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.
8
, No
.
6
,
Decem
ber
201
8
, p
p.
486
3
~
48
70
IS
S
N: 20
88
-
8708
,
DOI: 10
.11
591/
ijece
.
v8
i
6
.
pp
486
3
-
48
70
4863
Journ
al
h
om
e
page
:
http:
//
ia
es
core
.c
om/
journa
ls
/i
ndex.
ph
p/IJECE
Des
i
gning
a Novel High
P
er
f
orma
nce Four
-
to
-
T
wo
Compre
ssor
Cell Bas
ed on C
NTFET
Tec
hn
ology for Low
Vol
tages
Mehdi B
agher
iz
ad
eh
1
,
M
ona
Mora
di
2
,
Mo
s
tafa Tor
bi
3
1
,
3
Depa
rtment
of
Com
pute
r
Eng
i
nee
ring
,
R
afsa
nj
an
Bran
ch, I
slam
ic
Az
ad
Univ
ersi
t
y
,
Ir
an
2
Young Re
sea
r
c
her
and
E
li
t
e
Clu
b,
Roudeh
en
Br
a
nch,
Isl
amic
A
zad Unive
rsit
y
,
Ir
a
n
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
A
pr
15
, 201
8
Re
vised
Ju
l
2
9
,
201
8
Accepte
d
Aug
1
3
, 201
8
Com
pre
ss
or
ce
ll
is
ofte
n
p
la
c
e
d
in
criti
ca
l
p
a
th
of
m
ult
iplier
ci
rcu
it
s
to
per
form
par
tial
produc
t
sum
m
at
ion.
The
re
fore
,
i
t
play
s
a
sign
ifica
nt
role
in
det
ermining
the
ent
ir
e
p
erf
or
m
anc
e
of
m
ultipli
er
and
digital
s
y
stem
.
Respec
t
ing
to
t
he
ne
ce
ss
ity
of
low
power
des
ign
for
por
ta
bl
e
elec
tron
ic,
designi
ng
a
low
power
and
hig
h
-
per
form
anc
e
c
om
pre
ss
ors
see
m
s
to
be
a
good
soluti
on
to
over
come
of
th
ese
proble
m
s
for
computat
ions.
I
n
thi
s
pape
r
a
novel
high
p
erf
orm
anc
e
four
-
to
-
two
compre
ss
or
c
el
l
is
pro
posed
using
Carbon
Nanotub
e
Fiel
d
Eff
ec
t
T
ran
sistors
(CNTFETs)
te
chnol
og
y
.
The
new
ce
l
l
is
base
d
on
Majori
t
y
Fun
ct
ion
,
NO
R,
and
NAND
gat
es.
The
m
ai
n
adva
nt
age
of
proposed
design
in
compari
son
wi
th
form
er
ce
ll
s
i
s
the
ea
se
of
obta
ini
ng
CA
RRY
ou
tpu
t
b
y
m
e
ans
of
Majori
t
y
func
ti
on.
Sim
ula
ti
ons
hav
e
bee
n
done
with
32nm
te
chnol
og
y
node
using
S
y
nops
y
s
HS
PI
CE
software
.
Sim
ula
ti
on
resul
ts c
onfirm t
h
e
pr
iori
t
y
of
the prop
osed
cell compared
to
o
ther
stat
e
-
of
-
the
-
art
f
our
-
to
-
two
comp
ressor c
ells.
Ke
yw
or
d:
Ca
rbon
n
a
no
t
ub
e
f
ie
ld
e
ffec
t
t
ran
sist
or
(CN
TFET)
Four
-
to
-
t
w
o
c
om
pr
essor
High
s
pee
d
Lo
w
p
ower
Nano
el
ect
r
onic
Nanotech
nolo
gy
Copyright
©
201
8
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
:
Me
hd
i B
a
gh
e
ri
zadeh
,
Dep
a
rtm
ent o
f C
om
pu
te
r
E
ng
i
neer
i
ng,
Ra
fsanja
n
Bra
nch,
Islam
ic
A
zad
Un
i
ver
sit
y,
Ra
fsanja
n,
Ir
a
n
.
E
m
a
il
:
m
.b
agheriza
de
h@
s
rb
i
au.
ac.i
r
1.
INTROD
U
CTION
To
day,
in
VL
SI
syst
e
m
s
fast
arit
h
m
et
ic
com
pu
ta
ti
on
struc
tures
su
c
h
as
m
ul
ti
pliers
and
add
er
s
are
the
m
os
t
fr
equentl
y
util
iz
ed
c
ircuit
s
[
1
]
,
[
2
]
.
Mult
ipli
ers
are
the
m
os
t
sign
ific
ant
pa
rts
of
arit
hm
etic
c
ircuit
s
from
in
te
r
m
s
of
perform
ance
and
po
wer.
Digital
sign
al
process
or
s
a
nd
m
ic
ro
process
ors
rely
on
the
e
ff
ect
ive
i
m
ple
m
entat
io
n
of
floati
ng
point
un
it
s
a
nd
com
m
on
arit
hme
ti
c
log
ic
unit
s
to
perf
or
m
ded
ic
at
e
d
al
go
rithm
s
su
c
h
as
filt
erin
g
a
nd
co
nvol
ut
ion
.
I
n
m
any
of
these
ap
plica
ti
on
s
,
m
ulti
pliers
are
c
onside
re
d
as
the
c
riti
cal
pa
rt
dicta
ti
ng
the
overall
ci
rcu
it
pe
rfor
m
ance
wh
en
c
on
st
raine
d
by
com
pu
ta
ti
on
sp
ee
d
an
d
powe
r
consum
ption.
Mult
ipli
ers
co
m
m
on
ly
inclu
de
t
hr
ee
s
ub
functi
ons:
1)
par
ti
al
produ
ct
gen
erati
on;
2)
pa
rtia
l
product
reducti
on;
3)
f
inal
add
it
io
n
with
carry
pro
pag
at
in
g
[
3
]
.
Com
pr
esso
r
cel
ls
are
gen
erall
y
e
m
plo
ye
d
in
m
ulti
-
op
e
ra
nd
ad
ders
to
reduce
th
e
num
ber
of
opera
nds
a
nd
i
n
m
ulti
pliers.
These
cel
ls
a
r
e
us
e
d
t
o
redu
ce
the
nu
m
ber
of
pa
r
ti
al
pr
oducts.
I
n
the
desi
gn
of
the
m
ult
ipli
er
unit
s
the
fou
r
-
to
-
tw
o
com
pr
esso
rs
are
t
he
m
os
t
wide use
d
m
odules
[
4
]
-
[
6
]
.
The
ne
eds
of
scal
ing
down
t
he
siz
e
o
f
tran
sist
or
in
Nano
ranges
in
c
urren
t
MO
SFET
te
chnolo
gy
le
ads
to
s
om
e
chall
enges
s
uc
h
as
reli
abili
ty
,
powe
r
c
ons
um
pt
ion
,
le
ss
c
ontr
ol
of
t
he
gat
e,
le
aka
ge
pow
er
a
nd
high
li
thogra
phy
costs
[
7
]
-
[
9
]
.
He
nce,
t
o
ov
e
rc
om
e
th
ese
dif
ficult
ie
s
of
na
noscal
e
MOSFE
Ts,
ne
w
te
chnolo
gies
s
uch
as
Si
ng
le
-
Ele
ct
ron
Tra
ns
ist
or
(S
ET
),
Qu
a
ntu
m
-
Do
t
Ce
ll
ular
Au
t
om
at
a,
and
Ca
rbo
n
Nanotu
be
Fiel
d
E
ff
ect
T
ra
nsi
stor
(CNT
F
ET)
ha
ve
be
e
n
stu
died
in
m
any
li
te
ratur
es
[
8
]
.
Am
ong
these
te
chnolo
gies
CNTFE
T
seem
s
to
be
the
m
os
t
feasible
prom
isi
ng
s
uccess
or
du
e
t
o
it
s
re
m
a
rk
a
ble
featu
res
an
d
it
s
si
m
il
arities
with
M
OS
F
E
T
te
ch
no
l
og
y
[
8
]
,
[
10
]
.
C
NTFET
te
ch
no
l
og
y
be
nef
it
f
ro
m
balli
sti
c
op
er
at
ion
a
nd
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.
8
, N
o.
6
,
Dece
m
ber
2
01
8
:
486
3
-
48
70
4864
on
e
dim
ensional
band
str
uctu
re,
a
nd
l
ow
O
F
F
-
c
urren
t
a
nd
t
he
sam
e
m
ob
il
it
y
fo
r
bo
t
h
n
-
t
ype
CNTF
ET
and
p
-
ty
pe
CNTFET
.
This
featu
re
facil
it
ie
s
the
transistor
siz
ing
of
com
plex
ci
rcu
it
s
ea
sie
r
[
11
]
.
The
refor
e
,
CNTFET
ca
n
be
us
e
d
for
de
sign
i
ng
ene
r
gy
ef
fici
ent
inte
grat
ed
ci
rcu
it
s.
The
diam
et
er
of
the
CNT
rel
ie
s
on
the
chiral
ve
ct
or.
The
C
hiral
vecto
r
in
dicat
es
the
el
ect
ronic
char
act
erist
ic
s
and
a
rr
a
nge
m
ent
ang
le
of
carbo
n
at
om
s
a
lon
g
CNTs.
It
is
de
te
rm
ined
by
(n1,
n2)
integ
er
pair
.
If
12
3
(
)
n
n
i
i
Z
,
CNT
is
m
et
allic
,
oth
e
rw
ise
is
se
m
ic
on
duct
or
[
12
]
. T
he diam
eter
of
na
no
t
ub
e
s in n
a
no
m
et
er is co
m
pu
te
d us
ing
t
he
e
qu
at
i
on
(1):
22
1
2
1
2
0
.
0
7
8
3
C
N
T
D
n
n
n
n
(1)
It
worth
to
m
e
ntion
that
the
desire
d
thres
hold
volt
age
of
the
CNTFET
based
tra
ns
ist
or
s
c
ou
l
d
be
adjuste
d based
on the
diam
et
e
r of
na
no
t
ub
e
s
wh
ic
h
a
re lo
ca
t
ed unde
r
the
tr
ansisto
r gate
(
e
qu
at
io
n
2)
[
12
]
.
3
0
.
4
3
23
g
t
C
N
T
C
N
T
E
E
V
e
e
D
D
(2)
Wh
e
re,
(
0
.
2
4
9
)
nm
is
the
CNT
la
tt
ic
e
con
sta
nt,
E
is
the
energy
of
car
bon
band
i
n
the
ti
ght
bondin
g
m
od
el
, e
is t
he
el
ect
r
on ch
a
r
ge,
a
nd
C
N
T
D
is t
he diam
e
te
r
of n
a
notu
bes.
.
Com
pr
essor
cel
ls
are
gen
erall
y
em
plo
ye
d
in
m
ulti
-
op
eran
d
add
ers
to
red
uce
the
nu
m
ber
of
op
eran
ds
and
in
m
ulti
pliers.
The
fo
ur
-
to
-
two
com
pr
essor
com
pr
esses
five
par
ti
al
pr
od
ucts
bits
into
three.
This
com
pr
essor
has
fo
ur
inp
uts
cal
le
d
X1;
X2;
X
3
and
X4
and
two
ou
tpu
ts,
SUM
and
CARRY
al
on
g
with
a
CARRY
-
IN
(
Ci
n
)
and
a
CARRY
-
OU
T
(
Cou
t)
.
The
in
pu
t
Ci
n
is
the
ou
tpu
t
fr
om
the
pr
eviou
s
casca
ded
com
pr
essor
. Th
e Cou
t i
s the o
utp
ut to
the co
m
pr
essor
in
the n
ext stage.
In
the li
te
ratur
e
m
any
fo
ur
-
to
-
two
com
pr
essor
cel
ls hav
e
been
p
resen
te
d
pr
eviou
sly
[
13
]
-
[
15
]
. F
igu
re
1
sh
ow
s
the
structur
e
of
the
fo
ur
-
to
-
two
com
pr
essor
cel
l
pr
op
os
ed
in
[
13
]
.
The
m
ai
n
idea
of
this
design
or
iginate
d
fr
om
truth
ta
ble
of
the
fo
ur
-
to
-
two
com
pr
essor
.
This
structur
e
is
com
po
sed
of
a
sing
le
arr
ay
of
capaci
tors
and
CNTFET
-
based
netwo
rk
.
In
this
design
C
NTF
ET
with
diff
eren
t
thresh
old
vo
lt
age
are
us
ed
[1
1].
Fig
ure
1. The
Four
-
to
-
Tw
o
c
om
pr
esso
r
str
uc
ture
i
n
[
10]
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
Desig
ning
a
N
ovel Hi
gh Perf
or
m
ance F
our
-
to
-
Tw
o
C
ompr
essor
Ce
ll
B
ase
d on
...
(
Meh
di
Bagh
eri
z
adeh
)
4865
Figure
2,
e
xh
i
bits
the
str
uctu
re
of
the
f
our
-
to
-
tw
o
c
om
pr
e
sso
r
cel
l
discu
ssed
i
n
[
14
]
.
It
co
ns
ist
s
of
seve
nty
transis
tors.
It
has
ab
ou
t
th
ree
2
-
wa
y
XO
R
crit
ic
al
path
delay
.
This
com
pr
ess
or
is
base
d
on
ne
w
descr
i
ption o
f
l
og
ic
al
e
quat
ions
that
descr
i
be t
he
co
rr
e
spo
nding
f
un
ct
io
nalit
y [12].
F
i
g
ure
2.
T
he
F
ou
r
-
to
-
T
w
o
c
om
pr
e
s
s
or
s
t
r
u
c
t
ur
e
i
n
[
14
]
F
i
gu
r
e
3
de
m
ons
t
r
a
t
e
s
t
he
s
t
ruc
t
ur
e
of
f
ou
r
-
to
-
t
w
o
c
om
pr
e
s
s
or
c
e
l
l
pr
op
o
s
e
d
i
n
[
15
]
.
I
t
c
on
s
i
s
t
s
of
s
e
ve
nt
y
-
t
w
o
t
r
a
ns
i
s
t
or
s
.
T
he
c
r
i
t
i
c
al
pa
t
h
de
l
a
y
of
t
hi
s
c
omp
r
e
s
s
or
e
qu
a
l
s
t
o
t
he
de
l
a
y
s
of
"
on
e
‘
N
A
N
D
'
,
t
w
o
‘2
-
w
a
y
X
O
R
s
’
,
on
e
‘
2
-
w
a
y
X
N
O
R
’
"
ga
t
e
s
[
13
]
.
F
i
g
ur
e
3.
T
he
F
ou
r
-
to
-
T
w
o
c
om
pr
e
s
s
or
s
t
r
u
c
t
ur
e
i
n
[
15
]
In
t
his p
a
pe
r
a
new
f
our
-
to
-
t
w
o
c
om
pr
esso
r
c
el
l based
on M
ajorit
y Fu
nctio
n,
NO
R
, and N
AND gat
es
is
pr
ese
nted
.
T
he
m
ai
n
adv
a
nt
age
of
pr
opose
d
de
sig
n
in
c
om
par
ison
wit
h
pre
vious
cel
ls
i
s
that
the
C
OU
T
1
(CARRY
) ou
t
put o
btaine
d usi
ng Maj
or
it
y f
unct
ion
.
2.
RESEA
R
CH MET
HO
D
In
t
his
sect
ion
a
ne
w
desi
gn
for
a
high
pe
rfor
m
ance
f
our
-
to
-
tw
o
com
pr
ess
or
is
pre
sented
.
T
he
pro
po
se
d
f
our
-
to
-
tw
o
com
pr
e
sso
r
cel
l
is
bas
ed
on
the
eq
ua
ti
on
s
(3
)
,
(4)
,
and
(
5).
T
he
Bl
ock
dia
gr
am
of
the
pro
po
se
d desi
gn is s
how
n
i
n
F
igure
4.
X
1
X
2
X
3
X
4
C
i
n
S
U
M
M
a
j
o
r
i
t
y
F
u
n
c
t
i
o
n
C
O
U
T
1
2
-
i
n
p
u
t
X
O
R
F
u
l
l
A
d
d
e
r
2
-
i
n
p
u
t
X
O
R
X
4
C
i
n
C
O
U
T
2
Fig
ure
4. The
blo
c
k diag
ram
of prop
os
e
d
f
our
-
to
-
tw
o
c
ompress
or
To
obta
in
S
U
M
ou
tp
ut,
one
fu
ll
ad
der
an
d
two
2
-
in
put
X
OR
gates
are
e
m
plo
ye
d.
As
m
entioned
t
he
perform
ance
of
the
pro
po
sed
fou
r
-
to
-
t
wo
com
pr
ess
or
cel
l
de
pe
nds
on
the
m
et
hod
t
hat
use
s
the
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.
8
, N
o.
6
,
Dece
m
ber
2
01
8
:
486
3
-
48
70
4866
i
m
ple
m
entat
io
n
of
the
f
ull
ad
der,
X
OR,
a
nd
Ma
j
ori
ty
fu
nct
ion
blo
c
ks
.
T
o
increase
the
e
f
fici
ency
of
pro
po
s
e
d
cel
l,
fu
ll
add
e
r
wh
ic
h
is
desig
ned
in
[
16
]
has
been
use
d.
T
o
desig
n
this
f
ull
add
e
r,
the
uni
qu
e p
r
operti
es
of
the
CNTFET
tran
s
ist
or
s
a
re
util
iz
ed.
In
orde
r
to
pro
du
ce
t
he
outpu
ts
of
Ma
jori
ty
-
no
t,
N
AND,
an
d
N
OR
m
odules,
three
-
i
nput ca
pa
ci
tors netw
orks
a
nd in
ver
te
r
s w
it
h dif
fer
e
nt
thr
es
holds
are use
d.
1
2
3
4
5
S
U
M
X
X
X
X
X
(
3)
1
[
(
1
2
3
)
,
4
,
)
C
O
U
T
M
a
j
o
r
i
t
y
X
X
X
X
C
i
n
(4)
2
(
1
,
2
,
3
)
C
O
U
T
M
a
j
o
r
i
t
y
X
X
X
(5)
To
pro
duce
the
2
-
in
put
XO
R
f
un
ct
io
n,
6
tran
sist
or
s
in
sym
m
et
ric
m
ann
er ar
e
us
e
d.
Fig
ur
e
5
disp
la
ys
the 2
-
in
puts X
OR gat
e that e
m
plo
ys C
in and
X4 sig
nals as
inp
ut si
gn
al
s t
o gen
e
rate t
he
SU
M si
gn
al
.
Fig
ure
5. 1s
t
pr
opos
e
d
2
-
i
nput
s XOR
gate
Figure
6
s
hows
the
2
-
in
puts
X
OR
gate
t
hat
hires
S
(e
qu
at
io
n
6)
an
d
Ci
n
⨁
X4
si
gnal
s
as
in
pu
t
sig
nal
s
to g
e
ne
rate t
he SUM si
gn
al
.
S
=
X1
⨁
X2
⨁
X3
(6)
Fig
ure
6. 2n
d
pro
posed
2
-
in
pu
ts XO
R
g
at
e
The
C
ou
t
2
outpu
t
is
base
d
on
Ma
jority
of
X
1,
X
2,
a
nd
X3
inputs.
T
his
si
gn
al
is
em
plo
ye
d
as
one
of
the
el
e
m
ents
req
ui
red
to
pro
du
ce
t
he
SU
M
sign
al
.
T
his
m
et
ho
d
m
akes
the
propose
d
four
-
to
-
tw
o
str
uctu
re
m
or
e eff
ic
ie
nt.
In
t
he
pro
po
se
d
desig
n,
unli
ke
the
pr
e
vious
four
-
to
-
tw
o
c
om
pr
essor
st
ru
c
tures,
the
C
ou
t
1
ou
t
pu
t
i
s
ob
ta
ine
d
easi
l
y
us
ing
one
Ma
j
ori
ty
fu
nct
ion
.
T
he
tra
nsi
stor
le
vel
im
plem
entat
ion
of
the
pro
pose
d
cel
l
is
sh
ow
n
in
Fi
gu
re
7.
As
it
is
s
how
n
the
S
no
de
an
d
on
e
3
-
inputs
XO
R
functi
on,
are
us
e
d
f
or
both
S
U
M
and
Cou
t
1 ou
t
pu
ts.
It m
akes th
e
propose
d desig
n m
or
e eff
ect
ive.
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
Desig
ning
a
N
ovel Hi
gh Perf
or
m
ance F
our
-
to
-
Tw
o
C
ompr
essor
Ce
ll
B
ase
d on
...
(
Meh
di
Bagh
eri
z
adeh
)
4867
Fig
ure
7. The
s
chem
at
ic
o
f
th
e pro
posed
fo
ur
-
to
-
tw
o
c
om
pr
essor
3.
RESU
LT
S
A
ND AN
ALYSIS
To
pe
rfor
m
si
m
ula
ti
on
s,
the
Synopsys
HSPICE
too
l
ha
s
with
the
32nm
Co
m
pact
S
PI
CE
m
od
el
pr
ese
nted
i
n
[
17
]
,
[
18
]
has
be
en
us
e
d.
T
his
m
od
el
is
dev
el
op
e
d
f
or
M
OSFET
-
li
ke
C
NT
FETs
with
S
W
CNTs
as
their
cha
nnel
con
side
rs
non
-
ideal
pa
ra
m
et
er
su
ch
as
Parasit
ic
includi
ng
C
NT
[
19
]
,
Gate
resist
ances,
capaci
ta
nces,
Sour
ce/
Dr
ai
n,
and
Schot
tky
Ba
rr
ie
r
E
ff
ect
s
.
Sim
ulatio
n
ha
s
been
done
on
previ
ous
f
our
-
to
-
two
com
pr
esso
r
ce
ll
s
[
13
]
-
[
15
]
a
nd
t
he
res
ults
are
com
par
ed
with
the
pro
po
s
ed
desig
n.
The
de
la
y
of
ou
t
pu
t
sign
al
s
is
m
easur
e
d
fro
m
the
tim
e
that
inp
ut
sign
al
s
befo
re
the
buf
fers
rea
ch
to
50%
of
their
vo
lt
a
ge
le
vel
to
the
m
o
m
ent
that
outp
ut
sig
na
ls
reac
h
to
the
sam
e
le
vel.
Th
en
t
he
m
axi
m
um
delay
is
repor
te
d
as
the
de
la
y
of
the
ci
rcu
it
.
The
power
c
on
s
um
pt
ion
is
the
aver
a
ge
po
wer
c
on
s
u
m
ption
m
easur
e
d
du
rin
g
a
lon
g
tim
e
[
8
]
,
[
19
]
.
Finall
y
the
power
-
delay
pro
du
ct
(
PDP)
w
hich
c
om
pr
omi
ses
betwee
n
delay
and
power
c
on
s
um
ption
i
s
repor
te
d
as
a f
i
gure
of m
erit
. Tr
an
sist
or
s
are
adjust
ed
i
n
a
way that t
he
m
ini
m
u
m
PD
P c
ou
l
d be
ob
ta
i
ne
d.
Table
1
sho
ws
the
si
m
ulati
on
resu
lt
s
at
0.
65
V
supp
ly
vol
ta
ges
an
d
at
250M
Hz
op
e
rati
ng
fr
eq
ue
ncy
with
2.1f
F
out
pu
t
l
oad
capaci
ta
nce
in
r
oo
m
t
e
m
per
at
ure.
As
it
is
show
n
th
e
pro
posed
f
our
-
to
-
tw
o
c
om
pr
ess
or
structu
re
has
l
ow
delay
, low
powe
r
c
onsu
m
ption
,
and t
he be
st PDP i
n com
par
is
on w
it
h ot
her cel
ls.
Table
1.
T
he
sim
ula
ti
on
r
e
su
lt
s
at
0.65
V
a
nd
at
2
50 M
Hz, 2
.
1f
F
Desig
n
Delay
(E
-
10)
Po
wer
(E
-
7)
PDP
(E
-
1
5
J)
Desig
n
[
13
]
5
.11
2
1
7
.15
6
3
0
.36
5
8
3
Desig
n
[
14
]
5
5
.90
8
5
.76
9
9
3
.22
5
9
Desig
n
[
15
]
5
1
.32
7
5
.22
0
9
2
.67
9
7
Prop
o
sed
Desig
n
2
.49
1
0
4
.88
2
0
0
.12
1
6
1
Nowa
days,
VL
SI
ci
rc
uits
wit
h
the
capa
bili
ty
of
w
orkin
g
i
n
high
f
re
qu
e
nc
ie
s
are
re
quir
ed
[
19
]
.
All
desig
ns
are
si
m
ula
te
d
at
hig
her
f
re
qu
e
ncie
s
su
ch
as
10
0
MHz,
250M
Hz
and
50
0MHz.
Table
2
sho
ws
the
cel
l
perform
ance
ve
rsu
s
f
reque
nc
y
increm
ents.
It
can
be
re
al
iz
ed
that
th
e
pro
posed
struct
ur
e
has
t
he
best
perform
ance
am
on
g
al
l
com
p
ared
ci
rc
uits
at
high
f
reque
nc
y
op
er
at
ion.
Al
l
desig
ns
ha
ve
al
so
bee
n
sim
ulate
d
in
a
vast
range
of
te
m
per
at
ur
e
s
from
0
oC
to
100
oC
at
10
0MH
z
op
e
rati
ng
f
re
qu
e
ncy
an
d
2
.
1fF
loa
d
ca
pacit
or
t
o
exam
ine the im
m
un
it
y of
th
e circuit
s t
o
the
te
m
per
at
ure
noise
a
nd v
a
riat
ion
s
.
The
resu
lt
s
a
r
e
show
n
i
n
F
igure
8
T
he
r
esults
ex
hib
it
that
the
pro
po
s
ed
desi
gn
has
norm
al
functi
onal
it
y
i
n
sp
it
e
of
te
m
per
at
ur
e
var
ia
ti
ons.
T
he
pr
e
sent
ed
desig
n
al
so
su
r
passes
oth
e
r
designs
in
te
rm
s
of
perform
ance.
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.
8
, N
o.
6
,
Dece
m
ber
2
01
8
:
486
3
-
48
70
4868
Table
2.
T
he
sim
ula
ti
on
r
e
su
lt
s in
Dif
fer
e
nt
Op
e
rati
ng F
requen
ci
es
Desig
n
1
0
0
MHz
2
5
0
MHz
5
0
0
MHz
Delay(E
-
10)
Desig
n
[
13
]
5
.11
2
1
6
3
.68
9
8
1
.51
1
Desig
n
[
14
]
5
5
.90
8
5
.76
9
9
3
.22
5
9
Desig
n
[
15
]
5
1
.32
7
5
8
.40
1
7
4
.90
7
Prop
o
sed
Desig
n
2
.49
1
0
2
.84
4
7
3
3
.67
0
1
Av
erage Powe
r
Desig
n
[
13
]
7
.15
6
3
7
.23
4
4
7
.39
0
7
Desig
n
[
14
]
5
.76
9
9
5
.83
3
2
5
.99
1
4
Desig
n
[
15
]
5
.22
0
9
5
.2
842
5
.44
2
4
Prop
o
sed
Desig
n
4
.88
2
0
4
.93
3
2
5
.06
1
2
PDP(E
-
15)
Desig
n
[
13
]
0
.36
5
8
3
0
.42
1
0
0
.55
2
0
9
Desig
n
[
14
]
3
.22
5
8
3
.71
5
1
4
.88
3
6
Desig
n
[
15
]
2
.67
9
7
3
.08
6
0
4
.07
6
7
Prop
o
sed
Desig
n
0
.12
1
6
0
.14
0
3
0
.18
5
7
Figure
8. P
DP
ver
s
us
sup
ply Tem
per
at
ur
e
V
ariat
ion
s
The
sim
ulati
on
s
ha
ve
been
carried
out
at
0.5V,
0.65V,
0.85V,
1.1
V,
and
1.4
V
pow
er
s
upplies
at
100MHz
ope
ra
ti
ng
fr
e
qu
e
ncy
at
r
oom
te
m
per
at
ur
e ca
n
se
e i
n
Fi
g
ure
9.
Figure
9. P
DP
ver
s
us
sup
ply vo
lt
age
V
a
riat
ion
s
Diff
e
re
nt
capa
ci
tors
are
em
pl
oyed
as
the
ou
tpu
t
load
to
ex
a
m
ine
the
dr
iv
ing
po
wer
of
the
pro
posed
desig
n
m
or
e
preci
sel
y.
The
capaci
ty
of
outp
ut
load
is
co
nsi
der
e
d
as
1.5fF
,
2.1
FF,
2.8fF
,
3.6fF
an
d
4.5
fF
at
room
tem
per
atu
re
a
nd
100M
Hz
operati
ng
f
reque
ncy.
The
si
m
ulati
on
resu
lt
s
ar
e
dem
onstrat
ed
at
Figure
10
resp
ect
ively
.
0
1
2
3
4
5
6
7
8
0
10
20
30
40
50
60
70
80
90
1
0
0
PDP(E
-
15)
Temprature
Desig
n
[10
]
Desig
n
[11
]
Desig
n
[12
]
Pr
o
p
o
sed
Desig
n
0
0
.5
1
1
.5
2
2
.5
3
3
.5
4
4
.5
5
0
.5
0
.65
0
.85
1
.1
1
.4
PDP(E
-
15)
Desig
n
[10
]
Desig
n
[11
]
Desig
n
[12
]
Pr
o
p
o
sed
Desig
n
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
Desig
ning
a
N
ovel Hi
gh Perf
or
m
ance F
our
-
to
-
Tw
o
C
ompr
essor
Ce
ll
B
ase
d on
...
(
Meh
di
Bagh
eri
z
adeh
)
4869
Fig
ure
10. PD
P v
e
rs
us
L
oa
d Variat
io
ns
4.
CONCL
US
I
O
N
Com
pr
esso
r
ce
ll
is
on
e
of
th
e
m
os
t
i
m
po
rtant
ci
rcu
it
s
be
cause
of
it
s
ef
fect
on
the
m
ulti
plica
ti
on
process
wh
ic
h
is
fr
eq
ue
ntly
us
ed
in
the
di
gital
syst
e
m
.
In
t
his
pa
per
a
no
vel
four
-
to
-
tw
o
com
pr
essor
ce
ll
was
pr
ese
nted
us
in
g
Ca
pacit
or
ne
twork
an
d
Ma
jority
functi
on,
NA
N
D
an
d
N
OR
gates
is
presented
.
Em
pl
oying
Ma
j
ori
ty
fu
nct
ion
re
duced
it
s
delay
sign
ific
antly
.
The
pro
po
s
ed
cel
l
include
s
twenty
six
transist
or
s
a
nd
si
x
capaci
tors.
T
he
f
our
-
to
-
tw
o
com
pr
ess
or
cel
ls
was
sim
ulate
d
us
in
g
Synopsys
HSPICE
t
oo
l
with
32nm
CNTFET
C
ompact
SP
ICE
m
od
el
.
Sim
ulati
o
n
res
ults
co
nf
i
r
m
ed
the
pri
or
it
y
of
the
pro
pos
ed
cel
l
com
par
ed
to
oth
e
r
sta
te
-
of
-
the
-
a
rt fo
ur
-
to
-
t
wo co
m
pr
ess
or cel
ls.
ACKN
OWLE
DGE
MENT
This
pap
e
r
is
s
upporte
d
by m
entione
d
a
ff
il
ia
ti
on
s.
REFERE
NCE
S
[1]
M.
Baghe
ri
za
d
e
h,
et
a
l.
,
“
Design
an
AS
IP
for
edi
t
dista
n
ce
al
go
ri
thm
in
pat
t
ern
rec
ogni
ti
on,
”
20
14
22nd
Iranian
Confe
renc
e
on
E
le
c
tric
al
Engi
n
e
ering
(
ICEE
)
,
pp
.
971
-
975
,
2014
.
[2]
S.
Tha
kra
l
,
e
t
al.
,
“
Im
ple
m
ent
at
io
n
and
Anal
y
sis
o
f
Reve
rsible
log
i
c
Based
Arithme
ti
c
Log
ic
Unit
,
”
TEL
KOMNIKA
(
Tele
communic
ati
on
Computing
El
e
ct
ronics
and
Control)
,
vol. 14
,
pp
.
1292
-
1298
,
2016.
[3]
R.
S.
W
aters
an
d
E.
E.
Sw
artzl
a
nder
,
“
A
red
uc
e
d
complexi
t
y
W
al
l
ac
e
m
ult
iplier
red
uction,
”
IE
E
E
transacti
ons
o
n
Computers
,
vol.
59,
pp
.
1134
-
11
37,
2010
.
[4]
C.
H.
Chang
,
e
t
al.
,
“
Ultra
low
-
volt
ag
e
low
-
po
wer
CMO
S
4
-
2
and
5
-
2
compre
ss
or
s
for
fast
arithm
et
ic
ci
r
cui
ts,
”
IEE
E
Tr
ansacti
o
ns on
Circuits a
nd
Syste
ms
I:
R
e
gular P
apers
,
vo
l.
51
,
pp
.
1985
-
1
997,
2004
.
[5]
M.
G.
M
.
B
agheriz
ad
eh
and
M
.
Eshghi,
“A
n
AS
IP
for
Edi
t
Dist
a
nce
Al
gori
thm
i
n
Patt
ern
Rec
og
nit
ion,
”
The
22n
d
Iranian
Confe
re
nce
on
E
le
c
tric
a
l
Eng
ine
ering
(
ICEE
)
,
2014.
[6]
M.
Bhava
ni
,
e
t
a
l.
,
“
Del
a
y
Com
par
ison
for
16x16
Vedi
c
Mult
ipl
i
e
r
Us
ing
RCA
and
CLA,
”
In
t
ernati
onal
Journal
o
f
El
e
ct
rica
l
and
C
omputer
Engi
n
e
ering
(
IJE
CE)
,
v
ol.
6
,
p
p
.
1205
,
2
016.
[7]
S.
T
abr
izchi
,
et
al.
,
“
Design
a
novel
te
rn
ar
y
h
al
f
adde
r
and
m
ult
ipl
ie
r
bas
ed
on
ca
rbon
nano
-
tube
field
eff
ec
t
tra
nsistors (CNT
FETs),
”
Frontie
rs
,
vol. 1, 2016.
[8]
M.
B
aghe
r
izade
h,
e
t
al
.
,
“
Digit
a
l
counter
cell
d
e
sign
using
c
arb
o
n
nanot
ub
e
FET
s,
”
Journal
of
A
ppli
ed
Re
search
and
Technol
og
y
,
2017.
[9]
M.
S.
Ali,
“
Casca
ded
Ripp
le
C
ar
r
y
Adder
base
d
SR
CS
A
for
Eff
ic
ie
n
t
FIR
Filt
er
,
”
Indone
sian
Jo
urnal
of
Elec
tri
c
al
Engi
ne
ering
and
Computer
Sc
ie
n
ce
(
IJEECS)
,
vol
.
9
,
2018
.
[10]
Y.
S.
Mehra
bani a
nd
M.
Eshghi,
“
Noise
and
Proce
ss
Vari
at
ion To
le
ran
t,
Low
-
Pow
er,
High
-
Speed
,
and
Low
-
Ene
r
g
y
Full
Adders i
n
C
NF
ET
Technol
o
g
y
,
”
2016
.
[11]
A.
Panahi,
e
t
al
.
,
“
CNF
ET
-
base
d
appr
oximat
e
t
er
nar
y
add
ers
for
ene
rg
y
-
eff
icient
image
proc
essing
appl
i
cations,
”
Mic
ropr
oce
ss
ors
and
Mi
crosyste
ms
,
2016.
[12]
M.
Darvi
shi
and
M.
Baghe
ri
za
d
e
h,
Inte
rnat
ional
Journal
of
Mode
rn
Educ
ati
on
an
d
Computer
Sci
e
nce
(
IJM
ECS)
,
2017.
[13]
M.
Baghe
r
izade
h
and
M
.
Eshgh
i,
“
A
New
High
-
Speed
Carbon
Nanotube
Fi
el
d
Eff
ect
Tra
nsisto
r
-
Based
Struc
tur
e
for
4
-
to
-
2
Com
pre
ss
or
Cel
l
,
”
J
ournal
of
Comp
utat
ional
and
T
heore
tical
Nano
scie
nc
e
,
vol.
13
,
pp.
1006
-
1012
,
2016.
[14]
A.
Pis
hvai
e,
e
t
al.
,
“
Im
prove
d
CMO
S
(4;
2)
co
mpre
ss
or
de
signs
for
par
al
l
el
m
ultipli
ers
,
”
Comp
ut
ers
&
El
ec
tric
a
l
Engi
ne
ering
,
vol
.
38
,
pp
.
1703
-
1
716,
2012
.
0
0
.5
1
1
.5
2
2
.5
3
3
.5
4
4
.5
5
1
.5
2
.1
2
.8
3
.6
4
.5
PDP(E
-
15)
Desig
n
[10
]
Desig
n
[11
]
Desig
n
[12
]
Pr
o
p
o
sed
Desig
n
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.
8
, N
o.
6
,
Dece
m
ber
2
01
8
:
486
3
-
48
70
4870
[15]
D.
Bara
n
,
et
a
l.
,
“
Ene
rg
y
eff
i
cient
implemen
ta
t
ion
of
pa
rallel
CMO
S
m
ult
ipl
ie
rs
with
improv
ed
compress
ors,
”
Proce
ed
ings
of
t
he
16th
ACM
/IEEE
in
te
rnationa
l
sympos
ium
on
Low
power
el
ect
ronics
and
desig
n
,
pp.
147
-
152
,
2010.
[16]
M.
R.
Reshadi
ne
zha
d,
et
al.
,
“
An
ene
rg
y
-
eff
ic
i
ent
full
adde
r
c
el
l
u
sing
CNF
ET
te
c
hnolog
y
,
”
IE
ICE
transacti
ons
on
el
e
ct
ronics
,
vol
.
95,
pp
.
744
-
751
,
2012.
[17]
J.
Deng
and
H.
S.
P.
W
ong,
“
A
compac
t
SP
I
CE
m
odel
for
ca
rbon
-
nanot
ub
e
fie
ld
-
eff
e
ct
tra
n
sistors
inc
ludi
ng
nonide
a
li
ties
an
d
it
s
appl
icati
on
—
Part
I:
Model
of
the
int
rinsic
cha
nne
l
reg
ion
,
”
IEE
E
Tr
ansactions
on
El
ec
tron
Dev
ices
,
vo
l. 54, pp. 3186
-
3194,
2007.
[18]
J.
Deng
and
H.
S.
P.
W
ong,
“
A
compac
t
SP
I
CE
m
odel
for
ca
rbon
-
nanot
ub
e
fie
ld
-
eff
e
ct
tra
n
sistors
inc
ludi
ng
nonide
a
li
ties
an
d
it
s
app
licatio
n
—
Part
II:
Full
device
m
odel
and
c
i
rcu
i
t
p
erf
orm
anc
e
ben
ch
m
ark
ing,
”
IEEE
Tr
ansacti
ons on Electron
De
vi
c
e
s
,
vol. 54, pp. 31
95
-
3205,
2007
.
[19]
M.
Moradi
and
K.
Navi,
“
New
Curre
nt
-
Mode
T
ern
ar
y
Ful
l
Add
er
Circ
ui
ts
Base
d
on
Carbon
Nanotube
Fiel
d
Eff
ec
t
Tra
nsistor
Techn
olog
y
,
”
Journal
of
Computati
ona
l
and
Theore
ti
ca
l
Nanosci
enc
e
,
v
ol.
13
,
pp
.
327
-
3
37,
2016
.
BIOGR
AP
H
I
ES
OF
A
UTH
ORS
Mehdi
Baghe
ri
z
ade
h
recei
v
ed
his
B.
Sc.
degr
e
e
from
Shahid
Bahona
r
Univer
sit
y
in
computer
har
dware
engi
n
e
eri
ng.
He
is
al
so
rec
ei
v
ed
M.Sc.
and
Ph.D
degr
ee
in
computer
arc
hit
e
ct
ure
from
Scie
nc
e
and
Re
sea
rch
Bran
ch,
Islamic
Aza
d
Univer
sit
y
,
T
eh
ran
,
Ira
n
.
He
i
s
cur
ren
tly
an
As
sistant
Profess
or
in
Depa
rt
m
ent
of
Com
p
ute
r
Engi
n
e
eri
n
g,
Islamic
Aza
d
Univer
sit
y
,
Rafsa
nja
n
.
Ira
n
.
He
is
working
on
VLSI
and
ci
rcu
it
design
b
ase
d
on
CNF
ET
,
Appli
ca
t
ion
Speci
fi
c
Instru
ction
set
Proce
ss
or
,
and
W
SN
,
and
VA
NET.
E
-
m
ai
l:
m
.
bagheriz
ad
eh@srbia
u.
ac
.
ir
Mona
Moradi
r
ec
e
ive
d
h
er
B
.
S.
and
M.Sc.
deg
ree
s
in
Hardwar
e
Eng
ine
er
ing.
She
has
Ph.D.
degr
ee
in
2015
a
t
Scie
nc
e
and
Re
sea
rch
Bran
ch
of
Islamic
Aza
d
Univer
sit
y
under
supervision
of
Prof.
Keiva
n
Navi.
Her
rese
arch
int
ere
sts
incl
ude
VLSI,
Digit
al
Int
egr
ated
C
irc
uit
s
,
MV
L,
Arithmeti
c
Ci
rc
uit
s,
CNTFET
,
n
ano
elec
tron
ic,
h
ard
ware
se
cur
ity
and
Cr
y
p
togra
p
h
y
.
She
is
now
assistant
prof
essor a
t
Islamic
Az
ad
Univer
si
t
y
in Roudehe
n
Br
anch.
E
-
m
ai
l:
m
o.
m
ora
di@ri
au
.
a
c.ir
Mos
ta
fa
Tora
bi
rec
ei
v
ed
his
B.
Sc
degr
ee
in
C
om
pute
r
Software
Engi
ne
eri
ng
from
Mey
bod
Branc
h,
Isl
amic
Aza
d
unive
rsi
t
y
,
Me
y
bod
,
Ira
n
.
He
is
al
so
re
ce
iv
ed
M.Sc
degr
ee
from
Naja
fab
ad
Branc
h,
Isl
amic
Aza
d
Univ
ersity
,
Naj
afa
b
ad,
Ir
an
in
Com
pute
r
Software
Enginee
ring
.
He
is
cur
ren
t
l
y
an
as
sistant
profe
ss
o
r
in
Depa
rtme
nt
of
Com
puter
Enge
en
eri
ng
,
Islamic
Aza
d
Univer
sit
y
,
Rafs
anj
an
,
Ir
an. His
r
ea
serc
h
in
te
rests
are:
W
SN
,
VA
NET
and
Cl
aud Com
puti
ngs.
E
-
m
ai
l: m
.
tora
b
i
@ia
ura
fsan
ja
n.
a
c.
ir
Evaluation Warning : The document was created with Spire.PDF for Python.