I
nte
rna
t
io
na
l J
o
urna
l o
f
Rec
o
nfig
ura
ble a
nd
E
m
be
dd
e
d Sy
s
t
e
m
s
(
I
J
R
E
S)
Vo
l.
15
,
No
.
2
,
J
u
ly
20
26
,
p
p
.
269
~
2
8
0
I
SS
N:
2089
-
4864
,
DOI
:
1
0
.
1
1
5
9
1
/i
j
r
es
.
v
1
5
.
i
2
.
pp
2
6
9
-
280
269
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
r
es.ia
esco
r
e.
co
m
H
ig
h
-
pe
rfor
m
a
nc
e appro
x
i
m
a
te
M
AC
m
ul
tiplier
usi
ng
m
a
jori
ty
lo
g
ic
co
m
press
o
rs
for C
NNs
Selv
a
ra
s
a
n Ra
dh
a
k
r
is
h
na
n
1,
2
,
Su
dh
a
g
a
r
G
o
v
ind
ha
s
w
a
my
1
,
Ra
s
a
du
ra
i K
u
m
a
ra
v
el
2
1
D
e
p
a
r
t
me
n
t
o
f
E
l
e
c
t
r
o
n
i
c
s a
n
d
C
o
mm
u
n
i
c
a
t
i
o
n
E
n
g
i
n
e
e
r
i
n
g
,
B
h
a
r
a
t
h
I
n
st
i
t
u
t
e
o
f
H
i
g
h
e
r
Ed
u
c
a
t
i
o
n
a
n
d
R
e
se
a
r
c
h
,
C
h
e
n
n
a
i
,
I
n
d
i
a
2
D
e
p
ar
t
me
n
t
o
f
E
l
e
c
t
r
o
n
i
c
s a
n
d
C
o
mm
u
n
i
c
a
t
i
o
n
E
n
g
i
n
e
e
r
i
n
g
,
K
u
p
p
a
m E
n
g
i
n
e
e
r
i
n
g
C
o
l
l
e
g
e
,
K
u
p
p
a
m
,
I
n
d
i
a
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
No
v
2
8
,
2
0
2
5
R
ev
i
s
ed
Ma
r
6
,
2
0
2
6
A
cc
ep
ted
Ma
y
3
0
,
2
0
2
6
T
h
is
re
se
a
rc
h
p
re
se
n
ts
a
n
o
p
ti
m
iz
e
d
m
u
lt
ip
le
a
c
c
u
m
u
late
(M
A
C)
u
n
it
m
u
lt
ip
li
e
r
d
e
sig
n
f
o
r
e
ff
icie
n
t
c
o
n
v
o
lu
ti
o
n
a
l
n
e
u
ra
l
n
e
tw
o
rk
(CNN
)
o
p
e
ra
ti
o
n
s.
T
h
is
d
e
sig
n
m
a
in
l
y
f
o
c
u
se
s
o
n
m
a
k
in
g
th
e
m
u
lt
ip
li
e
r
s
y
ste
m
s
s
m
a
ll
e
r
b
y
u
sin
g
a
p
p
ro
x
im
a
t
e
m
a
jo
rit
y
c
o
m
p
re
ss
o
r
m
e
th
o
d
s
in
st
e
a
d
o
f
th
e
u
su
a
l
a
n
d
trad
it
i
o
n
a
l
a
p
p
r
o
x
im
a
te
m
e
th
o
d
s.
T
h
e
trad
it
i
o
n
a
l
a
p
p
ro
x
im
a
t
e
m
u
lt
ip
li
e
r
c
o
m
p
re
ss
o
r
tec
h
n
iq
u
e
s
a
r
e
lea
d
s
to
in
c
re
a
se
s
in
l
o
g
ic
si
z
e
,
c
rit
ica
l
p
a
th
d
e
lay
,
a
n
d
p
o
w
e
r
c
o
n
su
m
p
ti
o
n
;
h
o
w
e
v
e
r,
th
e
p
ro
p
o
se
d
re
se
a
rc
h
m
it
ig
a
tes
th
e
se
p
ro
b
lem
s
a
n
d
so
lv
e
s
th
e
m
w
it
h
a
n
o
v
e
lt
y
-
b
a
se
d
a
p
p
r
o
a
c
h
in
th
e
Da
d
d
a
m
u
lt
i
p
li
e
r
tec
h
n
iq
u
e
.
T
h
e
n
o
v
e
lt
y
o
f
th
is
a
p
p
ro
a
c
h
is
to
r
e
d
u
c
e
th
e
n
u
m
b
e
r
o
f
sta
g
e
s
in
th
e
m
u
lt
ip
li
e
r
d
e
sig
n
u
sin
g
4
:2
,
5
:
2
,
a
n
d
7
:2
c
o
m
p
re
ss
o
rs.
T
h
is
c
o
m
p
re
ss
o
r
is
d
e
sig
n
e
d
w
it
h
a
n
a
p
p
ro
x
im
a
t
e
m
e
th
o
d
u
sin
g
m
a
jo
rit
y
lo
g
ic;
c
o
m
p
a
r
e
d
to
th
is
trad
it
i
o
n
a
l
m
e
th
o
d
,
t
h
e
p
ro
p
o
se
d
m
a
jo
rit
y
a
p
p
ro
x
im
a
te
c
o
m
p
re
ss
o
r
m
e
th
o
d
p
ro
c
e
ss
e
d
les
s
e
rr
o
r
d
if
fe
re
n
c
e
s
in
m
u
lt
ip
li
c
a
ti
o
n
o
u
t
p
u
t
.
T
h
e
p
ro
p
o
se
d
a
p
p
ro
a
c
h
e
s
re
su
lt
e
d
i
n
sig
n
if
ica
n
t
re
d
u
c
ti
o
n
s
i
n
a
re
a
,
p
o
w
e
r,
a
n
d
d
e
la
y
re
lativ
e
to
trad
it
io
n
a
l
m
u
lt
ip
li
e
rs.
T
h
is
re
se
a
rc
h
c
o
m
p
a
re
d
se
v
e
n
u
n
iq
u
e
c
o
m
p
a
riso
n
s
o
f
M
A
C
-
b
a
se
d
m
u
lt
ip
li
e
r
a
rc
h
it
e
c
tu
re
,
a
n
d
it
w
il
l
h
a
v
e
b
e
e
n
d
e
v
e
lo
p
e
d
i
n
V
e
ril
o
g
h
a
rd
w
a
re
d
e
sc
rip
ti
o
n
lan
g
u
a
g
e
(HD
L
)
a
n
d
s
y
n
th
e
siz
e
d
o
n
th
e
X
il
i
n
x
V
e
rtex
-
5
F
P
G
A
,
p
ro
v
id
i
n
g
re
d
u
c
ti
o
n
s
o
f
5
8
.
4
%
i
n
lo
o
k
u
p
tab
le
(L
UT
)
a
n
d
7
6
.
2
%
in
o
c
c
u
p
ie
d
slice
s
,
a
n
d
p
r
o
v
in
g
les
s
p
o
w
e
r
c
o
n
su
m
p
ti
o
n
.
T
h
is
d
e
sig
n
is
a
h
ig
h
l
y
su
it
a
b
le
a
p
p
ro
a
c
h
f
o
r
re
a
l
-
ti
m
e
C
NN
a
n
d
d
ig
it
a
l
sig
n
a
l
p
ro
c
e
ss
i
n
g
(
DSP
)
a
p
p
li
c
a
ti
o
n
s
.
K
ey
w
o
r
d
s
:
A
p
p
r
o
x
i
m
ate
m
u
lt
ip
lier
C
o
n
v
o
lu
tio
n
al
n
e
u
r
al
n
e
t
w
o
r
k
Dad
d
a
m
u
ltip
lier
Fu
ll a
d
d
er
Mu
ltip
le
ac
c
u
m
u
late
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Selv
ar
asa
n
R
ad
h
ak
r
i
s
h
n
a
n
Dep
ar
t
m
en
t o
f
E
lectr
o
n
ics a
n
d
C
o
m
m
u
n
icat
io
n
E
n
g
i
n
ee
r
in
g
B
h
ar
ath
I
n
s
tit
u
te
o
f
Hi
g
h
er
E
d
u
ca
tio
n
a
n
d
R
esear
c
h
C
h
e
n
n
ai,
I
n
d
ia
E
m
ail:
ar
asa
n
ec
e
@
g
m
ail.
co
m
1.
I
NT
RO
D
UCT
I
O
N
T
h
e
co
n
v
o
lu
tio
n
al
n
eu
r
al
n
e
t
w
o
r
k
(
C
NN)
w
il
l
p
r
io
r
itize
ad
v
an
ce
m
e
n
t
s
in
th
e
late
s
t
ar
tif
icial
in
telli
g
e
n
ce
an
d
m
ac
h
i
n
e
lear
n
in
g
tec
h
n
iq
u
es,
b
ased
o
n
n
u
m
er
o
u
s
ap
p
licatio
n
s
.
T
h
ese
ap
p
licatio
n
s
in
cl
u
d
e
o
b
j
ec
t
d
etec
tio
n
,
im
a
g
e
i
m
p
r
o
v
e
m
e
n
t,
m
ed
ical
i
m
ag
e
d
ia
g
n
o
s
tics
,
an
d
o
th
er
s
i
m
ilar
ap
p
licatio
n
s
[
1
]
,
[
2
]
.
Her
e,
C
NN
is
p
ar
tic
u
lar
l
y
s
k
il
led
in
th
e
p
r
o
ce
s
s
o
f
ex
tr
ac
t
in
g
i
m
ag
e
f
ea
t
u
r
es
f
r
o
m
h
ig
h
-
d
i
m
en
s
io
n
a
l
d
ata,
an
d
it
ca
n
f
ac
ili
tate
th
e
p
r
o
d
u
ctio
n
o
f
ex
ac
t
d
ec
is
io
n
s
in
r
ea
l
-
t
i
m
e
i
m
ag
e
ap
p
licatio
n
s
[
3
]
.
I
n
p
ar
ticu
lar
,
th
ese
ap
p
licatio
n
s
r
eq
u
ir
ed
h
ar
d
w
ar
e
ac
ce
ler
ato
r
s
th
at
w
er
e
ab
le
to
s
u
cc
ess
f
u
ll
y
m
a
n
a
g
e
th
e
p
r
o
ce
s
s
in
g
n
ee
d
s
o
f
C
NN
o
p
er
atio
n
s
.
E
m
p
lo
y
in
g
th
e
m
i
n
co
n
j
u
n
ctio
n
w
ith
ed
g
e
co
m
p
u
ti
n
g
en
v
ir
o
n
m
e
n
ts
w
as
p
ar
ticu
lar
l
y
ef
f
ec
tiv
e
[
1
]
,
[
2
]
.
I
t
is
also
p
o
s
s
ib
le
to
co
m
b
i
n
e
t
h
e
m
u
ltip
le
ac
c
u
m
u
la
te
(
M
A
C
)
o
p
er
atio
n
s
w
it
h
t
h
i
s
tech
n
iq
u
e
to
b
o
o
s
t
th
e
p
e
r
f
o
r
m
an
ce
o
f
t
h
e
C
NN
in
ter
m
s
o
f
in
cr
ea
s
ed
s
p
ee
d
,
an
d
ef
f
icie
n
c
y
i
n
h
ar
d
w
ar
e
w
h
ile
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
9
-
4864
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
,
Vo
l.
15
,
No
.
2
,
J
u
l
y
202
6
:
269
-
2
8
0
270
m
ai
n
tai
n
in
g
a
b
alan
ce
d
er
r
o
r
ac
cu
m
u
lat
io
n
in
t
h
e
M
A
C
o
p
er
atio
n
s
,
as
d
is
cu
s
s
ed
b
y
P
ar
k
et
a
l.
[
4
]
,
an
d
also
ad
d
r
ess
ed
b
y
E
d
av
o
o
r
et
a
l
.
[
5
]
,
w
h
o
d
esig
n
ed
ap
p
r
o
x
i
m
ate
m
u
lt
ip
lier
s
f
o
r
ef
f
icie
n
t
n
e
u
r
al
n
et
w
o
r
k
ac
ce
ler
atio
n
.
T
h
e
tr
ad
itio
n
al
m
et
h
o
d
,
w
h
ic
h
u
s
es
p
r
ec
is
e
D
ad
d
a,
W
allac
e
tr
ee
,
an
d
Ved
ic
m
u
ltip
lier
s
,
u
s
u
all
y
lead
s
to
m
o
r
e
co
m
p
licated
lo
g
ic
an
d
h
i
g
h
er
p
o
w
er
co
n
s
u
m
p
t
io
n
[
6
]
–
[
8
]
.
Fo
r
in
s
tan
ce
,
t
h
e
Dad
d
a
an
d
W
allac
e
m
u
ltip
lier
s
w
ill
r
eq
u
ir
e
s
e
v
er
a
l
s
tag
es
f
o
r
th
e
r
ed
u
ctio
n
o
f
p
ar
tial
p
r
o
d
u
cts,
an
d
th
e
W
allac
e
tr
ee
,
in
p
ar
ticu
lar
,
w
il
l
also
in
te
g
r
ate
w
it
h
p
ar
allel
p
r
ef
ix
ad
d
er
s
to
p
r
o
v
id
e
v
er
y
f
a
s
t
o
p
er
atio
n
s
,
as
d
is
cu
s
s
ed
b
y
Mo
m
en
i
et
a
l.
[
6
]
,
an
d
f
u
r
th
er
elab
o
r
ated
b
y
Stro
llo
et
a
l.
[
7
]
an
d
J
ian
g
e
t
a
l.
[
8
]
,
w
h
o
an
a
l
y
ze
d
th
e
tr
ad
e
-
o
f
f
s
b
et
w
ee
n
d
if
f
er
en
t
m
u
lt
ip
lier
ar
ch
itect
u
r
es.
T
h
e
f
ac
t
th
at
th
e
s
e
p
r
o
ce
s
s
es
w
ill
ta
k
e
u
p
s
ev
er
al
lo
g
ical
ad
d
er
s
,
h
al
f
ad
d
er
s
,
an
d
co
m
p
r
ess
o
r
s
t
y
p
icall
y
r
es
u
lt
s
in
a
n
i
n
cr
ea
s
e
in
t
h
e
a
m
o
u
n
t
o
f
ar
ea
,
d
elay
,
a
n
d
p
o
w
er
co
n
s
u
m
p
tio
n
.
T
h
ese
d
etails
ab
o
u
t
co
m
p
r
es
s
o
r
an
d
m
u
ltip
lier
s
tr
u
ctu
r
e
w
er
e
ex
p
lai
n
ed
b
y
B
ala
et
a
l.
[
9
]
,
w
it
h
ad
d
itio
n
al
d
is
cu
s
s
io
n
b
y
G
h
a
s
e
m
za
d
e
h
et
a
l.
[
1
0
]
an
d
Fath
i
et
a
l.
[
1
1
]
,
w
h
o
f
o
cu
s
ed
o
n
h
i
g
h
-
s
p
ee
d
an
d
ef
f
icien
t
co
m
p
r
ess
o
r
d
esig
n
s
.
I
t
is
n
ec
es
s
ar
y
f
o
r
th
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
to
lo
w
er
th
e
n
u
m
b
er
o
f
lo
g
ic
g
ate
s
to
less
e
n
t
h
e
co
m
p
le
x
it
y
o
f
t
h
e
l
o
g
ic
s
ize
[
1
2
]
,
[
1
3
]
.
T
h
e
s
ta
n
d
ar
d
ap
p
r
o
ac
h
o
f
a
f
u
ll
ad
d
er
w
il
l
n
ee
d
u
p
to
f
i
v
e
lo
g
ic
g
ates,
an
d
4
:2
co
m
p
r
es
s
o
r
s
w
ill
u
tili
ze
u
p
to
1
0
l
o
g
ic
g
ates,
s
i
m
ilar
l
y
,
th
e
p
r
o
p
o
s
e
d
tech
n
iq
u
e
o
f
an
ap
p
r
o
x
im
a
te
m
aj
o
r
it
y
ad
d
er
w
il
l
o
cc
u
p
y
o
n
l
y
t
w
o
lo
g
ic
g
ates,
an
d
u
s
i
n
g
th
is
ap
p
r
o
x
i
m
ate
f
u
ll
ad
d
er
[
1
4
]
in
4
:2
,
5
:
2
,
an
d
7
:
2
co
m
p
r
ess
o
r
s
,
it
w
ill
r
ed
u
ce
n
u
m
b
er
o
f
lo
g
ic
g
ate
s
an
d
r
ed
u
ce
s
th
e
co
m
p
lex
it
y
o
f
m
u
ltip
licatio
n
.
T
h
e
g
o
al
o
f
th
is
s
t
u
d
y
i
s
to
r
ed
u
ce
m
u
lt
ip
licatio
n
co
m
p
le
x
it
y
b
y
u
s
i
n
g
m
aj
o
r
it
y
lo
g
ic
co
m
p
r
es
s
o
r
s
.
An
ad
d
itio
n
al
a
s
p
ec
t to
co
n
s
id
er
is
th
e
m
u
ltip
li
er
d
esig
n
t
h
at
i
n
co
r
p
o
r
ates M
AC
ar
ch
itect
u
r
e.
T
h
e
ap
p
licatio
n
s
o
f
d
ig
ital
s
ig
n
a
l
p
r
o
ce
s
s
in
g
(
DSP
)
r
eq
u
ir
e
ca
r
ef
u
l
d
esig
n
an
d
o
p
tim
izatio
n
to
r
eso
lv
e
p
o
w
er
s
a
v
i
n
g
i
s
s
u
es
wh
ile
m
ai
n
tain
in
g
a
n
ac
ce
p
tab
le
er
r
o
r
r
ate
[
3
]
,
[
1
5
]
,
[
1
6
]
.
T
h
e
ev
alu
a
tio
n
o
f
er
r
o
r
m
etr
ics,
w
h
ic
h
i
n
t
h
i
s
ca
s
e
i
n
clu
d
es
er
r
o
r
r
ate
(
E
R
%),
m
ea
n
er
r
o
r
d
is
tan
ce
(
ME
D)
,
n
o
r
m
alize
d
m
ea
n
er
r
o
r
d
is
tan
ce
(
NM
E
D
)
,
an
d
m
ea
n
r
elativ
e
er
r
o
r
d
is
tan
ce
(
MR
E
D
)
,
is
d
o
n
e
to
d
em
o
n
s
tr
ate
th
e
ar
ch
itec
tu
r
e
's
r
esil
ien
ce
i
n
s
it
u
atio
n
s
w
h
er
e
ap
p
r
o
x
im
a
te
co
m
p
u
tatio
n
i
s
a
cc
ep
tab
le,
as
d
escr
ib
ed
b
y
Z
e
n
d
eg
a
n
i
et
a
l.
[
1
5
]
,
an
d
also
d
is
cu
s
s
ed
b
y
An
s
ar
i
et
a
l.
[
3
]
an
d
Mr
az
ek
et
a
l.
[
1
6
]
,
w
h
o
e
v
alu
a
ted
er
r
o
r
m
etr
ics
f
o
r
v
ar
io
u
s
ap
p
r
o
x
im
a
te
m
u
ltip
lier
s
.
Deta
iled
in
f
o
r
m
at
io
n
o
n
th
e
p
r
ec
is
e
an
d
p
r
o
p
o
s
ed
m
aj
o
r
ity
lo
g
ic
co
m
p
r
ess
o
r
s
f
o
r
4
:2
,
5
:2
,
an
d
7
:2
co
n
f
ig
u
r
atio
n
s
i
s
p
r
o
v
id
ed
in
th
e
r
e
m
a
in
i
n
g
p
ar
t
o
f
th
i
s
s
t
u
d
y
,
w
h
ich
is
s
ec
tio
n
1
.
Sectio
n
2
d
escr
ib
e
th
e
o
v
er
v
ie
w
o
f
e
x
a
ct
an
d
ap
p
r
o
x
i
m
ate
co
m
p
r
es
s
o
r
s
,
th
e
r
e
m
ai
n
d
er
o
f
t
h
is
p
ap
er
is
o
r
g
an
ized
as
f
o
llo
w
s
.
Sectio
n
3
p
r
esen
ts
t
h
e
m
e
th
o
d
o
lo
g
y
f
o
r
th
e
e
x
ac
t
m
u
ltip
lier
d
esi
g
n
u
s
i
n
g
an
ap
p
r
o
x
im
a
te
ad
d
er
ar
ch
itect
u
r
e,
in
clu
d
i
n
g
t
h
e
ar
ch
itect
u
r
al
f
r
a
m
e
w
o
r
k
an
d
o
p
er
atio
n
al
p
r
in
cip
les.
Sectio
n
4
d
is
cu
s
s
es
t
h
e
ex
p
er
i
m
e
n
tal
r
es
u
lt
s
an
d
p
er
f
o
r
m
a
n
ce
e
v
alu
a
tio
n
,
h
ig
h
li
g
h
t
in
g
th
e
e
f
f
ec
ti
v
e
n
ess
o
f
t
h
e
p
r
o
p
o
s
ed
d
esig
n
i
n
ter
m
s
o
f
ac
c
u
r
ac
y
,
p
o
w
er
co
n
s
u
m
p
t
io
n
,
ar
ea
u
tili
za
t
io
n
,
a
n
d
co
m
p
u
tatio
n
al
ef
f
icie
n
c
y
.
Fin
a
ll
y
,
s
ec
tio
n
5
co
n
clu
d
es t
h
e
p
ap
er
b
y
s
u
m
m
ar
izin
g
t
h
e
k
e
y
f
i
n
d
in
g
s
a
n
d
o
u
tli
n
i
n
g
p
o
ten
t
ial
d
ir
ec
tio
n
s
f
o
r
f
u
t
u
r
e
r
esear
ch
.
2.
O
VE
RVI
E
W
O
F
E
XACT A
ND
AP
P
RO
XIM
AT
E
CO
M
P
RE
SS
O
RS
C
o
m
p
r
ess
o
r
s
i
m
p
r
o
v
e
th
e
d
esi
g
n
o
f
h
ar
d
w
ar
e
b
y
r
ed
u
cin
g
th
e
n
u
m
b
er
o
f
s
ta
g
es
a
n
d
p
ar
tial
p
r
o
d
u
cts
in
t
h
e
m
u
ltip
licatio
n
o
p
er
atio
n
s
,
a
s
d
escr
ib
ed
b
y
Z
en
d
e
g
a
n
i
et
a
l.
[
1
5
]
an
d
E
s
p
o
s
ito
e
t
a
l.
[
1
7
]
.
I
n
th
is
co
m
p
r
es
s
o
r
ar
ch
itectu
r
e,
th
er
e
ar
e
tw
o
ca
teg
o
r
ies
o
f
t
h
e
d
esig
n
:
e
x
ac
t
an
d
ap
p
r
o
x
i
m
ate
m
e
th
o
d
s
,
as
d
is
cu
s
s
ed
b
y
Mr
az
e
k
et
a
l.
[
1
6
]
.
Du
e
to
th
e
f
ac
t
th
a
t
th
e
e
x
ac
t
co
m
p
r
ess
o
r
w
ill
ta
k
e
u
p
m
o
r
e
lo
g
ic
ar
ea
,
it
is
p
r
ef
er
r
ed
f
o
r
ap
p
licatio
n
s
th
at
r
eq
u
ir
e
a
h
ig
h
er
lev
el
o
f
ac
cu
r
ac
y
,
as
ex
p
lain
ed
b
y
Mo
m
e
n
i
et
a
l.
[
6
]
.
Ho
w
e
v
er
,
in
m
an
y
ap
p
licatio
n
s
,
t
h
is
h
i
g
h
er
lev
el
o
f
ac
cu
r
ac
y
is
n
o
t
r
eq
u
ir
ed
;
in
s
tead
,
th
e
m
ai
n
p
r
io
r
it
y
is
m
i
n
i
m
izi
n
g
lo
g
ic
s
ize
an
d
p
o
w
er
co
n
s
u
m
p
t
io
n
.
Fo
r
s
u
c
h
ca
s
es,
ap
p
r
o
x
i
m
ate
co
m
p
r
ess
o
r
s
ar
e
u
tili
ze
d
,
as
r
ep
o
r
t
ed
b
y
Z
en
d
e
g
an
i
e
t
a
l.
[
1
5
]
an
d
E
s
p
o
s
ito
et
a
l.
[
1
7
]
.
A
f
u
n
d
a
m
e
n
tall
y
w
id
e
-
r
an
g
in
g
ad
d
er
-
b
ased
co
m
p
r
es
s
o
r
d
esig
n
is
co
m
m
o
n
l
y
u
s
ed
;
th
e
co
m
p
r
es
s
o
r
lo
g
ic
s
i
ze
d
ep
en
d
s
u
p
o
n
th
e
u
n
d
er
l
y
i
n
g
ad
d
er
d
esig
n
.
W
h
e
n
th
e
a
d
d
er
o
cc
u
p
ies
m
o
r
e
lo
g
ic,
th
e
co
m
p
r
ess
o
r
also
in
c
r
ea
s
es
in
lo
g
ic
a
n
d
p
o
w
er
co
n
s
u
m
p
tio
n
.
T
o
ad
d
r
ess
th
is
,
w
e
in
tr
o
d
u
ce
d
a
n
o
v
el
ad
d
er
a
r
ch
itectu
r
e
in
t
h
i
s
d
esig
n
,
w
h
ich
m
i
n
i
m
izes
lo
g
ic
s
iz
e
an
d
also
in
tr
o
d
u
ce
s
ap
p
r
o
x
im
atio
n
i
n
th
e
o
u
tp
u
t,
as
s
ee
n
in
th
e
w
o
r
k
s
o
f
E
d
a
v
o
o
r
et
a
l.
[
5
]
an
d
Stro
llo
et
a
l.
[
7
]
.
A
d
d
itio
n
all
y
,
t
h
i
s
ap
p
r
o
ac
h
m
i
n
i
m
ize
s
th
e
co
m
p
le
x
it
y
o
f
t
h
e
h
ar
d
w
ar
e,
as
w
ell
a
s
t
h
e
a
m
o
u
n
t
o
f
p
o
w
er
co
n
s
u
m
ed
a
n
d
d
el
a
y
e
x
p
er
ien
ce
d
.
T
h
er
e
ar
e
th
r
ee
d
is
ti
n
ct
d
esi
g
n
s
f
o
r
th
e
p
r
o
p
o
s
ed
co
m
p
r
ess
o
r
m
eth
o
d
in
tr
o
d
u
ce
d
:
an
ap
p
r
o
x
i
m
at
e
f
u
ll
ad
d
er
-
b
ased
co
m
p
r
es
s
o
r
,
an
ap
p
r
o
x
i
m
ate
f
u
ll
ad
d
er
[
1
8
]
–
[
2
0
]
w
ith
a
m
u
ltip
le
x
er
-
b
a
s
ed
co
m
p
r
ess
o
r
,
an
d
a
m
aj
o
r
it
y
g
ate
-
b
ased
co
m
p
r
ess
o
r
[
2
1
]
–
[
2
5
]
ea
ch
o
f
t
h
ese
d
esi
g
n
s
f
ea
tu
r
es
i
ts
o
w
n
u
n
iq
u
e
co
llecti
o
n
o
f
ar
c
h
itect
u
r
al
ele
m
e
n
ts
.
T
h
e
p
er
f
o
r
m
a
n
ce
o
f
ea
ch
o
f
t
h
ese
co
m
p
r
es
s
o
r
s
w
a
s
ev
al
u
ated
a
n
d
co
m
p
ar
ed
to
tr
ad
itio
n
al
ex
ac
t
co
m
p
r
es
s
o
r
s
.
2
.
1
.
E
x
a
ct
co
m
press
o
r
E
x
ac
t
co
m
p
r
es
s
o
r
s
w
er
e
d
ev
el
o
p
ed
to
ca
r
r
y
o
u
t
ar
i
th
m
etic
o
p
er
atio
n
s
w
it
h
e
x
ac
t
p
r
ec
is
io
n
,
en
s
u
r
in
g
th
at
t
h
e
s
u
m
a
n
d
ca
r
r
y
b
its
ar
e
ca
lcu
lated
p
r
ec
is
el
y
an
d
w
it
h
o
u
t
a
n
y
lo
s
s
o
f
b
its
.
Utilizatio
n
o
f
th
e
s
e
co
m
p
r
es
s
o
r
s
is
w
id
e
s
p
r
ea
d
in
ap
p
l
icatio
n
s
w
h
er
e
ac
c
u
r
ac
y
is
o
f
t
h
e
u
t
m
o
s
t
i
m
p
o
r
tan
ce
,
as
h
i
g
h
lig
h
ted
b
y
Mo
m
en
i
et
a
l.
[
6
]
an
d
Gh
asem
za
d
eh
et
a
l.
[
1
0
]
.
A
lt
h
o
u
g
h
th
e
y
ar
e
ex
tr
e
m
el
y
r
el
iab
le,
ac
cu
r
ate
co
m
p
r
ess
o
r
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
I
SS
N:
2089
-
4864
Hig
h
-
p
erfo
r
ma
n
ce
a
p
p
r
o
xima
te
MAC mu
ltip
lier
u
s
in
g
ma
jo
r
ity
…
(
S
elva
r
a
s
a
n
R
a
d
h
a
kris
h
n
a
n
)
271
m
i
g
h
t
n
ee
d
a
s
ig
n
i
f
ica
n
t
a
m
o
u
n
t
o
f
h
ar
d
w
ar
e
r
eso
u
r
ce
s
,
w
h
ic
h
ca
n
r
es
u
lt
in
m
o
r
e
p
o
w
er
co
n
s
u
m
p
tio
n
,
in
cr
ea
s
ed
lo
g
ic
s
ize,
a
n
d
g
r
ea
ter
co
m
p
u
tatio
n
al
late
n
c
y
co
m
p
ar
ed
to
th
eir
ap
p
r
o
x
i
m
ate
eq
u
iv
a
len
t
s
,
as
d
is
cu
s
s
ed
b
y
Fat
h
i
et
a
l.
[
1
1
]
,
R
ed
d
y
et
a
l.
[
1
2
]
,
an
d
Z
e
n
d
eg
a
n
i
et
a
l.
[
1
5
]
.
T
h
e
ex
ac
t
co
m
p
r
ess
o
r
ar
ch
itect
u
r
es
e
m
p
lo
y
ed
i
n
th
e
m
u
ltip
lier
r
ed
u
ctio
n
s
ta
g
e
h
a
v
e
b
ee
n
s
h
o
w
n
i
n
Fi
g
u
r
e
1
.
T
h
ese
4
:2
,
5
:2
,
an
d
7
:
2
ar
ch
itect
u
r
es
ar
e
th
e
m
o
s
t
co
m
m
o
n
t
y
p
e
s
o
f
e
x
ac
t
co
m
p
r
es
s
o
r
d
esig
n
s
w
h
ich
f
o
r
m
b
asic
u
n
i
ts
i
n
r
ed
u
ci
n
g
t
h
e
p
ar
tial
-
p
r
o
d
u
cts
s
i
n
ce
th
e
y
co
m
b
in
e
a
p
lu
r
ali
t
y
o
f
in
p
u
ts
i
n
t
o
f
e
w
er
o
u
tp
u
ts
w
i
th
o
u
t
ch
a
n
g
in
g
t
h
e
ar
ith
m
eti
c
co
r
r
ec
tn
ess
.
T
h
e
f
o
llo
w
in
g
lo
g
ical
(
1
)
ar
e
r
elev
an
t to
th
is
d
es
ig
n
.
=
1
⊕
2
⊕
3
⊕
4
⊕
=
(
(
1
⊕
2
⊕
3
)
·
4
)
+
(
·
(
(
1
⊕
2
⊕
3
)
⊕
4
)
)
=
(
1
·
2
)
+
(
3
·
(
1
⊕
2
)
)
(
1
)
B
ased
o
n
th
e
co
n
f
i
g
u
r
atio
n
o
f
t
h
e
4
:2
co
m
p
r
es
s
o
r
s
,
wh
ich
ca
n
b
e
s
ee
n
i
n
Fig
u
r
e
1
(
a)
,
th
is
ar
ch
itect
u
r
e
is
d
ev
elo
p
ed
w
it
h
a
5
-
in
p
u
t
a
n
d
3
-
o
u
tp
u
t
d
esi
g
n
,
in
cl
u
d
in
g
ca
r
r
y
-
i
n
an
d
ca
r
r
y
-
o
u
t,
as
d
escr
ib
ed
b
y
G
h
ase
m
za
d
eh
et
a
l.
[
1
0
]
a
n
d
R
ed
d
y
et
a
l.
[
1
2
]
.
T
h
e
n
ex
t
ar
ch
itectu
r
e
(
Fig
u
r
e
1
(
b
)
,
th
e
5
:2
co
m
p
r
ess
o
r
s
,
h
as
f
iv
e
i
n
p
u
t
b
its
(
1
,
2
,
3
,
4
,
5
)
co
m
b
i
n
ed
w
it
h
t
w
o
ca
r
r
y
-
i
n
s
i
g
n
a
ls
(
1
an
d
2
)
.
T
h
ese
5
:2
co
m
p
r
es
s
o
r
s
ar
e
ab
le
to
im
p
r
o
v
e
u
p
o
n
th
e
ca
p
ab
ilit
ies
o
f
t
h
e
4
:2
c
o
m
p
r
es
so
r
s
an
d
also
r
ed
u
ce
th
e
n
u
m
b
er
o
f
p
ar
tial
p
r
o
d
u
cts
in
m
u
ltip
lier
d
esig
n
,
as
d
is
cu
s
s
ed
b
y
Fat
h
i
et
a
l.
[
1
1
]
an
d
E
d
av
o
o
r
et
a
l.
[
5
]
.
I
n
th
e
ca
s
e
o
f
th
e
5
:2
co
m
p
r
ess
o
r
s
,
th
e
lo
g
ical
e
q
u
atio
n
s
ar
e
p
r
esen
ted
i
n
(
2
)
.
=
1
⊕
2
⊕
3
⊕
4
⊕
5
⊕
1
⊕
2
5
=
(
5
·
2
)
+
(
2
·
(
4
⊕
1
⊕
(
1
⊕
2
⊕
3
)
)
+
(
(
4
⊕
1
⊕
(
1
⊕
2
⊕
3
)
)
·
5
)
1
=
(
1
·
2
)
+
(
3
·
(
1
⊕
2
)
)
2
=
(
4
·
1
)
+
(
1
·
(
1
⊕
2
⊕
3
)
)
+
(
(
1
⊕
2
⊕
3
)
·
4
)
(
2
)
T
h
er
ef
o
r
e,
th
e
5
:2
c
o
m
p
r
ess
o
r
s
ar
e
co
n
s
id
er
ed
b
etter
f
o
r
m
u
lt
ip
licatio
n
o
p
er
atio
n
s
,
alth
o
u
g
h
i
t
r
eq
u
ir
es
m
o
r
e
lo
g
ic
ar
ea
d
u
e
t
o
th
e
n
u
m
b
er
o
f
f
u
l
l
ad
d
er
s
u
s
ed
in
its
d
e
s
ig
n
,
as
a
n
al
y
ze
d
b
y
E
d
av
o
o
r
et
a
l.
[
5
]
an
d
Fat
h
i
et
a
l.
[
1
1
]
.
Si
m
i
lar
l
y
,
th
e
7
:2
co
m
p
r
ess
o
r
m
et
h
o
d
also
u
s
es
f
i
v
e
f
u
l
l
ad
d
er
s
in
its
d
esig
n
,
as
v
er
if
ied
in
F
ig
u
r
e
1
(
c)
.
T
h
is
ar
ch
i
tectu
r
e
h
as
s
e
v
en
in
p
u
t
b
it
s
(
1
,
2
,
3
,
4
,
5
,
6
,
7
)
an
d
t
w
o
ca
r
r
y
-
in
s
(
ci
n
1
an
d
cin
2
)
,
r
esu
lti
n
g
i
n
f
o
u
r
o
u
tp
u
ts
(
,
,
1
,
2
)
,
as
d
escr
ib
e
d
b
y
G
h
ase
m
za
d
eh
et
a
l.
[
1
0
]
an
d
Fat
h
i
et
a
l.
[
1
1
]
.
T
h
e
d
etails o
f
t
h
is
lo
g
ic
ar
e
b
r
ief
l
y
e
x
p
lain
ed
in
t
h
e
f
o
llo
w
i
n
g
lo
g
ical
(
3
)
.
=
1
⊕
2
⊕
3
⊕
4
⊕
5
⊕
6
⊕
7
⊕
=
(
(
1
⊕
2
⊕
3
⊕
4
⊕
5
⊕
6
⊕
7
)
·
1
)
+
(
c
in
_
2
(
(
x1
⊕
x2
⊕
x3
⊕
x4
⊕
x5
⊕
x6
⊕
x7
)
⊕
c
in
_
1
)
)
1
=
(
1
·
2
)
+
(
3
·
(
1
⊕
2
)
)
2
=
(
(
1
⊕
2
⊕
3
)
·
4
)
+
(
5
·
(
(
1
⊕
2
⊕
3
)
⊕
4
)
)
3
=
(
(
1
⊕
2
⊕
3
⊕
4
⊕
5
)
·
6
)
+
(
7
·
(
(
1
⊕
2
⊕
3
⊕
4
⊕
5
)
⊕
6
)
)
(
3
)
F
u
l
l
A
d
d
er
F
u
l
l
A
d
d
er
x
1
x
2
x
3
x
4
c
i
n
S
u
m
C
a
r
r
y
C
o
u
t
F
u
l
l
A
d
d
er
F
u
l
l
A
d
d
e
r
x
1
x
2
x
3
x
4
c
i
n
1
S
u
m
C
a
r
r
y
C
o
u
t
1
F
u
l
l
A
d
d
er
x
5
c
i
n
2
C
o
u
t
2
F
u
l
l
A
d
d
er
F
u
l
l
A
d
d
er
x
1
x
2
x
3
c
i
n
1
S
u
m
C
a
r
r
y
C
o
u
t
1
F
u
l
l
A
d
d
er
c
i
n
2
C
o
u
t
2
x
4
x
5
x
6
x
7
F
u
l
l
A
d
d
er
F
u
l
l
A
d
d
er
(
a
)
(
b
)
(
c
)
(
a)
(
b
)
(
c)
Fig
u
r
e
1
.
A
r
ch
itectu
r
e
s
o
f
co
m
p
r
e
s
s
o
r
cir
cu
it
s
:
(
a)
4
:2
ex
ac
t
co
m
p
r
es
s
o
r
,
(
b
)
5
:
2
ex
ac
t
co
m
p
r
e
s
s
o
r,
an
d
(
c)
7
:2
ex
ac
t
co
m
p
r
ess
o
r
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
9
-
4864
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
,
Vo
l.
15
,
No
.
2
,
J
u
l
y
202
6
:
269
-
2
8
0
272
2
.
2
.
Appro
x
i
m
a
t
e
co
m
pre
s
s
o
r
T
h
er
e
is
a
h
ar
d
w
ar
e
-
e
f
f
icie
n
t
alter
n
ati
v
e
to
p
r
ec
is
e
co
m
p
r
e
s
s
o
r
s
k
n
o
w
n
as
ap
p
r
o
x
i
m
ate
co
m
p
r
e
s
s
o
r
s
.
T
h
ese
co
m
p
r
ess
o
r
s
w
er
e
d
ev
e
lo
p
ed
p
r
im
ar
il
y
f
o
r
ap
p
licatio
n
s
t
h
at
ca
n
to
ler
ate
m
o
d
est
co
m
p
u
tat
io
n
al
er
r
o
r
s
,
as
d
is
cu
s
s
ed
b
y
R
ed
d
y
et
a
l.
[
1
2
]
an
d
J
o
o
q
et
a
l.
[
1
3
]
.
On
t
h
e
o
th
er
h
an
d
,
th
ese
s
y
s
te
m
s
s
ac
r
if
ice
ac
cu
r
ac
y
i
n
ex
ch
a
n
g
e
f
o
r
co
n
s
id
er
ab
le
r
ed
u
ctio
n
s
i
n
h
ar
d
w
ar
e
co
m
p
le
x
it
y
,
p
o
w
er
co
n
s
u
m
p
tio
n
,
a
n
d
lat
en
c
y
,
a
s
s
h
o
w
n
b
y
E
d
av
o
o
r
et
a
l.
[
5
]
an
d
Stro
llo
et
a
l.
[
7
]
.
A
s
a
r
es
u
lt
o
f
t
h
e
ar
ch
itect
u
r
al
d
esi
g
n
th
at
in
co
r
p
o
r
ates
ap
p
r
o
x
i
m
ate
f
u
ll
ad
d
er
s
,
ap
p
r
o
x
im
a
te
co
m
p
r
e
s
s
o
r
s
p
r
o
v
id
e
o
u
tp
u
t
s
t
h
at
ar
e
ap
p
r
o
x
i
m
ate
eq
u
i
v
al
en
t,
as
p
r
esen
ted
b
y
E
d
av
o
o
r
et
a
l.
[
5
]
an
d
J
o
o
q
e
t
a
l.
[
1
3
]
.
T
h
e
(
4
)
c
o
n
tain
s
t
h
e
lo
g
ical
eq
u
atio
n
s
o
f
an
ap
p
r
o
x
i
m
ate
co
m
p
lete
ad
d
er
,
w
h
ich
ar
e
r
esp
o
n
s
ib
le
f
o
r
th
e
cr
ea
tio
n
o
f
th
e
s
u
m
an
d
ca
r
r
y
v
ar
iab
les.
=
1
⊕
2
⊕
=
2
·
(
4
)
T
h
e
d
esig
n
ar
ch
itect
u
r
e
o
f
th
e
ap
p
r
o
x
i
m
ate
f
u
ll
ad
d
er
[
1
4
]
,
as
s
h
o
w
n
i
n
Fi
g
u
r
e
2
u
s
es
o
n
l
y
t
w
o
lo
g
i
c
g
ates
(
XO
R
an
d
AND
g
ate
s
)
,
as
p
r
o
p
o
s
ed
b
y
R
ed
d
y
et
a
l.
[
1
2
]
an
d
J
o
o
q
et
a
l.
[
1
3
]
.
C
o
m
p
ar
ed
to
th
e
ex
ac
t
m
et
h
o
d
,
th
e
tr
ad
itio
n
al
ar
c
h
itectu
r
e
u
s
es
f
i
v
e
lo
g
ic
g
at
es;
th
is
r
ed
u
ct
io
n
co
n
tr
ib
u
te
s
to
lo
w
er
p
o
w
er
co
n
s
u
m
p
tio
n
,
r
ed
u
ce
d
ar
ea
,
an
d
f
as
ter
co
m
p
u
tat
io
n
ti
m
es,
as
d
e
m
o
n
s
tr
ated
b
y
E
d
av
o
o
r
et
a
l.
[
5
]
an
d
Stro
llo
et
a
l.
[
7
]
.
Usi
n
g
t
h
e
d
esi
g
n
o
f
an
ap
p
r
o
x
i
m
ate
f
u
ll
ad
d
er
[
1
4
]
,
T
h
e
ap
p
r
o
x
im
a
te
-
ad
d
e
r
-
b
ased
co
m
p
r
ess
o
r
ar
ch
itect
u
r
es
u
s
ed
in
t
h
e
m
u
ltip
lier
d
esi
g
n
ar
e
s
h
o
w
n
i
n
Fi
g
u
r
e
3
.
T
h
e
f
ig
u
r
e
s
h
o
w
s
th
e
s
u
m
m
ar
y
o
f
ap
p
r
o
x
im
a
te
ad
d
itio
n
p
r
in
cip
l
es
u
s
ed
to
r
ea
lis
e
co
m
p
r
e
s
s
o
r
s
o
f
d
if
f
er
en
t
i
n
p
u
t
-
o
u
tp
u
t
co
n
f
ig
u
r
at
io
n
s
,
4
:2
,
5
:2
an
d
7
:2
co
m
p
r
ess
o
r
s
.
t
h
e
p
r
o
p
o
s
ed
m
eth
o
d
i
n
te
g
r
ated
th
e
ar
ch
itect
u
r
e
o
f
4
:2
,
5
:2
,
an
d
7
:2
co
m
p
r
ess
o
rs
a
n
d
d
em
o
n
s
tr
ated
t
h
eir
p
er
f
o
r
m
a
n
ce
,
as
s
h
o
w
n
i
n
Fi
g
u
r
e
3
(
a)
4
:
2
co
m
p
r
ess
o
r
s
,
F
ig
u
r
e
3
(
b
)
5
:
2
co
m
p
r
ess
o
r
s
,
a
n
d
Fig
u
r
e
3
(
c)
7
:2
c
o
m
p
r
es
s
o
r
s
,
f
o
llo
w
i
n
g
t
h
e
s
tr
ate
g
ie
s
d
is
cu
s
s
ed
b
y
E
d
av
o
o
r
et
a
l.
[
5
]
an
d
J
o
o
q
et
a
l.
[
1
3
]
.
A
B
C
i
n
S
u
m
C
a
r
r
y
Fig
u
r
e
2
.
A
p
p
r
o
x
i
m
a
te
f
u
ll a
d
d
er
d
esig
n
A
p
p
r
o
x
i
m
a
t
e
F
u
l
l
A
d
d
er
A
p
p
r
o
x
i
m
a
t
e
F
u
l
l
A
d
d
er
x
1
x
2
x
3
x
4
c
i
n
S
u
m
C
a
r
r
y
C
o
u
t
A
p
p
r
o
x
i
m
a
t
e
F
u
l
l
A
d
d
er
A
p
p
r
o
x
i
m
a
t
e
F
u
l
l
A
d
d
er
x
1
x
2
x
3
x
4
c
i
n
1
S
u
m
C
a
r
r
y
C
o
u
t
1
A
p
p
r
o
x
i
m
a
t
e
F
u
l
l
A
d
d
er
x
5
c
i
n
2
C
o
u
t
2
A
p
p
r
o
x
i
m
a
t
e
F
u
l
l
A
d
d
er
A
p
p
r
o
x
i
m
a
t
e
F
u
l
l
A
d
d
er
x
1
x
2
x
3
c
i
n
1
S
u
m
C
a
r
r
y
C
o
u
t
1
A
p
p
r
o
x
i
m
a
t
e
F
u
l
l
A
d
d
e
r
c
i
n
2
C
o
u
t
2
x
4
x
5
x
6
x
7
A
p
p
r
o
x
i
m
a
t
e
F
u
l
l
A
d
d
er
A
p
p
r
o
x
i
m
a
t
e
F
u
l
l
A
d
d
er
(
a
)
(
b
)
(
c
)
(
a)
(
b
)
(
c)
Fig
u
r
e
3
.
A
p
p
r
o
x
i
m
a
te
-
ad
d
er
-
b
ased
ar
ch
itectu
r
es
:
(
a)
4
:2
ap
p
r
o
x
i
m
ate
co
m
p
r
ess
o
r
,
(
b
)
5
:2
ap
p
r
o
x
im
ate
co
m
p
r
es
s
o
r
,
an
d
(
c)
7
:2
a
p
p
r
o
x
i
m
ate
co
m
p
r
ess
o
r
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
I
SS
N:
2089
-
4864
Hig
h
-
p
erfo
r
ma
n
ce
a
p
p
r
o
xima
te
MAC mu
ltip
lier
u
s
in
g
ma
jo
r
ity
…
(
S
elva
r
a
s
a
n
R
a
d
h
a
kris
h
n
a
n
)
273
T
h
e
co
r
e
o
f
ap
p
r
o
x
im
ate
f
u
ll a
d
d
er
b
ased
4
:2
c
o
m
p
r
es
s
o
r
lo
g
ical
(
5
)
ar
e
as
:
=
1
⊕
2
⊕
3
⊕
4
⊕
=
4
.
=
2
.
(
5
)
f
o
r
5
:2
co
m
p
r
ess
o
r
s
u
s
i
n
g
ap
p
r
o
x
i
m
ate
f
u
ll
ad
d
er
[
1
4
]
,
th
e
f
o
llo
w
in
g
is
t
h
e
lo
g
ica
l
eq
u
a
tio
n
th
at
s
h
o
u
ld
b
e
u
s
ed
(
6
)
.
=
1
⊕
2
⊕
3
⊕
4
⊕
5
⊕
1
⊕
2
=
5
.
2
1
=
2
.
3
2
=
4
.
1
(
6
)
W
ith
r
eg
ar
d
to
7
:2
co
m
p
r
ess
o
r
s
u
s
in
g
ap
p
r
o
x
i
m
ate
f
u
ll a
d
d
er
,
th
e
f
o
llo
w
i
n
g
i
s
th
e
lo
g
ical
(
7
)
.
=
1
⊕
2
⊕
3
⊕
4
⊕
5
⊕
6
⊕
7
⊕
1
⊕
2
=
1
.
2
1
=
(
2
.
3
)
⊕
(
5
.
6
)
⊕
(
(
4
⊕
5
⊕
6
)
.
7
)
)
2
=
(
5
.
6
)
.
(
(
4
⊕
5
⊕
6
)
.
7
)
)
(
7
)
T
h
e
4
:2
,
5
:
2
,
an
d
7
:2
m
eth
o
d
s
o
f
ap
p
r
o
x
i
m
ate
co
m
p
r
es
s
o
r
d
esig
n
,
p
r
o
v
id
e
an
ef
f
ici
en
t
p
ar
tial
p
r
o
d
u
ct
r
ed
u
ctio
n
in
m
u
ltip
lier
s
.
2
.
3
.
Appro
x
i
m
a
t
e
w
it
h
m
ux
-
ba
s
ed
co
m
pres
s
o
r
T
h
e
ap
p
r
o
x
i
m
ate
w
it
h
m
u
x
-
b
ased
co
m
p
r
ess
o
r
s
ar
e
a
f
u
r
t
h
er
en
h
a
n
ce
m
e
n
t
o
f
ap
p
r
o
x
i
m
at
e
co
m
p
r
es
s
o
r
d
esig
n
s
,
in
co
r
p
o
r
atin
g
m
u
ltip
lex
er
s
to
i
m
p
r
o
v
e
r
eso
u
r
ce
e
f
f
icien
c
y
a
n
d
co
m
p
u
tat
io
n
s
p
ee
d
,
as
d
is
cu
s
s
ed
b
y
E
d
av
o
o
r
et
a
l.
[
5
]
an
d
J
o
o
q
et
a
l.
[1
3]
.
A
s
p
r
esen
ted
b
y
E
d
av
o
o
r
et
a
l.
[
5
]
,
b
y
in
te
g
r
ati
n
g
m
u
ltip
lex
er
s
,
th
e
s
e
co
m
p
r
ess
o
r
s
d
y
n
a
m
icall
y
h
a
n
d
le
th
e
s
elec
tio
n
o
f
i
n
ter
m
ed
iate
r
esu
l
ts
,
f
u
r
th
er
r
ed
u
ci
n
g
lo
g
ic
co
m
p
le
x
it
y
a
n
d
d
ela
y
w
h
ile
m
ai
n
tai
n
i
n
g
a
n
ap
p
r
o
x
i
m
ate
o
u
tp
u
t.
T
h
ese
co
m
p
r
e
s
s
o
r
s
p
r
eser
v
e
th
e
ap
p
r
o
x
im
a
te
f
u
ll
ad
d
er
at
th
eir
co
r
e
b
u
t
ad
d
m
u
ltip
lex
er
lo
g
ic
to
h
an
d
le
s
p
ec
if
ic
i
n
p
u
t
co
n
f
i
g
u
r
atio
n
s
,
allo
w
i
n
g
f
o
r
a
m
o
r
e
ef
f
icien
t
r
ed
u
ctio
n
o
f
p
ar
tial
p
r
o
d
u
cts,
as
d
em
o
n
s
tr
ated
in
b
o
th
E
d
a
v
o
o
r
et
a
l.
[
5
]
an
d
J
o
o
q
et
a
l.
[
1
3
]
.
T
h
e
m
u
ltip
le
x
er
-
b
ased
d
esi
g
n
e
n
s
u
r
es
th
a
t
o
n
l
y
cr
itical
i
n
p
u
t
s
co
n
tr
ib
u
te
to
th
e
co
m
p
u
tatio
n
at
ea
ch
s
tag
e,
m
i
n
i
m
izi
n
g
u
n
n
ec
es
s
ar
y
lo
g
ic
o
p
er
atio
n
s
an
d
en
h
a
n
cin
g
t
h
e
o
v
er
all
ef
f
icie
n
c
y
o
f
t
h
e
co
m
p
r
es
s
o
r
,
as
s
h
o
w
n
i
n
t
h
e
h
ar
d
w
ar
e
e
v
al
u
atio
n
s
b
y
E
d
av
o
o
r
et
a
l.
[
5
]
.
T
h
ese
d
esi
g
n
s
ar
e
i
m
p
le
m
e
n
ted
ac
r
o
s
s
4
:2
,
5
:2
,
an
d
7
:2
co
m
p
r
ess
o
r
ar
ch
itectu
r
es,
ea
ch
o
p
ti
m
ized
f
o
r
s
p
ec
if
ic
i
n
p
u
t
c
o
n
f
i
g
u
r
atio
n
s
.
T
h
is
f
o
llo
w
s
t
h
e
s
a
m
e
p
r
o
ce
d
u
r
e
as
th
e
ap
p
r
o
x
i
m
ate
f
u
ll
ad
d
er
[
1
4
]
co
m
p
r
ess
o
r
tech
n
iq
u
e,
b
u
t
t
h
e
b
o
o
lean
eq
u
atio
n
f
o
r
th
e
i
n
ter
n
al
ar
c
h
i
tectu
r
e
o
f
t
h
is
co
m
p
r
ess
o
r
is
f
u
r
t
h
er
r
ed
u
ce
d
to
p
r
o
d
u
ce
th
e
f
i
n
al
s
u
m
o
u
tp
u
t,
ef
f
ec
tiv
e
l
y
r
ed
u
cin
g
d
ela
y
a
n
d
h
ar
d
w
ar
e
co
m
p
lex
i
t
y
.
T
h
e
lo
g
ical
(
8
)
s
h
o
u
ld
b
e
u
s
ed
f
o
r
4
:2
co
m
p
r
es
s
o
r
s
u
s
i
n
g
an
ap
p
r
o
x
i
m
ate
m
u
x
-
b
a
s
ed
f
u
ll a
d
d
er
,
as sh
o
w
n
b
y
E
d
av
o
o
r
et
a
l.
[
5
]
.
=
(
4
⊕
(
1
⊕
2
⊕
3
)
)
+
(
4
.
(
2
.
3
)
)
=
(
4
.
(
2
.
3
)
)
+
4
(
8
)
Si
m
i
lar
l
y
,
in
th
e
5
:2
co
m
p
r
es
s
o
r
s
,
in
ter
m
ed
iate
r
es
u
lt
s
ar
e
r
o
u
ted
th
r
o
u
g
h
a
m
u
ltip
le
x
er
,
en
s
u
r
i
n
g
ef
f
icien
t
co
m
p
u
tatio
n
o
f
s
u
m
an
d
ca
r
r
y
o
u
tp
u
t
s
(
C
o
u
t1
an
d
C
o
u
t2
)
w
ith
r
ed
u
ce
d
lo
g
ic
d
ep
th
,
as
d
etailed
b
y
E
d
av
o
o
r
et
a
l
.
[
5
]
an
d
J
o
o
q
et
a
l.
[
1
3
]
.
T
h
e
(
9
)
w
ill
b
e
u
s
ed
f
o
r
5
:2
co
m
p
r
ess
o
r
s
f
o
llo
w
i
n
g
t
h
e
m
et
h
o
d
o
lo
g
y
d
is
cu
s
s
ed
in
t
h
ese
w
o
r
k
s
.
=
(
⊕
(
1
⊕
2
⊕
3
)
⊕
4
⊕
5
)
)
+
(
.
(
4
.
5
)
)
=
(
.
(
4
.
5
)
)
+
(
.
(
2
.
3
)
)
(
9
)
T
h
e
m
o
r
e
co
m
p
le
x
d
esig
n
o
f
7
:2
ar
ch
itect
u
r
es,
w
h
ic
h
u
s
e
m
u
ltip
le
ap
p
r
o
x
i
m
ate
f
u
ll
ad
d
er
s
to
co
m
p
u
te
in
ter
m
ed
iate
s
u
m
s
(
s
x
1
an
d
s
x
2
)
,
an
d
m
u
ltip
le
x
er
s
to
d
y
n
a
m
ical
l
y
m
an
a
g
e
th
e
f
l
o
w
o
f
th
e
s
e
r
esu
lt
s
,
p
r
o
d
u
ce
s
th
e
f
in
a
l
s
u
m
an
d
ca
r
r
y
o
u
tp
u
ts
,
as
d
escr
ib
ed
b
y
E
d
av
o
o
r
et
a
l.
[
5
]
an
d
J
o
o
q
et
a
l.
[
1
3
]
.
T
h
is
p
r
o
ce
s
s
is
d
etailed
i
n
t
h
e
lo
g
ica
l
(
1
0
)
.
T
h
e
ar
ch
itect
u
r
e
d
esig
n
o
f
t
h
e
ap
p
r
o
x
i
m
ate
m
u
x
-
b
ased
co
m
p
r
e
s
s
o
r
,
s
h
o
w
n
in
Fig
u
r
e
4
,
p
r
o
v
id
es
a
s
ca
lab
le
s
o
lu
tio
n
th
at
en
s
u
r
e
s
m
i
n
i
m
al
h
ar
d
w
ar
e
co
m
p
lex
it
y
,
m
a
k
i
n
g
it
id
ea
l
f
o
r
lar
g
e
-
s
ca
le
in
p
u
t
co
n
f
i
g
u
r
atio
n
s
,
a
s
d
is
cu
s
s
ed
b
y
J
o
o
q
et
a
l.
[
1
3
]
.
T
h
e
p
r
o
p
o
s
ed
4
:2
ap
p
r
o
x
i
m
ate
co
m
p
r
ess
o
r
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
9
-
4864
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
,
Vo
l.
15
,
No
.
2
,
J
u
l
y
202
6
:
269
-
2
8
0
274
e
m
p
lo
y
s
o
n
e
ap
p
r
o
x
i
m
ate
f
u
l
l
ad
d
er
an
d
tw
o
m
u
ltip
lex
er
s
to
g
en
e
r
ate
th
e
S
u
m
an
d
C
a
r
r
y
o
u
tp
u
ts
(
F
ig
u
r
e
4
(
a)
)
.
T
h
e
p
r
o
p
o
s
ed
5
:2
ap
p
r
o
x
i
m
ate
co
m
p
r
ess
o
r
u
tili
ze
s
th
r
ee
ap
p
r
o
x
i
m
ate
f
u
ll
ad
d
er
s
an
d
tw
o
m
u
lt
ip
lex
er
s
(
Fig
u
r
e
4
(
b
)
)
.
T
h
e
p
r
o
p
o
s
ed
7
:2
ap
p
r
o
x
i
m
ate
co
m
p
r
es
s
o
r
co
n
s
is
ts
o
f
f
o
u
r
ap
p
r
o
x
i
m
ate
f
u
ll
ad
d
er
s
an
d
t
w
o
m
u
ltip
lex
er
s
(
Fig
u
r
e
4
(
c)
)
.
T
h
e
ar
ch
itect
u
r
e
ac
h
ie
v
es
h
i
g
h
er
co
m
p
r
ess
io
n
ca
p
ab
ilit
y
w
h
ile
m
ai
n
tain
in
g
lo
w
p
o
w
er
co
n
s
u
m
p
tio
n
an
d
h
ig
h
co
m
p
u
tatio
n
al
ef
f
icie
n
c
y
.
A
cr
o
s
s
all
t
h
r
ee
ar
ch
itect
u
r
es
,
th
e
u
s
e
o
f
m
u
ltip
lex
er
s
r
ed
u
ce
s
t
h
e
cr
itical
p
ath
d
ela
y
a
n
d
p
o
w
er
co
n
s
u
m
p
tio
n
,
w
h
ile
t
h
e
ap
p
r
o
x
i
m
ate
lo
g
ic
m
ai
n
tai
n
s
ac
ce
p
tab
le
ac
cu
r
ac
y
le
v
el
s
f
o
r
er
r
o
r
-
to
ler
an
t
ap
p
licatio
n
s
,
as d
e
m
o
n
s
tr
ated
b
y
b
o
th
E
d
av
o
o
r
et
a
l.
[
5
]
an
d
J
o
o
q
et
a
l.
[
1
3
]
.
=
(
⊕
(
(
(
1
⊕
2
⊕
3
)
⊕
4
⊕
5
)
⊕
6
⊕
7
)
)
+
(
.
(
6
.
7
)
)
=
(
.
(
6
.
7
)
)
+
(
.
(
4
.
5
)
)
(
1
0
)
(
a)
(
b
)
(
c)
Fig
u
r
e
4
.
P
r
o
p
o
s
ed
m
u
l
tip
lex
e
r
-
b
ased
ap
p
r
o
x
im
a
te
co
m
p
r
ess
o
r
ar
ch
itectu
r
es: (
a)
4
:2
ap
p
r
o
x
i
m
ate
m
u
x
b
ased
co
m
p
r
es
s
o
r
,
(
b
)
5
:
2
ap
p
r
o
x
i
m
ate
m
u
x
b
ased
co
m
p
r
ess
o
r
,
an
d
(
c)
7
:
2
ap
p
r
o
x
i
m
ate
m
u
x
b
as
ed
co
m
p
r
ess
o
r
2
.
4
.
M
a
j
o
rit
y
g
a
t
e
-
ba
s
ed
co
m
p
re
s
s
o
r
T
h
e
p
r
o
p
o
s
ed
d
esig
n
o
f
m
aj
o
r
ity
g
ate
-
b
ased
co
m
p
r
ess
o
r
s
p
r
esen
ts
an
e
f
f
ec
ti
v
e
a
n
d
o
p
ti
m
al
alter
n
ati
v
e
to
p
r
ev
io
u
s
co
m
p
r
e
s
s
o
r
d
esig
n
s
.
T
h
i
s
d
esi
g
n
u
t
ili
ze
s
a
v
er
y
s
m
al
l n
u
m
b
er
o
f
3
-
i
n
p
u
t
m
aj
o
r
it
y
lo
g
ic
g
ates
in
th
e
co
m
p
r
ess
o
r
d
esi
g
n
,
w
h
ic
h
r
ed
u
ce
s
t
h
e
to
tal
l
o
g
ic
s
ize
a
n
d
p
o
w
er
co
n
s
u
m
p
tio
n
,
as
s
h
o
w
n
b
y
R
ed
d
y
et
a
l.
[
1
2
]
an
d
J
o
o
q
e
t
a
l
.
[
1
3
]
.
T
h
e
ex
p
r
ess
io
n
o
f
m
aj
o
r
ity
g
ate
lo
g
ic
w
it
h
m
u
lt
i
-
i
n
p
u
t
ex
p
r
es
s
io
n
is
(
1
1
)
.
(
,
,
)
=
(
.
)
+
(
.
)
+
(
.
)
(
1
1
)
T
h
e
o
p
e
r
atio
n
s
o
f
4
:2
co
m
p
r
es
s
o
r
r
eq
u
ir
ed
t
w
o
m
aj
o
r
ity
lo
g
i
c
g
ate
is
e
x
p
r
ess
ed
in
(
1
2
)
.
=
(
(
3
.
4
)
+
(
4
.
)
+
(
3
.
)
)
.
(
1
+
2
)
+
(
1
.
2
)
=
4
=
3
(
1
2
)
T
h
e
5
:2
co
m
p
r
ess
o
r
u
tili
ze
s
t
h
r
ee
la
y
er
o
f
m
aj
o
r
it
y
lo
g
ic
g
ates,
an
d
th
e
e
x
p
r
ess
io
n
o
f
5
in
p
u
t
a
n
d
2
o
u
tp
u
t
b
ased
co
m
p
r
ess
o
r
u
s
i
n
g
m
aj
o
r
it
y
lo
g
ic
i
s
(
1
3
)
.
=
(
(
4
(
5
+
1
)
+
(
5
.
1
)
)
.
(
3
+
2
)
+
(
3
.
2
)
.
(
2
+
1
)
+
(
2
.
1
)
=
5
1
=
4
2
=
3
(
1
3
)
Fu
r
t
h
er
m
o
r
e,
th
e
lo
g
ical
eq
u
at
io
n
o
f
t
h
e
7
:2
co
m
p
r
es
s
o
r
d
esi
g
n
th
at
m
a
k
e
s
u
s
e
o
f
th
e
m
aj
o
r
it
y
lo
g
ic
ap
p
r
o
ac
h
an
d
h
as
th
e
ca
p
ac
it
y
to
ac
co
m
m
o
d
ate
s
ev
e
n
p
r
i
m
ar
y
i
n
p
u
ts
an
d
t
w
o
o
u
tp
u
t
s
is
d
e
m
o
n
s
tr
ated
in
(
1
4
)
,
as
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
I
SS
N:
2089
-
4864
Hig
h
-
p
erfo
r
ma
n
ce
a
p
p
r
o
xima
te
MAC mu
ltip
lier
u
s
in
g
ma
jo
r
ity
…
(
S
elva
r
a
s
a
n
R
a
d
h
a
kris
h
n
a
n
)
275
d
is
cu
s
s
ed
b
y
R
ed
d
y
et
a
l.
[
1
2
]
an
d
J
o
o
q
et
a
l.
[
1
3
]
.
I
n
th
is
p
ar
ticu
lar
in
s
tan
ce
,
it
m
a
k
es
u
s
e
o
f
f
o
u
r
la
y
er
s
o
f
m
aj
o
r
ity
g
ates.
T
h
e
p
r
o
p
o
s
ed
m
aj
o
r
ity
-
lo
g
ic
b
a
s
ed
co
m
p
r
es
s
o
r
ar
ch
itectu
r
es
f
o
r
p
ar
tial
p
r
o
d
u
ct
r
ed
u
ctio
n
in
th
e
ap
p
r
o
x
i
m
ate
m
u
ltip
lier
d
esig
n
ar
e
s
h
o
w
n
i
n
Fi
g
u
r
e
5
.
T
h
ese
co
m
p
r
es
s
o
r
s
tr
u
ct
u
r
es
ar
e
th
e
b
asic
b
u
ild
i
n
g
b
lo
ck
s
o
f
t
h
e
r
ed
u
ctio
n
s
ta
g
e
w
h
ic
h
u
s
e
m
aj
o
r
ity
-
g
ate
lo
g
ic
to
ac
h
iev
e
t
h
e
cir
c
u
it
i
m
p
le
m
en
tatio
n
a
n
d
lo
w
er
h
ar
d
w
ar
e
co
m
p
le
x
it
y
,
p
o
w
er
co
n
s
u
m
p
tio
n
an
d
co
m
p
u
tat
io
n
e
f
f
ic
ien
c
y
.
T
h
e
s
tr
u
ct
u
r
e
o
f
t
h
e
p
r
o
p
o
s
ed
4
:2
m
aj
o
r
ity
-
lo
g
ic
co
m
p
r
ess
o
r
is
illu
s
tr
ated
in
Fi
g
u
r
e
5
(
a)
.
T
h
e
in
p
u
t
b
its
g
o
th
r
o
u
g
h
a
s
m
all
m
aj
o
r
ity
-
g
ate
n
et
w
o
r
k
w
h
er
e
th
e
s
u
m
a
n
d
ca
r
r
y
o
u
tp
u
ts
ar
e
ca
lcu
lated
w
it
h
m
in
i
m
u
m
lo
g
ic
ele
m
e
n
t
s
.
T
h
e
p
r
o
p
o
s
ed
5
:
2
m
aj
o
r
ity
-
lo
g
ic
co
m
p
r
ess
o
r
ca
n
b
e
s
ee
n
as
an
e
x
p
an
s
io
n
o
f
th
e
m
aj
o
r
it
y
-
g
ate
co
n
ce
p
t
b
y
o
n
e
m
o
r
e
in
p
u
t
b
it
an
d
is
d
ep
icted
in
Fig
u
r
e
5
(
b
)
.
B
y
u
s
i
n
g
th
e
ca
s
ca
d
ed
m
aj
o
r
i
t
y
-
g
ate
s
s
tr
u
ct
u
r
e
,
w
e
g
e
t
th
e
ef
f
ec
tiv
e
r
ea
lizatio
n
o
f
o
u
tp
u
t
b
its
w
it
h
t
h
e
s
a
m
e
s
i
m
p
lic
it
y
o
f
i
m
p
le
m
e
n
tatio
n
an
d
s
h
o
r
t
cr
itical
p
at
h
.
Fi
g
u
r
e
5
(
c)
illu
s
tr
ate
s
t
h
e
7
:2
m
aj
o
r
it
y
-
lo
g
ic
co
m
p
r
es
s
o
rs
,
w
h
ic
h
ca
n
o
p
er
ate
w
it
h
i
n
cr
ea
s
ed
in
p
u
t
b
its
n
u
m
b
er
,
p
r
o
v
id
in
g
f
u
r
t
h
er
p
ar
tial
p
r
o
d
u
cts co
m
p
r
es
s
io
n
.
M
a
j
o
r
i
t
y
x
3
x
4
c
i
n
M
a
j
o
r
i
t
y
x
1
x
2
S
u
m
C
a
r
r
y
C
o
u
t
M
a
j
o
r
i
t
y
x
5
x
4
C
i
n
1
M
a
j
o
r
i
t
y
x
3
C
i
n
2
S
u
m
C
a
r
r
y
C
o
u
t
1
C
o
u
t
2
M
a
j
o
r
i
t
y
x
1
x
2
M
a
j
o
r
i
t
y
x
7
x
6
C
i
n
1
M
a
j
o
r
i
t
y
x
5
C
i
n
2
S
u
m
C
a
r
r
y
C
o
u
t
2
C
o
u
t
1
M
a
j
o
r
i
t
y
x
4
x
3
M
a
j
o
r
i
t
y
x
1
x
2
(
a
)
(
c
)
(
b
)
(
a)
(
b
)
(
c)
Fig
u
r
e
5
.
P
r
o
p
o
s
ed
m
aj
o
r
ity
-
l
o
g
ic
-
b
ased
co
m
p
r
es
s
o
r
s
(
a)
4
:
2
m
aj
o
r
it
y
-
lo
g
ic
co
m
p
r
es
s
o
r
,
(
b
)
5
:2
m
aj
o
r
ity
-
lo
g
ic
co
m
p
r
ess
o
r
,
an
d
(
c)
7
:2
m
aj
o
r
ity
-
lo
g
ic
co
m
p
r
ess
o
r
3.
E
XAC
T
M
UL
T
I
P
L
I
E
R
DE
SI
G
N
US
I
N
G
AP
P
RO
XIM
AT
E
A
DDE
R
ARCH
I
T
E
CT
URE
T
h
is
s
ec
tio
n
p
r
esen
ts
t
h
e
m
et
h
o
d
o
lo
g
ical
f
r
a
m
e
w
o
r
k
ad
o
p
te
d
f
o
r
th
e
ex
ac
t
m
u
ltip
lier
d
esi
g
n
u
s
i
n
g
a
n
ap
p
r
o
x
im
a
te
ad
d
er
ar
ch
itectu
r
e
an
d
p
r
o
p
o
s
ed
A
p
p
r
o
x
i
m
ate
m
u
ltip
lier
d
esig
n
.
T
h
e
ex
ac
t
m
et
h
o
d
o
f
m
u
ltip
lier
d
esig
n
,
w
h
ich
is
co
m
m
o
n
l
y
u
s
ed
in
p
r
ec
is
io
n
-
cr
itical
ap
p
licatio
n
s
s
u
c
h
a
s
i
m
ag
e
an
d
s
ig
n
al
p
r
o
ce
s
s
in
g
,
ten
d
s
to
o
cc
u
p
y
a
s
i
g
n
i
f
ica
n
tl
y
lar
g
er
lo
g
ic
ar
ea
,
as
r
ep
o
r
te
d
b
y
Mo
m
en
i
et
a
l
.
[
6
]
.
T
r
ad
itio
n
al
m
u
lt
ip
lier
s
s
u
c
h
as
W
allac
e
tr
ee
,
Dad
d
a,
Ved
ic,
a
n
d
ar
r
ay
m
u
ltip
lier
s
also
u
s
e
t
h
is
e
x
ac
t
co
m
p
u
tatio
n
s
eq
u
e
n
ce
.
A
co
m
m
o
n
is
s
u
e
w
it
h
ex
ac
t
m
u
ltip
licatio
n
is
in
cr
ea
s
ed
lo
g
ic
s
ize
an
d
p
o
w
er
co
n
s
u
m
p
tio
n
.
Fo
r
th
at
r
ea
s
o
n
,
r
ec
en
t
r
esear
ch
h
as
f
o
cu
s
ed
o
n
v
ar
io
u
s
ap
p
r
o
x
i
m
atio
n
-
b
ased
m
u
ltip
licatio
n
tech
n
iq
u
es
[
1
]
,
[
4
]
,
[
1
5
]
,
[
1
6
]
.
R
ec
en
t
s
t
u
d
ies,
in
cl
u
d
in
g
th
o
s
e
b
y
[
4
]
,
[
1
5
]
,
[
1
]
,
[
1
6
]
h
av
e
d
e
m
o
n
s
tr
ated
th
e
ef
f
ec
ti
v
en
e
s
s
o
f
ap
p
r
o
x
i
m
atio
n
tech
n
iq
u
e
s
as
f
ea
s
ib
le
r
ep
lace
m
en
ts
to
r
ed
u
ce
lo
g
ic
ar
ea
,
p
o
w
er
co
n
s
u
m
p
tio
n
,
an
d
d
ela
y
.
T
o
ad
d
r
es
s
th
ese
ch
al
len
g
es,
t
h
e
p
r
o
p
o
s
ed
tech
n
iq
u
e
in
t
h
i
s
s
t
u
d
y
i
n
tr
o
d
u
ce
s
a
n
ap
p
r
o
x
i
m
a
te
f
u
ll
ad
d
er
[
1
4
]
tech
n
iq
u
e,
it’
s
in
s
p
ir
ed
b
y
t
h
e
w
o
r
k
o
f
Stro
llo
et
a
l.
[
7
]
,
w
h
i
ch
p
r
o
d
u
ce
s
th
e
s
a
m
e
s
u
m
o
u
t
p
u
t
as
a
co
n
v
en
tio
n
al
f
u
ll
ad
d
er
b
u
t
s
i
m
p
lif
ie
s
th
e
ca
r
r
y
g
e
n
er
atio
n
.
Fi
g
u
r
e
6
s
h
o
w
s
th
e
tr
ad
itio
n
al
d
esi
g
n
o
f
Dad
d
a
8
×
8
m
u
lt
ip
lier
i
ts
h
ig
h
li
g
h
ts
t
h
ese
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
9
-
4864
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
,
Vo
l.
15
,
No
.
2
,
J
u
l
y
202
6
:
269
-
2
8
0
276
i
m
p
r
o
v
e
m
en
t
s
.
W
h
ile
t
h
i
s
d
es
ig
n
tak
e
n
f
r
o
m
t
h
e
co
n
v
en
tio
n
al
Dad
d
a
m
u
lt
ip
lier
ar
ch
itect
u
r
e,
its
p
r
o
d
u
ce
s
6
4
p
ar
tial
p
r
o
d
u
cts
an
d
t
y
p
icall
y
r
eq
u
ir
es
f
o
u
r
r
ed
u
ct
io
n
s
ta
g
e
s
,
th
e
p
r
o
p
o
s
ed
d
esig
n
m
i
n
i
m
izes
th
e
n
u
m
b
er
o
f
s
tag
e
s
b
y
e
m
p
lo
y
i
n
g
a
co
m
b
i
n
atio
n
o
f
h
al
f
ad
d
er
s
an
d
ap
p
r
o
x
i
m
ate
f
u
l
l
ad
d
er
s
,
ef
f
ec
tiv
el
y
r
ed
u
ci
n
g
in
ter
n
a
l
co
m
p
le
x
it
y
.
A
l
m
o
s
t
F
u
l
l
A
d
d
e
r
Ha
l
f
&
F
u
l
l
A
d
d
er
A
c
c
u
m
u
l
a
t
o
r
2
N
+
1
F
i
n
a
l
O
u
t
p
u
t
Fig
u
r
e
6
.
8
×
8
Dad
d
a
m
u
ltip
lic
atio
n
u
s
in
g
ap
p
r
o
x
i
m
ate
f
u
ll a
d
d
er
3
.
1
.
P
r
o
po
s
ed
a
pp
ro
x
i
m
a
t
e
m
u
lt
ipl
ier
des
ig
n u
s
ing
4
:
2
,
5
:
2
,
a
nd
7
:
2
co
m
pre
s
s
o
rs a
rc
hite
ct
ure
T
h
e
p
r
o
p
o
s
ed
a
p
p
r
o
ac
h
o
f
th
is
r
esear
ch
u
s
i
n
g
ap
p
r
o
x
i
m
at
e
m
u
ltip
lier
ar
ch
itect
u
r
e
is
s
p
ec
if
icall
y
d
esig
n
ed
to
r
ed
u
ce
th
e
n
u
m
b
er
o
f
s
tag
es
i
n
t
h
e
m
u
ltip
lica
tio
n
p
r
o
ce
s
s
.
T
o
ac
h
iev
e
th
is
,
4
:2
,
7
:2
,
an
d
5
:2
m
aj
o
r
ity
lo
g
ic
co
m
p
r
es
s
o
r
s
ar
e
ad
d
itio
n
all
y
i
n
te
g
r
ate
d
in
to
t
h
e
m
u
lt
ip
lier
d
esig
n
ar
ch
itect
u
r
e,
en
ab
lin
g
a
h
i
g
h
lev
el
o
f
co
m
p
r
es
s
io
n
o
f
p
ar
ti
al
p
r
o
d
u
cts
i
n
f
e
w
er
s
tep
s
,
a
s
d
is
c
u
s
s
ed
b
y
[
1
0
]
,
[
1
1
]
.
T
h
is
m
ax
i
m
u
m
i
n
p
u
t
r
ed
u
ctio
n
en
h
a
n
ce
s
b
o
th
s
p
e
ed
an
d
ef
f
ic
ien
c
y
i
n
th
r
ee
l
ev
els
o
f
m
u
ltip
lier
d
esi
g
n
th
at
co
m
b
in
e
e
x
ac
t,
ap
p
r
o
x
im
a
te,
an
d
tr
an
s
d
u
ce
tech
n
iq
u
es.
C
o
m
p
ar
ed
to
th
e
tr
ad
itio
n
al
d
esig
n
o
f
m
u
lt
ip
lier
s
u
s
i
n
g
o
n
l
y
ap
p
r
o
x
im
a
te
4
:2
co
m
p
r
es
s
o
r
ar
ch
itectu
r
e,
t
h
e
p
r
o
p
o
s
ed
d
esig
n
s
i
g
n
if
ica
n
tl
y
r
ed
u
ce
s
th
e
lo
g
ic
s
ize
an
d
co
m
p
u
tatio
n
al
co
m
p
le
x
it
y
t
h
r
o
u
g
h
t
h
is
in
n
o
v
at
iv
e
ap
p
r
o
ac
h
,
as
d
escr
ib
ed
b
y
E
d
a
v
o
o
r
et
a
l.
[
5
]
an
d
f
u
r
t
h
er
v
alid
ated
in
R
ed
d
y
et
a
l.
[
1
2
]
.
No
tab
ly
,
as
s
h
o
w
n
in
Fi
g
u
r
e
7
,
th
e
m
u
ltip
lier
r
eq
u
ir
es
o
n
l
y
t
w
o
s
ta
g
es
i
n
th
e
8
×
8
m
u
ltip
licatio
n
ar
ch
itect
u
r
e.
Sp
ec
if
icall
y
,
i
n
t
h
e
p
r
o
p
o
s
e
d
ap
p
r
o
x
i
m
ate
m
u
lt
ip
lier
,
s
ta
g
e
1
is
ex
cl
u
s
i
v
el
y
h
an
d
led
b
y
7
:2
an
d
5
:2
co
m
p
r
ess
o
r
s
,
w
h
ich
r
ed
u
ce
t
h
e
n
u
m
b
er
o
f
p
ar
tial
p
r
o
d
u
cts;
as
a
r
e
s
u
lt,
t
h
e
n
u
m
b
er
o
f
in
ter
m
ed
iate
s
ig
n
al
s
p
ass
ed
to
s
tag
e
2
is
m
in
i
m
a
l.
Stag
e
2
f
u
r
th
er
co
m
p
r
ess
e
s
th
e
d
ata
u
s
i
n
g
e
x
ac
t
4
:2
co
m
p
r
es
s
o
r
s
,
f
u
ll
ad
d
er
s
,
an
d
h
al
f
ad
d
er
s
,
m
ai
n
tai
n
i
n
g
ac
cu
r
ac
y
w
h
er
e
it
is
m
o
s
t
cr
itical.
T
h
is
co
m
b
i
n
atio
n
o
f
o
p
tim
ized
tech
n
iq
u
e
s
en
s
u
r
es
a
co
m
p
ac
t d
esi
g
n
w
it
h
lo
w
la
t
en
c
y
a
n
d
en
er
g
y
e
f
f
icien
c
y
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
I
SS
N:
2089
-
4864
Hig
h
-
p
erfo
r
ma
n
ce
a
p
p
r
o
xima
te
MAC mu
ltip
lier
u
s
in
g
ma
jo
r
ity
…
(
S
elva
r
a
s
a
n
R
a
d
h
a
kris
h
n
a
n
)
277
4
:
2
C
o
m
p
r
e
s
s
o
r
Ha
l
f
&
F
u
l
l
A
d
d
e
r
E
x
a
c
t
C
o
m
p
r
e
s
s
o
r
E
x
a
c
t
A
p
p
r
o
x
i
m
a
t
e
T
r
a
n
s
d
u
c
e
5
:
2
C
o
m
p
r
e
s
s
o
r
7
:
2
C
o
m
p
r
e
s
s
o
r
A
c
c
u
m
u
l
a
t
o
r
2
N
+
1
F
i
n
a
l
O
u
t
p
u
t
Fig
u
r
e
7
.
P
r
o
p
o
s
ed
h
ig
h
er
b
it
c
o
m
p
r
es
s
o
r
b
ased
8
×
8
Dad
d
a
m
u
ltip
lier
4.
RE
SU
L
T
S
T
h
e
p
e
r
f
o
r
m
a
n
ce
an
d
er
r
o
r
an
al
y
s
i
s
o
f
th
e
p
r
o
p
o
s
ed
ap
p
r
o
x
im
ate
m
u
ltip
lier
d
esig
n
,
i
m
p
le
m
en
ted
in
Ver
ilo
g
h
ar
d
w
ar
e
d
escr
ip
tio
n
lan
g
u
a
g
e
(
HD
L
)
,
w
as
s
y
n
t
h
es
ized
u
s
in
g
Xili
n
x
Ver
tex
-
5
F
P
GA
an
d
co
m
p
ar
ed
f
o
r
u
ti
lizatio
n
r
esu
lts
s
u
ch
as
s
lice
r
eg
is
ter
s
,
s
lice
lo
o
k
u
p
ta
b
le
s
(
L
UT
s
)
,
o
cc
u
p
ied
s
lice
s
,
b
o
n
d
ed
I
OB
,
d
elay
,
an
d
p
o
w
er
.
T
h
e
co
m
p
ar
ativ
e
p
er
f
o
r
m
an
ce
r
es
u
lts
o
f
th
e
MA
C
-
b
ased
8
×8
a
p
p
r
o
x
i
m
at
e
Dad
d
a
m
u
ltip
lier
d
esig
n
e
m
p
lo
y
in
g
ap
p
r
o
x
i
m
a
t
e
co
m
p
r
es
s
o
r
s
ar
e
p
r
esen
ted
i
n
T
ab
le
1
.
I
n
ad
d
itio
n
,
er
r
o
r
m
etr
ics
—
in
c
lu
d
i
n
g
an
al
y
s
is
w
it
h
t
h
e
v
er
if
ica
tio
n
m
eth
o
d
i
n
s
i
m
u
lat
io
n
u
s
i
n
g
m
o
d
elsi
m
—
w
er
e
e
m
p
lo
y
ed
;
th
e
p
ar
a
m
eter
s
an
al
y
ze
d
i
n
cl
u
d
e
E
R
%,
ME
D,
NM
E
D,
an
d
MRED,
as s
u
p
p
o
r
ted
b
y
th
e
w
o
r
k
o
f
[
3
]
,
[
8
]
.
I
n
a
v
ital
r
o
le
o
f
o
p
ti
m
izi
n
g
d
ig
ital
ar
ith
m
etic
cir
cu
its
b
y
r
ed
u
c
in
g
in
ter
m
ed
iate
r
es
u
lts
d
u
r
in
g
co
m
p
u
tatio
n
s
,
v
ar
io
u
s
co
m
p
r
ess
o
r
d
esig
n
s
o
f
th
e
p
r
o
p
o
s
ed
m
et
h
o
d
o
lo
g
y
—
s
u
c
h
as
ex
ac
t,
ap
p
r
o
x
im
a
te,
th
o
s
e
in
co
r
p
o
r
atin
g
m
u
l
tip
lex
er
s
,
a
n
d
m
aj
o
r
ity
g
ate
s
—
o
f
f
er
a
tr
ad
eo
f
f
b
et
w
ee
n
u
tili
za
t
io
n
an
d
p
er
f
o
r
m
an
ce
.
T
ab
le
2
s
h
o
w
s
t
h
e
cr
itical
p
er
f
o
r
m
a
n
ce
m
etr
ics
ac
r
o
s
s
d
if
f
er
e
n
t
co
m
p
r
es
s
o
r
d
esig
n
s
to
ev
alu
ate
th
eir
ef
f
ec
tiv
e
n
e
s
s
.
T
h
ese
r
esu
lts
s
h
o
w
th
at
ap
p
r
o
x
i
m
ate
co
m
p
r
es
s
o
r
s
,
p
ar
ticu
lar
l
y
m
u
ltip
le
x
er
-
an
d
m
aj
o
r
it
y
lo
g
ic
b
ased
ar
ch
itectu
r
e
s
,
ac
h
iev
e
n
o
tab
le
r
ed
u
ctio
n
s
i
n
L
UT
u
s
ag
e,
o
cc
u
p
ied
s
lices,
a
n
d
d
ela
y
,
as
d
e
m
o
n
s
tr
ated
b
y
E
d
av
o
o
r
et
a
l.
[
5
]
an
d
R
ed
d
y
et
a
l.
[
1
2
]
.
E
r
r
o
r
an
aly
s
is
e
s
ti
m
ati
n
g
th
e
er
r
o
r
an
aly
s
is
a
n
d
it
s
r
eliab
ilit
y
u
s
i
n
g
ap
p
r
o
x
i
m
ate
m
u
lti
p
lier
s
,
th
i
s
w
o
r
k
q
u
a
n
ti
f
ies
th
e
er
r
o
r
s
in
tr
o
d
u
ce
d
b
y
ap
p
r
o
x
i
m
ate
m
et
h
o
d
s
,
w
h
ic
h
i
s
es
s
e
n
tial
to
e
n
s
u
r
e
th
e
d
es
ig
n
m
ee
t
s
th
e
ac
cu
r
ac
y
r
eq
u
ir
e
m
en
t
s
o
f
C
NN
ap
p
licatio
n
s
,
as
d
e
m
o
n
s
t
r
ated
b
y
Ki
m
et
a
l.
[
1
]
an
d
A
n
s
ar
i
et
a
l.
[
3
]
.
Key
er
r
o
r
m
etr
ics
—
E
R
,
ME
D,
NM
E
D,
an
d
MRED
—
p
r
o
v
id
e
a
s
tr
u
ctu
r
ed
f
r
a
m
e
w
o
r
k
to
ass
ess
an
d
co
m
p
ar
e
th
e
ef
f
ec
tiv
e
n
e
s
s
o
f
v
ar
io
u
s
ap
p
r
o
x
i
m
ate
co
m
p
u
t
in
g
ar
ch
itect
u
r
es,
as
d
is
cu
s
s
ed
b
y
An
s
ar
i
et
a
l.
[
3
]
an
d
J
ian
g
et
a
l.
[
8
]
.
T
h
ese
er
r
o
r
m
e
tr
ics
ar
e
ch
o
s
e
n
f
o
r
th
e
ap
p
r
o
x
i
m
ati
o
n
s
tr
ate
g
y
,
w
h
ic
h
tr
ad
es
o
f
f
b
et
w
ee
n
e
f
f
icien
c
y
an
d
p
r
ec
is
io
n
,
an
d
d
ir
ec
tly
af
f
ec
t
th
e
s
elec
tio
n
o
f
ap
p
r
o
x
im
ate
v
al
u
es
i
n
s
p
ec
if
ic
ca
s
es,
esp
ec
iall
y
in
er
r
o
r
r
esil
ien
t a
p
p
licatio
n
s
s
u
c
h
as i
m
ag
e
p
r
o
ce
s
s
i
n
g
an
d
I
o
T
s
y
s
t
e
m
s
,
as r
ep
o
r
ted
b
y
Ki
m
et
a
l.
[
1
]
an
d
Ham
m
ad
et
a
l.
[
2
]
.
T
h
is
w
o
r
k
is
ab
le
to
m
in
i
m
ize
er
r
o
r
s
w
h
ile
ac
h
ie
v
in
g
g
ai
n
s
in
p
o
w
er
,
ar
ea
,
an
d
s
p
ee
d
.
Fo
r
in
s
tan
ce
,
a
h
ig
h
er
E
R
%
i
m
p
lie
s
f
r
eq
u
e
n
t
co
m
p
u
tatio
n
a
l
d
ev
iatio
n
a
n
d
m
a
y
b
e
u
n
s
u
itab
le
f
o
r
p
r
ec
is
io
n
-
cr
itical
tas
k
s
,
w
h
ile
a
lo
w
er
E
R
%
in
d
icate
s
g
r
ea
ter
r
eliab
ilit
y
an
d
s
tab
ilit
y
,
ev
e
n
f
o
r
ap
p
r
o
x
i
m
ate
d
esi
g
n
s
,
a
s
an
al
y
ze
d
b
y
E
d
av
o
o
r
et
a
l.
[
5
]
an
d
J
ian
g
e
t
a
l.
[
8
]
.
T
h
e
an
al
y
s
i
s
o
f
t
h
e
e
r
r
o
r
ass
ess
m
en
t
f
r
a
m
e
w
o
r
k
is
w
id
el
y
ad
o
p
ted
in
r
ec
en
t
ap
p
r
o
x
i
m
ate
r
esear
c
h
,
i
n
clu
d
i
n
g
t
h
e
w
o
r
k
s
o
f
Ha
m
m
a
d
et
a
l.
[
2
]
,
A
n
s
ar
i
et
a
l.
[
3
]
,
an
d
J
ian
g
e
t
a
l.
[
8
]
.
T
h
e
f
r
eq
u
en
c
y
er
r
o
r
an
al
y
s
i
s
o
f
E
R
% is
(
1
5
)
,
w
ith
t
h
e
n
u
m
b
e
r
o
f
in
co
r
r
ec
t v
er
s
u
s
th
e
to
tal
n
u
m
b
er
o
f
o
u
tp
u
ts
.
T
ab
le
3
an
al
y
ze
s
th
e
e
v
alu
a
ti
o
n
o
f
er
r
o
r
m
etr
ics
ac
r
o
s
s
m
u
ltip
le
ap
p
r
o
x
i
m
ate
m
u
ltip
lier
d
esig
n
i
n
w
h
ic
h
h
i
g
h
lig
h
t
s
th
e
in
h
er
e
n
t
tr
ad
eo
f
f
b
et
w
ee
n
ac
c
u
r
ac
y
a
n
d
ap
p
r
o
x
i
m
ate
v
al
u
es.
I
n
th
is
ar
ch
itectu
r
e
s
ig
n
i
f
ica
n
tl
y
i
m
p
r
o
v
e
t
h
e
er
r
o
r
b
eh
av
io
r
an
d
o
b
s
er
v
ed
th
r
o
u
g
h
t
h
e
i
n
te
g
r
atio
n
o
f
ad
v
an
ce
d
tech
n
iq
u
es,
in
cl
u
d
in
g
m
aj
o
r
it
y
lo
g
ic
co
m
p
r
ess
o
r
lo
g
ic.
T
h
e
co
n
v
en
tio
n
al
lo
g
ic
d
esig
n
ex
h
ib
its
a
h
i
g
h
er
lev
el
o
f
er
r
o
r
at
6
.
1
%,
in
d
icatin
g
a
h
i
g
h
er
r
an
g
e
o
f
ex
ac
t c
o
m
p
u
tatio
n
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
9
-
4864
I
n
t J
R
ec
o
n
f
i
g
u
r
ab
le
&
E
m
b
ed
d
ed
Sy
s
t
,
Vo
l.
15
,
No
.
2
,
J
u
l
y
202
6
:
269
-
2
8
0
278
T
ab
le
1
.
P
er
f
o
r
m
a
n
ce
an
al
y
s
is
o
f
MA
C
b
ased
8
×
8
ap
p
r
o
x
i
m
ate
Dad
d
a
m
u
ltip
lier
d
esi
g
n
u
s
in
g
ap
p
r
o
x
i
m
ate
co
m
p
r
es
s
o
rs
D
e
si
g
n
L
U
T
FF
D
e
l
a
y
(
n
s)
P
o
w
e
r
(
W
)
C
o
n
v
e
n
t
i
o
n
a
l
d
e
si
g
n
[
1
7
]
2
3
8
1
9
8
1
5
.
4
6
3
0
.
6
1
3
4
:
2
a
p
p
r
o
x
i
mat
e
F
A
1
2
4
51
4
.
6
4
8
0
.
6
1
6
4
:
2
a
p
p
r
o
x
i
mat
e
F
A
-
mu
x
1
3
5
62
4
.
6
4
8
0
.
6
1
7
4
:
2
a
p
p
r
o
x
i
mat
e
F
A
-
maj
o
r
i
t
y
c
o
mp
r
e
sso
r
1
2
2
50
4
.
6
4
8
0
.
6
1
2
P
r
o
p
o
se
d
c
o
mp
r
e
sso
r
u
si
n
g
a
p
p
r
o
x
i
mat
e
f
u
l
l
a
d
d
e
r
1
1
1
50
2
.
5
2
5
0
.
6
1
4
P
r
o
p
o
se
d
c
o
mp
r
e
sso
r
u
si
n
g
a
p
p
r
o
x
i
mat
e
w
i
t
h
mu
x
1
0
1
47
2
.
5
2
5
0
.
6
1
2
P
r
o
p
o
se
d
c
o
mp
r
e
sso
r
u
si
n
g
m
a
j
o
r
i
t
y
c
o
mp
r
e
sso
r
99
51
2
.
5
2
5
0
.
6
0
6
T
ab
le
2
.
C
o
m
p
ar
is
o
n
s
an
d
an
a
l
y
s
i
s
o
f
o
n
l
y
co
m
p
r
es
s
o
rs
D
e
si
g
n
L
U
T
S
l
i
c
e
s
I
O
B
s
D
e
l
a
y
(
n
s)
F
a
n
-
o
u
t
Ex
a
c
t
4
:
2
6
4
8
5
.
7
7
9
1
.
7
3
Ex
a
c
t
5
:
2
9
5
11
6
.
7
8
2
1
.
7
5
Ex
a
c
t
7
:
2
15
6
13
6
.
8
5
7
1
.
8
3
A
p
p
r
o
x
i
mat
e
4
:
2
4
3
8
4
.
3
8
6
1
.
4
4
A
p
p
r
o
x
i
mat
e
5
:
2
3
2
11
5
.
1
8
0
1
.
1
7
A
p
p
r
o
x
i
mat
e
7
:
2
4
3
13
5
.
1
8
0
1
.
1
2
A
p
p
r
o
x
i
mat
e
w
i
t
h
m
u
x
4
:
2
3
2
6
5
.
3
7
7
1
.
5
7
A
p
p
r
o
x
i
mat
e
w
i
t
h
m
u
x
5
:
2
3
1
8
6
.
5
6
0
1
.
4
0
A
p
p
r
o
x
i
mat
e
w
i
t
h
m
u
x
7
:
2
4
2
10
7
.
7
4
3
1
.
5
0
M
a
j
o
r
i
t
y
4
:
2
1
1
8
4
.
3
0
7
1
.
3
3
M
a
j
o
r
i
t
y
5
:
2
2
1
11
4
.
9
9
2
1
.
3
3
M
a
j
o
r
i
t
y
7
:
2
2
1
13
5
.
6
7
8
1
.
2
7
T
ab
le
3
.
E
r
r
o
r
an
aly
s
is
o
f
ap
p
r
o
x
i
m
ate
m
u
ltip
lier
D
e
si
g
n
ER
%
M
ED
N
M
ED
M
R
ED
C
o
n
v
e
n
t
i
o
n
a
l
a
p
p
r
o
x
i
mat
e
f
u
l
l
a
d
d
e
r
6
.
1
1
.
1
1
.
4
×
10
-
3
2
.
3
×
10
-
3
4
:
2
c
o
mp
r
e
sso
r
u
si
n
g
a
p
p
r
o
x
i
mat
e
f
u
l
l
a
d
d
e
r
5
.
4
1
.
0
1
.
2
×
10
-
3
2
.
0
×
10
-
3
4
:
2
c
o
mp
r
e
sso
r
u
si
n
g
a
p
p
r
o
x
i
mat
e
f
u
l
l
a
d
d
e
r
w
i
t
h
m
u
x
4
.
8
0
.
9
1
.
1
×
10
-
3
1
.
8
×
10
-
3
4
:
2
c
o
mp
r
e
sso
r
u
si
n
g
m
a
j
o
r
i
t
y
g
a
t
e
3
.
2
0
.
8
0
.
0
8
×
10
-
3
1
.
5
×
10
-
3
P
r
o
p
o
se
d
c
o
mp
r
e
sso
r
u
si
n
g
a
p
p
r
o
x
i
mat
e
f
u
l
l
a
d
d
e
r
3
.
0
0
.
8
1
.
7
×
10
-
3
1
.
2
×
10
-
3
P
r
o
p
o
se
d
c
o
mp
r
e
sso
r
u
si
n
g
a
p
p
r
o
x
i
mat
e
f
u
l
l
a
d
d
e
r
w
i
t
h
mu
x
2
.
6
0
.
7
1
.
6
×
10
-
3
1
.
1
×
10
-
3
P
r
o
p
o
se
d
c
o
mp
r
e
sso
r
u
si
n
g
m
a
j
o
r
i
t
y
g
a
t
e
1
.
9
0
.
6
0
.
4
×
10
-
3
0
.
8
×
10
-
3
5.
CO
NCLU
SI
O
N
T
h
is
r
esear
ch
i
n
tr
o
d
u
ce
s
an
in
n
o
v
ati
v
e
ap
p
r
o
x
i
m
atio
n
-
b
ased
MA
C
m
u
ltip
lier
ar
ch
it
ec
tu
r
e
th
a
t
s
ig
n
i
f
ica
n
tl
y
e
n
h
a
n
ce
s
co
m
p
u
t
atio
n
al
ef
f
icie
n
c
y
f
o
r
co
n
v
o
l
u
t
io
n
o
p
er
atio
n
s
in
C
NN
ap
p
licatio
n
s
.
T
h
is
d
esig
n
lev
er
ag
e
s
ad
v
a
n
ce
d
4
:2
,
5
:2
,
an
d
7
:2
co
m
p
r
ess
o
r
s
u
tili
zi
n
g
a
p
p
r
o
x
im
a
te
co
m
p
r
ess
o
r
s
,
ap
p
r
o
x
i
m
ate
w
it
h
m
u
x
b
ased
co
m
p
r
ess
o
r
an
d
m
aj
o
r
it
y
co
m
p
r
ess
o
r
s
,
ac
h
ie
v
in
g
m
ar
k
ed
i
m
p
r
o
v
e
m
en
t
o
v
er
c
o
n
v
e
n
tio
n
al
ac
cu
r
at
e
Dad
d
a
m
u
ltip
lier
s
p
ar
tial
p
r
o
d
u
cts
r
ed
u
ctio
n
s
.
T
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
s
d
e
m
o
n
s
tr
ate
a
n
o
tab
le
r
ed
u
cti
o
n
in
lo
g
ic
s
ize,
ar
ea
,
an
d
p
o
w
er
co
n
s
u
m
p
tio
n
w
h
ile
m
ai
n
tai
n
in
g
co
m
p
u
ta
tio
n
al
ac
c
u
r
ac
y
.
T
h
is
w
o
r
k
i
m
p
le
m
en
ta
tio
n
o
n
th
e
Xili
n
x
Ver
tex
-
5
FP
GA
p
r
o
v
es
th
e
p
er
f
o
r
m
an
ce
e
v
alu
at
io
n
o
f
L
U
T
r
ed
u
ctio
n
s
u
p
to
5
8
.
4
%,
an
o
cc
u
p
ied
s
lice
o
f
7
6
.
2
%,
an
d
a
laten
c
y
o
f
6
0
.
5
%
.
Mo
r
eo
v
er
,
th
is
r
esear
ch
i
n
clu
d
e
s
th
e
d
etailed
er
r
o
r
an
al
y
s
i
s
u
s
in
g
m
etr
ic
s
s
u
ch
as
E
R
,
ME
R
,
NM
E
D,
an
d
MRED
an
d
p
r
o
v
es
th
e
lo
w
e
s
t
er
r
o
r
v
alu
es
ac
r
o
s
s
all
m
etr
ic
s
u
s
i
n
g
m
aj
o
r
ity
lo
g
ic
-
b
ased
co
m
p
r
ess
o
r
s
,
w
h
ic
h
h
ig
h
li
g
h
ts
th
eir
s
u
itab
ilit
y
f
o
r
ap
p
r
o
x
i
m
ati
n
g
co
m
p
u
ti
n
g
e
n
v
ir
o
n
m
e
n
t
s
.
Fu
r
th
er
r
esear
ch
m
a
y
e
x
te
n
d
th
i
s
w
o
r
k
to
h
ig
h
er
o
r
d
er
m
u
l
ti
p
lier
s
(
1
6
×
16
an
d
32
×
3
2
)
,
ex
p
lo
r
e
d
y
n
a
m
ic
ap
p
r
o
x
i
m
atio
n
tech
n
iq
u
es,
a
n
d
ad
d
itio
n
all
y
i
n
te
g
r
ate
w
it
h
C
NN
ap
p
licatio
n
s
to
f
u
r
t
h
er
ac
h
ie
v
e
t
h
e
b
en
e
f
its
o
f
ap
p
r
o
x
im
a
tio
n
.
F
UND
I
N
G
I
NF
O
RM
AT
I
O
N
Au
t
h
o
r
s
s
tate
n
o
f
u
n
d
i
n
g
i
n
v
o
l
v
ed
.
AUTHO
R
CO
NT
RIB
UT
I
O
NS ST
A
T
E
M
E
NT
T
h
is
j
o
u
r
n
al
u
s
e
s
th
e
C
o
n
tr
ib
u
to
r
R
o
les
T
ax
o
n
o
m
y
(
C
R
ed
iT
)
to
r
ec
o
g
n
ize
in
d
i
v
id
u
al
au
th
o
r
co
n
tr
ib
u
tio
n
s
,
r
ed
u
ce
au
t
h
o
r
s
h
ip
d
is
p
u
tes,
an
d
f
ac
ilit
ate
co
lla
b
o
r
atio
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.