ear t
o
per
f
o
r
m
a
nce of OF
DM
wi
t
h
con
v
e
n
t
i
onal
SLM
t
echni
qu
e. The
dra
w
ba
ck of the
DSLM is th
at th
e
PAPR
sh
ows
no
reductio
n
if it is cal
cu
lated
in
t
h
e
cu
rren
t sym
b
o
l
p
eri
o
d
du
ration
as m
o
st o
f
t
h
e sy
m
b
o
l
en
erg
y
lies
in the t
w
o suc
cessive sym
bol pe
riods.
Whereas in
GFDM , Z. Sh
arifian et
al
[7]
ca
rri
e
d
out
a
n
i
nve
st
i
g
at
i
on
Pol
y
n
o
m
i
al
-ba
s
ed
C
o
m
p
an
di
ng Tech
ni
q
u
e (PC
T
)
.
F
o
r SE
FDM
, I. Isam
and I. Da
rwaz
eh
[
1
5]
has
c
o
nd
uct
e
d
sev
e
ral PAPR redu
ction
techn
i
qu
es
su
ch
as
C
l
i
ppi
n
g
, Sel
e
ct
i
v
e M
a
ppi
ng
(SLM), Pa
rtial Tra
n
sm
it Transform
(
P
TS)
and
a
new
techn
i
qu
e,
Slid
in
g
W
i
ndow
(
S
LW
)
in
the SEFD
M syst
e
m
. A
f
ter
n
u
m
er
ical si
m
u
lati
o
n
s
, it
sho
w
e
d
t
h
at
S
L
W
ha
d s
h
o
w
n
a g
r
eat
am
oun
t
of
PA
PR
r
e
d
u
c
t
i
on
po
we
r
wi
t
h
o
u
t
si
de
ef
fe
ct
s.
Wh
ereas, to
date, in
GIDFT n
-
OFDM syste
m
, PAPR
an
al
y
s
i
s
has n
o
t
bee
n
f
o
rm
al
ly
st
udy
y
e
t
.
Desp
ite th
e imp
o
rtan
t of PAPR redu
ction
in
m
u
lticarri
er, t
h
ere ha
s bee
n
n
o
resea
r
c
h
i
n
im
pl
em
ent
i
ng a
n
y
PAPR tech
n
i
qu
es in
t
h
is GIDFT
n
-
OFDM
. In
th
is p
a
pe
r,
t
h
e G
I
D
F
T n
-
OD
FM
sy
st
em
m
odel
an
d P
A
PR
wi
l
l
b
e
d
e
scri
b
e
d i
n
section
2
.
In
section
3
,
th
e
p
r
op
o
s
ed
PAPR redu
ction
techn
iqu
es will
b
e
d
iscu
s
sed
.
Fo
r
sectio
n
4
sim
u
latio
n
resu
lt and
d
i
scu
ssi
on
will b
e
presen
ted and
th
e con
c
l
u
sio
n
in
section
5
.
2.
SYSTE
M
MO
DEL
2.1.
Generaliz
ed Inverse
Discrete Four
ier Tran
sfor
m Non-Orthogonal
Frequency Divisi
on
Multiple
xing (
G
IDFT
n-OF
DM)
Thi
s
sect
i
on e
xpl
ai
ni
ng t
h
e
p
rom
i
s
i
ng t
echn
i
que o
f
GID
F
T
n-
OFDM
sy
st
em
. GIDFT
n-
OFDM
can
trans
m
it large data inform
ation
by exploiting the ort
hogonality with huge
ove
rl
apping am
ong those
subcarriers
[13]. The sam
p
le is smaller
o
r
less th
an
th
e sub
c
arrier and
m
a
tr
ix
size o
f
th
is t
ech
n
i
q
u
e
tran
sfo
r
m
e
d
t
o
rect
angul
ar
s
i
ze or ‘
fat
’ rat
h
er t
h
an s
qua
re
si
ze such as i
n
OFDM
. Fi
g
u
re
1 s
h
o
w
s t
h
e
bl
ock
di
agram
of basi
c
GIDFT
n-
OF
D
M
sy
stem
.
Fi
gu
re
1.
B
l
oc
k
di
ag
ram
of b
a
si
c GI
DF
T
n-
OF
DM
t
ra
n
sm
it
t
e
r
The
n-
OF
DM
si
gnal
i
s
e
x
pre
ssed as:
Evaluation Warning : The document was created with Spire.PDF for Python.
In
d
onesi
a
n
J
E
l
ec En
g &
C
o
m
p
Sci
ISS
N
:
2
5
0
2
-
47
52
PAPR Re
duction Tec
hni
que
s
in Ge
ner
a
lized Inverse
Discre
te Fourier… (Iram M
a
isar
ah
Mok
h
tar)
1
047
(1
)
Where
i
s
t
o
t
a
l
num
ber of
subcarri
er,
,
is th
e
subcarri
er in the
n-
OFDM
sy
m
bol, while
√
1
,
whereas
∆
is the guard fre
quency of the s
u
bcarriers, and
i
s
t
h
e n-OF
D
M
sym
bol
dur
at
i
on. I
n
orde
r t
o
ha
ve n
on-
ort
h
og
onal
si
gnal
,
t
h
e
m
u
st be less than one so that
∆
is narrower c
o
mpared to
OFDM.
Th
erefore th
e b
an
d
w
id
th
will
b
e u
s
ed
efficien
tly as
sh
o
w
n
i
n
(2
) as
will deter
m
ines the b
a
ndwidth efficiency
fact
or and
gi
ve
n as (
3
)
wi
t
h
i
s
t
h
e t
oot
al
num
ber o
f
sam
p
li
ng.
(
2
)
(
3
)
The
discrete ti
m
e
GI
DFT
n
-
OF
DM
is e
x
p
r
esses as
(4
)
whe
r
e
is th
e
ti
m
e sa
m
p
le o
f
th
e first sym
b
o
l
of
in (1
) wi
th
0
,
1
,
2
,
…
1
and
is
the norm
alization fact
or. For
m
o
re
sim
p
lific
ation, t
h
e
received signal ca
n
be e
x
presse
d a
s
(
5
)
whe
r
e
,
,…
is t
h
e
received signal
vect
or,
,
,…
is th
e inp
u
t
sign
al
vecto
r
,
,
,…
is zero-m
ean com
p
lex
White
Ga
us
sian noise
with va
riance
,
2.
2.
Peak
t
o
Aver
a
reg
e P
o
wer
Rati
o (PAP
R)
Peak to
Av
erag
e
Power Ratio
(PAPR) issue m
u
st b
e
taken
care seriously as it will l
ead
t
o
h
uge
flu
c
tu
ation
in
m
u
l
ticarrier mo
du
latio
n
system o
f
OF
DM
[
1
6]
. PA
PR
occ
u
rre
d f
rom
the cum
u
lative peaks that
were
produ
ce
wh
en
m
u
ltip
le sin
u
so
i
d
s are
ad
d
e
d
t
o
g
e
t
h
er to
form
th
e
mu
lticarrier[17
]
.
Wh
en all th
e
h
i
gh
er
peak
of the m
u
lticarrier
m
eet
each othe
r at t
h
e sam
e
ti
me, t
h
e high PAPR will occu
rred[18]. The PAPR of the
t
r
ansm
i
t
t
e
d OF
DM
si
g
n
al
ca
n
be
defi
ned
as:
(
6
)
The PAPR
i
n
d
B
i
s
defi
ned as,
(7
)
3.
PROPOSE
D
PAP
R
REDUCTIO
N
TECHN
I
QUES
IN
GID
FT
N-O
F
DM
S
Y
STEM
In
th
is section
,
th
ree typ
es
o
f
PAPR
redu
ction
tech
n
iqu
es that were pro
p
o
s
ed
will b
e
d
e
scrib
e
d
.
Th
e
techniq
u
es
use
d
we
re p
r
e
v
io
u
s
ly
used in
OF
DM
we
re
app
lied
to
GIDFT
n-OFDM system
. It is
to
in
v
e
stig
ate
the efficiency
of the tec
hni
ques in t
h
is ne
w
s
y
ste
m
and to a
n
alyse the
perform
ance of the
PAPR.
3.
1.
Clipping
Clipping is the sim
p
lest and easy technique that
has been done to re
duce PAPR [19]. The basic
k
nowledg
e o
f
clip
p
i
ng
is where it will
clip
th
e a
m
p
litu
d
e
lev
e
l to
a cer
tain
v
a
lu
e th
at h
a
s b
een
set. It is
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
502
-47
52
I
ndo
n
e
sian
J Elec Eng
& Com
p
Sci, V
o
l. 10
,
No
.
3
,
Jun
e
2
018
:
10
45
–
1
052
1
048
determ
ine as Clipping Le
vel (CL) [18]or
can be calle
d cut-off value which decides
t
h
e
e
x
ceedi
n
g
pea
k
s
to be
cl
i
pped
[
1
7]
. T
h
i
s
cl
i
p
pi
n
g
pr
ocess ca
n
be
d
e
fi
ne
d as:
(
8
)
In th
is
p
a
p
e
r, clip
p
ing
lev
e
l is set as
and
it i
s
a
p
o
s
itiv
e real n
u
m
b
e
r and
∠
is th
e an
g
l
e of
. The
n
,
i
s
G
I
D
F
T
n-
O
F
DM
cl
i
p
pe
d
si
gnal
Des
p
i
t
e
s o
f
t
h
e
si
m
p
lest
t
ech
ni
q
u
e,
cl
i
ppi
n
g
t
e
c
hni
que
t
e
n
d
t
o
p
r
od
u
ce h
i
g
h
bit error
rate (B
ER)
d
egrad
ati
o
n
as it w
ill clip
p
e
d m
o
st o
f
t
h
e i
m
p
o
r
tan
t
sign
al d
a
ta
[2
0
]
.
3.
2.
Sel
ecti
v
e
M
a
p
p
i
n
g.
Selectiv
e Map
p
i
ng
is wid
el
y
u
s
ed
in
m
u
lticarri
er system to reduce
th
e PAPR
[1
5]
,[
19]
. T
h
i
s
tech
n
i
qu
e
reduced
th
e
PAPR with
ou
t requ
ired
m
o
re po
we
r and
su
fferi
n
g fro
m
d
ata rate lo
ss.
In
itial idea of
SLM
i
s
t
o
const
r
uct
al
t
e
rna
t
i
v
e t
r
ansm
it
seque
nces f
r
om
t
h
e sam
e
data sou
r
ce and
t
h
e seque
nce
of t
h
e
tran
sm
it
ted
data v
ectors
will b
e
m
u
lti
p
lied
with
the d
a
ta
v
ectors b
y
so
m
e
ran
d
o
m
p
h
a
se
,
,…,
where
or
1,
. Wh
en
t
h
e m
u
ltip
licatio
n
is do
n
e
, in
o
r
der to
tran
sform
th
e sig
n
a
l
frequ
ency d
o
m
ain
to
ti
m
e
d
o
m
a
i
n
,
th
e m
u
ltip
lied
d
a
ta
will g
o
th
ro
ugh
Inv
e
rse Fast
Fou
r
ier Transfo
r
m
(IFFT). Th
e resu
lt will
g
en
e
rate th
e data b
l
o
c
k
o
f
GIDFT
n
-
OFDM syste
m
th
at h
a
s
d
i
fferen
t
ti
m
e
d
o
m
ain
sig
n
a
ls an
d
th
is
pro
cess will ch
an
g
e
th
e PAPR prop
er
ties. After th
e IFFT p
r
o
c
ess, th
e
lowest
PAPR
a
m
ong the i
n
de
pende
n
t data
blocks
and t
h
e c
a
ndi
date wit
h
t
h
e lowest
PAP
R
will be c
o
m
p
are
d
and
to be selected
for t
r
ansm
ission as
.
Fi
gu
re
2
s
h
o
w
s t
h
e
bl
ock
di
ag
ram
of
se
l
ect
i
v
e m
a
ppi
n
g
technique.
Fi
gu
re
2.
B
l
oc
k
di
ag
ram
of G
IDF
T
n-
OF
D
M
fo
r S
L
M
t
e
c
hni
que
3.3.
Parti
a
l Trans
mit Sequence
(PTS)
In
Part
i
a
l
Tra
n
sm
i
t
Seque
nce
(PTS
) ,
t
h
e i
n
p
u
t
dat
a
bl
oc
k
o
f X
i
s
di
vi
de
d i
n
t
o
a
N
o
f
di
sj
oi
nt
sy
m
bol
sub
s
eq
ue
nce o
r
s
u
b
-
bl
oc
ks [
1
8]
,[
21]
.
T
h
i
s
c
a
n be re
prese
n
t
e
d by
t
h
e
vect
o
r
…
,
therefore:
(
9
)
The
n
, t
h
e s
u
b
b
l
ocks
were transform
ed
in
to
ti
m
e
-d
o
m
ain
p
a
rtial tran
sm
its
sequ
en
ces throu
g
h
t
h
e IFFT
process
.
T
h
es
e pa
rtial se
que
nces
are
in
depe
n
d
ent
l
y
r
o
t
a
t
e
d
by
p
h
ase
fact
o
r
s
,
∈
|
,…
. Nex
t
, t
h
e m
u
ltip
lied
resu
lt sig
n
a
l
sub
s
eq
u
e
nces are
su
mm
e
d
as,
(
1
0
)
Evaluation Warning : The document was created with Spire.PDF for Python.
In
d
onesi
a
n
J
E
l
ec En
g &
C
o
m
p
Sci
ISS
N
:
2
5
0
2
-
47
52
PAPR Re
duction Tec
hni
que
s
in Ge
ner
a
lized Inverse
Discre
te Fourier… (Iram M
a
isar
ah
Mok
h
tar)
1
049
Ass
u
m
i
ng that there a
r
e
W=4
phase
angl
es to
be
allowed. Befo
re tran
sm
itt
in
g
the sig
n
a
l, side
in
fo
rm
atio
n
will b
e
ad
d
e
d
.
After th
at, th
e PAPR is calcu
late
d
fo
r each
resultin
g
sequ
en
ce
an
d
th
en
th
e lowest
PAPR
of the
signal se
que
n
ce
was selected a
n
d transm
itted as
. Fi
gu
re
3 s
h
ows
t
h
e
bl
oc
k
di
ag
ram
of t
h
e
PTS t
e
c
hni
que
.
Fi
gu
re
3.
B
l
oc
k
di
ag
ram
of G
I
DF
T
n-
OF
D
M
base
PTS
t
echni
que
3.
SIM
U
LATI
O
N
AN
D RES
U
LTS
A
NAL
Y
SIS
Th
is section
presen
ts th
e simu
latio
n
resu
lts
o
f
GI
DFT
n-
O
F
DM
sy
st
em
wi
t
h
t
h
re
e di
f
f
e
rent
t
y
pes
of
PAPR
re
d
u
ct
i
on
usi
n
g M
A
TLAB
si
m
u
l
a
ti
on. Ta
bl
e 1 s
u
m
m
a
ri
zed t
h
e sim
u
l
a
t
i
on param
e
t
e
rs for PAP
R
analy
s
is
in GI
DFT n
-
O
F
DM
sy
stem
.
Tabl
e 1. Si
m
u
lat
i
on
Pa
ram
e
t
e
rs
Para
m
eters
Type/Value
M
odulation
BPSK
Nu
m
b
e
r
of sa
m
p
le
s (
)
32
Nu
m
b
er
of blocks
8192
Nu
m
b
e
r
of
subcarr
iers (
)
⁄
Over
lapping factor
0.
2,
0.
4,
0.
6,
0.
8
T
y
pes of PAPR
Clipping,
SL
M
,
PT
S
In t
h
i
s
si
m
u
l
a
t
i
on
, we
are
usi
ng
3
2
num
ber
of sam
p
l
e
s,
w
h
ere we vari
e
d
num
ber of
su
bcar
ri
ers,
ba
sed
o
n
t
h
e
ove
rl
ap
pi
n
g
fa
ct
or,
α
. Th
is al
lo
wed
m
o
re inform
at
io
n
t
o
be tran
sm
itted
in
th
e system
with
t
h
e sam
e
num
ber
of sam
p
l
e
s
. C
l
i
ppi
n
g
, S
L
M
and
PTS
of P
A
PR
red
u
c
t
i
on t
ech
ni
q
u
e
i
s
appl
i
e
d i
n
t
h
i
s
GI
DFT
-
n
O
F
D
M
sy
stem
with va
riety
of
α
to
in
v
e
stig
ate th
e relatio
n
s
h
i
p
o
f
PAPR
redu
ctio
n
with
th
e
ove
rl
ap
pi
n
g
fa
ct
or.
Fro
m
Fig
u
re
4, it sho
w
s th
e
resu
lt
o
f
GIDFT n-
OFDM
wi
th
ou
t PAPR an
d with im
p
l
e
m
en
tatio
n
of
t
h
ree
t
y
pes
o
f
P
A
PR
re
duct
i
on
t
ech
ni
q
u
es
.
It
s
h
o
w
s
t
h
a
t
at
10
, C
l
i
p
pi
n
g
ha
s t
h
e
w
o
rs
t
pe
rf
orm
a
nce
com
p
ared to the ori
g
inal
GIDFT n-OFDM
perform
a
nce as
the overla
ppi
ng fact
or i
n
crea
ses. E
v
en though it is
a sim
p
l
e
t
echn
i
que,
i
t
cann
o
t
be
use i
n
t
h
i
s
sy
st
em
. B
e
si
des t
h
at
, PT
S al
so s
h
o
w
s a
p
r
om
i
s
i
ng per
f
o
r
m
a
nce
whi
c
h
6 dB
P
A
PR
re
d
u
ct
i
o
n
. SLM
t
ech
ni
q
u
e s
h
o
w
s i
m
pr
ovem
e
nt
res
u
l
t
com
p
ared t
o
ot
he
rs
whi
c
h i
s
8
dB
PAPR redu
ction
s
. After
th
at, a
si
m
u
latio
n
w
ith
m
a
n
y
ch
o
ices of
ov
er
lapp
in
g f
actor
α
wa
s d
one
in
GI
D
F
T n
-
OFDM with SLM technique. In Figure
5, as expect
ed, it
sh
owed
the resu
lt
of SL
M decreased
as the
ove
rl
ap
pi
n
g
f
act
ors
α
i
n
c
r
e
a
sed.
T
h
e
ov
erl
a
p
p
i
n
g
fact
ors
c
o
r
r
elated with sign
ifican
t PAPR
red
u
c
tion
perform
a
nces. It also shows that
α
=0.
2
has
gi
ve
n t
h
e m
o
st
red
u
ct
i
on
PA
PR
val
u
es
wi
t
h
re
duct
i
o
n
of
9 dB
.
This
happe
n
ed beca
use, at
α
=0.
2
, t
h
e
ove
r
l
appi
n
g
fact
o
r
of
t
h
e
fre
qu
en
cy is 80
% wh
er
e
the
num
b
er of
subcarriers
are 160 a
n
d s
u
bcarrier,
is fixed at 32. He
nc
e, the
bigg
er
t
h
e d
a
ta
in
pu
t, th
e
ov
erlap
p
i
ng
will in
crease an
d th
e
frequ
ency sep
a
ration
∆
is decrea
sed.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
502
-47
52
I
ndo
n
e
sian
J Elec Eng
& Com
p
Sci, V
o
l. 10
,
No
.
3
,
Jun
e
2
018
:
10
45
–
1
052
1
050
Fi
gu
re
4.
P
A
P
R
per
f
o
rm
ance usi
n
g
C
l
i
ppi
n
g
,
PTS
an
d
SL
M
PAPR
re
duc
t
i
on t
ech
ni
ques
Fi
gu
re
5.
P
A
P
R
per
f
o
rm
ance o
f
SLM
t
e
c
h
ni
que
s wi
t
h
di
ffe
rent
val
u
es
o
f
α
.
4.
CO
NCL
USI
O
N
In
t
h
is p
a
p
e
r, t
h
e Peak
to Av
erag
e
Power Ratio
(P
APR) in
GIDFT
n-OFDM syste
m
was investigate
d
and
anal
y
s
ed
.
Sim
u
l
a
t
i
ons o
f
GI
DFT
n
-
O
F
DM
sy
st
em
w
i
t
hout
a
n
y
P
A
PR
red
u
ct
i
o
n
as wel
l
as wi
t
h
t
h
ree
PAPR redu
ctio
n
techn
i
qu
es
were p
e
rform
to
i
m
p
r
o
v
e th
e PAPR v
alu
es. Th
e resu
lts o
f
th
e lowest PAPR
val
u
es f
o
r al
l
sim
u
l
a
t
i
on were
recor
d
e
d
an
d
anal
y
s
ed usi
n
g
C
C
D
F. Si
m
u
lat
i
on res
u
l
t
s
sh
ow
, SLM
has
gi
ve
n
t
h
e best
PA
PR
val
u
e o
f
com
p
are
d
t
o
ot
her
PAPR
re
d
u
ct
i
o
n t
echni
que
s t
h
at
has bee
n
p
e
rf
orm
i
n
t
h
i
s
pape
r
.
B
e
si
des t
h
at
,
we ca
n c
o
ncl
ude
t
h
at
t
h
e l
e
vel
of
PA
PR
re
duct
i
o
n
va
ri
es wi
t
h
t
h
e
l
e
vel
o
f
fre
q
u
e
ncy
ove
rlapping i
n
dicated as
α
. As
th
e ov
er
lapp
i
n
g
fr
equ
e
n
c
y increase
,
the
PAPR
reducti
on also
decrea
se i
n
great
v
a
lu
es. In
additio
n
,
even
t
h
ou
gh
th
e
ov
erlap
p
i
n
g
p
e
r
cen
tag
e
is in
crease,
o
n
l
y PTS and
SLM still
m
a
n
a
g
e
t
o
redu
ce th
e
PAPR d
e
sp
ite o
f
l
a
rg
e i
n
pu
t d
a
ta. Th
erefo
r
e,
thi
s
shows t
h
at,
GIDFT
n-
OFDM syste
m
efficiency
can
b
e
im
p
r
o
v
ed
b
y
u
s
ing
SLM an
d
PTS t
ech
n
i
q
u
e
s. In
t
h
e fu
ture
work, th
e sim
u
latio
n
and
an
alysis will be
extended at the receiver pa
rt in or
der to investigate the Bit Error Rate
(BER) of the wh
ole GIDFT n-OFDM
sy
st
em
wi
t
h
re
duct
i
o
n t
e
c
h
ni
que
o
f
PAPR
.
REFERE
NC
ES
[
1
]
J.
G.
Andrews
et
al.
, “
W
hat
will 5G b
e
?
,
”
I
E
EE
J.
Sel. Areas Commun.
,
vol.
32,
no.
6,
pp.
1065–10
82,
201
4.
[2]
G.
W
under
et al.
, “
5
GNOW
:
Non-ort
hogo
nal, asy
n
chronous wavefor
m
s for future
m
obile
a
pplications,”
I
E
EE Commun.
Mag.
,
vol.
52,
no.
2,
pp.
97–1
05,
201
4.
[
3
]
S. Isa
m
and I.
Dar
wazeh,
“
P
recoded spectrally efficien
t
FDM
system,
”
IE
E
E
I
n
t. Sy
m
p
.
Per
s
.
I
ndoor
M
ob.
Radio Co
m
m
un.
PI
M
RC, pp.
99–10
4,
2010.
Evaluation Warning : The document was created with Spire.PDF for Python.
In
d
onesi
a
n
J
E
l
ec En
g &
C
o
m
p
Sci
ISS
N
:
2
5
0
2
-
47
52
PAPR Re
duction Tec
hni
que
s
in Ge
ner
a
lized Inverse
Discre
te Fourier… (Iram M
a
isar
ah
Mok
h
tar)
1
051
[4]
W
.
Kozek and A. F. Molisch, “
N
on
orthogo
nal pulses
h
apes fo
r m
u
ltic
arri
er co
m
m
unication
s
in d
oubly
dis
p
ersive channels,”
I
E
EE
J.
Sel.
A
r
eas
Commun.
,
vol.
16,
no.
8,
pp.
1579–
15
89,
1998.
[5]
A.
M.
E
l
shir
kasi,
M
.
U.
Siddiqi,
and M
.
H.
Habaebi,
“
G
en
era
lized
d
iscrete fo
u
r
ier tra
n
s
fo
rm b
a
s
ed
min
imiza
tio
n
o
f PAPR in
OFDM systems,
” Pr
oc. -
5th I
n
t. Co
nf. Com
put. Co
m
m
un.
Eng. Em
er
g. Technol. via Co
m
p
-
U
nica
tion Con
v
er
g. I
CCCE 2014,
vol.
6,
pp.
205–2
0
8,
2015.
[
6
]
T. Xu and I. Da
rw
azeh,
“Spectrally Efficient FDM
:
Spect
rum Saving Techni
que for 5G?,”
Pr
oc.
1st I
n
t.
Conf
.
5G Ubiquitous
Connect.
,
2014.
[
7
]
Z. Sharif
ian, M. J.
O
m
idi, A. Farhan
g, and H. Saeedi-s
ourck, “
P
olyno
m
ia
l-Based Co
m
p
ress
ing and Iterative Expanding for
PAPR Reduction in GFDM,
”
pp.
518–523,
20
15.
[
8
]
B. Saltzberg, “
Perf
o
r
m
ance of
an Ef
f
icient Parall
el Dat
a
Trans
m
ission Syste
m
,”
Commun.
Technol.
I
E
EE
Trans.
,
vol.
15,
no.
6,
pp.
805–8
11,
19
67.
[9]
P. A
m
ini, R. Chen,
B. Fa
rhang-borouj
eny, and S. Me
m
b
er, “
F
ilterbank Multica
rri
er Co
mm
un
ications for Underwater Acousti
c
Channels,
”
vol.
40
,
no.
1,
pp.
115–13
0,
2015.
[10]
S.
N.
Pr
e
m
nath,
D.
Wasden,
S.
K. Kaser
a
,
N.
Patwar
i,
and B.
Far
h
ang-
Bor
oujeny
,
“B
ey
ond OFDM
: Best-
e
ffor
t
dy
nam
i
c
spectru
m
ac
cess using filterbank
m
u
lt
icarrie
r
,”
IE
EE/
AC
M Tra
n
s
. Netw.
,
v
ol.
21,
no.
3,
pp.
869–8
82,
201
3.
[11]
G. Fettweis,
M. Krondor
f, and
S. Bitt
ner, “
GFDM - Ge
neralized
frequenc
y
division m
u
ltiple
xing,”
IEEE Veh.
Technol. Conf.
,
pp.
1–4,
200
9.
[12]
J.
W
a
ng,
X.
Yang, X.
H.
Z
h
ang,
and D.
L
i
,
“The prefix
design an
d perfor
mance analys
is of
DF
T-
based overlapped frequency
division multiplexing (
O
vFDM
-DF
T
) system,
” IW
SDA
07 -
3r
d I
n
t.
W
o
r
k
.
Signal Des. I
t
s
Appl.
Co
m
m
un.
,
no. I
c
, pp.
361–
364,
200
7.
[13]
N.
Ah
m
a
d
,
S.
K.
Sy
ed-
Y
usof,
N.
Fis
a
l,
K.
Anwar
,
and T
.
Matsu
m
oto,
“Soft-
feedback MMSE
equalization fo
r
non-
orth
ogo
nal
frequency division
multiplexing
(
n
-O
FDM) signal detection,”
2012
16th I
n
t.
I
T
G
W
o
r
k
.
S
m
ar
t Antennas,
W
S
A
2012
, pp
. 2
4
8
–
255,
201
2.
[14]
S. S.
K. C.
Bulusu, H.
Shaiek, D.
Roviras, R.
Zayani, a
nd C. Laetitia,
“
R
e
duction of PAPR for FBMC-OQA
M
Syste
m
s using
Disper
sive SLM
Technique,
”
vol.
1,
no.
2,
pp.
568–5
72
.
[
15]
S. Isa
m
and I. Da
r
w
azeh,
“Peak to average power
rati
o reduction in
spe
c
trally efficient FDM systems,”
2011 18t
h I
nt.
Co
nf.
T
e
leco
m
m
un
.
I
C
T
2011,
no.
2,
pp.
36
3–36
8,
2011.
[16]
P. Malathi and P.
T. Vanathi,
“Imp
r
o
ved
in
terlea
ving
tech
n
i
q
u
e for PAPR red
u
ctio
n
i
n
OFDM-MIM
O system
,
” Pr
oc.
-
2nd
Asia I
n
t.
Conf.
M
o
del.
Sim
u
lation,
AM
S 2008,
pp.
253
–258,
20
08.
[17]
M. G. Su
m
ithra an
d M. Saru
m
a
thi,
“
C
omparative anal
ysis of PAPR reduction
techniques for multicarrier syst
ems,”
2012 I
n
t.
Conf.
Co
m
put.
Com
m
u
n
.
I
n
form
atics
,
pp.
1–5,
2012.
[18]
Y.
Rah
m
atallah and S.
M
ohan,
“Pe
a
k-
T
o
-
A
ver
a
ge Po
wer
Ratio Reduction in OFDM
Sy
stem
s
: A Sur
v
ey
And T
a
xonom
y
,
”
vol.
15,
no.
4,
pp.
1567
–15
92,
201
3.
[19]
K.
M
h
atr
e
and U. P.
Khot,
“E
fficient
Selective M
a
pp
ing PAPR Reduction T
echnique,
”
P
rocedia Comput.
S
c
i.
,
vol.
45,
pp.
620–
627,
20
15.
[20]
H.
F.
M
a
hdi,
M.
I
s
m
a
il,
and R.
Nor
d
in,
“Co
m
par
i
son of
Channel State Inf
o
r
m
ation Esti
m
a
tion Using SL
M
and Clipping-
based
PAPR Reduction M
e
thods,
”
P
r
ocedia Technol.
,
vol.
11,
no.
I
c
eei,
pp.
955–961,
20
13.
[21]
S.
Ham
e
ed L
,
“A
Co
mb
in
ed
Weig
h
t
i
n
g
an
d
PTS Tech
n
i
q
u
e for PAPR R
ed
u
ctio
n
i
n
OFDM S
i
g
n
a
l
s,”
2
nd
I
n
t.
Conf.
C
u
r
r
.
T
r
ends E
ng.
Technol.
I
CCTE
T
’
14,
pp
.
228–23
2,
2014.
BIOGRAP
HI
ES OF
AUTH
ORS
Iram Maisarah Mokhtar
recei
ved a diplom
a in Electr
i
c Com
m
unication Eng
i
neer
ing from
Universiti T
e
kn
ologi Malay
s
i
a
i
n
2011, B. Eng (
H
orns) in Electri
cal
Tel
ecom
m
unication Degree
from
Universiti Teknologi Malay
s
ia too
in 2014. She is
currentl
y
a Master student
b
y
research in
Universiti T
e
kn
ologi Mala
ysi
a
a
nd current rese
ar
ch inter
e
sts incl
ude wireless co
m
m
unication in
in future com
m
unica
tion which i
s
5G, m
u
lticarri
er s
y
s
tem
, P
A
P
R
anal
ys
is
, and
P
A
P
R
reduction
techn
i
ques.
Nor
u
lhus
na Ah
mad
graduat
e
d f
rom
Universiti
T
e
knologi M
a
lay
s
ia (UTM)
in 200
1 with BSc
in
Ele
c
tri
cal Engi
neering
.
She received her Master
degr
ee
of Electrical Engi
neeri
ng
(Tel
ecom
m
uni
cat
i
on) a
nd
Ph
D i
n
El
ect
ri
cal
Engi
nee
r
i
n
g i
n
2
0
0
3
and 2
0
1
4
,
respectively.
Currently, she i
s
a lecturer at Razak School
of Scienc
e and
Adva
nce
d
Tech
nol
ogy
,
U
T
M
KL.
He
r
expe
rt
i
s
e i
s
o
n
t
h
e area
o
f
di
gi
t
a
l
si
gnal
pr
ocessi
n
g
a
n
d
wireless c
o
mmunication.
He
r research i
n
tere
st
s are
on
f
u
t
u
re c
o
m
m
uni
cat
i
on s
u
c
h
as
5
G
an
d
co
gn
itiv
e rad
i
o
fo
cu
sin
g
o
n
erro
r co
rr
ectin
g
co
d
e
s,
tu
rb
o
eq
u
a
lizatio
n
,
OFDM,
reso
u
r
ce al
l
o
ca
t
i
on,
net
w
o
r
k
c
odi
ng
an
d c
o
o
p
erat
i
v
e
com
m
uni
cat
i
o
n.
Haz
ilah Mad Kaidi
receiv
e
d the B. Eng (Horns)
in Elect
r
i
cal En
gineer
ing (teleo
mmunication),
Universiti
Tekn
ologi Mal
a
y
s
ia
(UTM – 2006), the M
.
Sc. d
e
gr
ee i
n
T
e
lecom
m
unication
and
Inform
ation Eng
i
neer
ing at Univ
ersiti T
e
knolog
i MARA (UiTM
– 2008), and th
e Ph.D. degree
from
the Universiti Teknolog
i Malay
s
ia (UTM – 2015)
. She i
s
currentl
y
a senior lectur
er at
UTM Razak School of Enginee
ring and Advanced
Te
chnolog
y
,
Universiti Tekn
ologi Mala
y
s
ia
Kuala Lum
pur.
Her res
ear
ch in
t
e
res
t
s
inc
l
ude m
obile
and wireless communications, error
control
coding, relay
n
e
tworks, coope
r
a
tiv
e communications, H
y
br
id
ARQ Cross Lay
e
r
Design
and
iter
ativ
e r
ece
iver
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
502
-47
52
I
ndo
n
e
sian
J Elec Eng
& Com
p
Sci, V
o
l. 10
,
No
.
3
,
Jun
e
2
018
:
10
45
–
1
052
1
052
M
o
hd A
z
r
i
M
o
hd
I
z
har
receiv
e
d his
M
.
Eng. degree in e
l
e
c
tri
c
al engin
eering (
c
om
m
unications
)
from the Univer
sity
of Sheffield, U.K.,
in 2008
where he r
e
ceiv
ed the Mappin
medal award
for
outstanding
ac
a
d
em
ic perfo
rm
ance
and t
h
e Ins
titut
e
of
El
ec
tri
cal
and
El
ectro
nics Eng
i
neers
(IEEE)
priz
e fo
r the b
e
st com
m
unications-re
la
ted fin
a
l
ye
ar
p
r
ojec
t. He
re
cei
ved the
P
h
.D.
degree in
electri
cal eng
i
neering f
r
om
Univers
iti Teknologi Malay
s
ia (UTM
), Malay
s
ia, in 2014.
He has been a senior lecturer with UTM Kuala Lumpur campus
since 2014.
His current research
inter
e
sts include channel
coding
, coding theor
y
,
joint source-
chan
nel coding
, cod
e
d modulation,
cooperative com
m
unications, co
gnitive rad
i
o, Io
T and
quan
t
um
com
m
unications.
Norliz
a Mohamed
received h
e
r PhD in El
ectrical
Eng
i
neeri
ng from
Universiti T
e
knolog
i
Malay
s
ia (UTM) in 2013
. She
is one
of th
e lecturer in
Universi
ti Tekno
logi Ma
lay
s
ia. Besides
that, Dr. Norliza is an Assist
ant
Secretar
y
for Persatuan Keba
jikan AnNajah Selangor (PKAS), a
Non-Government Organization
focusing on ch
arity
ac
ti
vi
ti
e
s.
He
r ma
i
n
re
se
arc
h
a
r
ea
i
s i
n
optic
al
com
m
u
nica
tions
in
clu
d
es
opti
cal
wi
re
le
ss c
o
mmu
ni
c
a
t
i
on,
Radio
over Fib
e
r
,
optoelectron
i
cs devices,
RF
mixer design,
and
frequency
con
v
ers
i
on. Dr Norliz
a als
o
has
expanded
her
res
earch
interest in
Agricultur
e
Technolog
y
.
Evaluation Warning : The document was created with Spire.PDF for Python.