Int
ern
at
i
onal
Journ
al of Ele
ctrical
an
d
Co
mput
er
En
gin
eeri
ng
(IJ
E
C
E)
Vo
l.
10
,
No.
1
,
Febr
uar
y
2020
, pp. 71
9~72
7
IS
S
N: 20
88
-
8708
,
DOI: 10
.11
591/
ijece
.
v10
i
1
.
pp719
-
727
719
Journ
al h
om
e
page
:
http:
//
ij
ece.i
aesc
or
e.c
om/i
nd
ex
.ph
p/IJ
ECE
A
t
est
a
rchite
cture
d
esig
n
f
or
So
Cs
u
sing
a
t
am
m
eth
od
D.
R.
V.
A.
Sh
arath
Kum
ar
1
,
C
h.
Srini
vas
Kum
ar
2
,
Rag
am
ay
i
S
.
3
,
P.
Sa
mp
ath
Ku
mar
4
,
K.
S
ai
K
um
ar
5
,
Sk
Hasa
ne
Ahamm
ad
6
1,4,5,6
Depa
rtment
of
Elec
tron
ic
s
an
d
Com
m
unic
at
io
n
Engi
n
ee
ring
2,3
Depa
rtment
of
Mathe
m
atics
1, 2
KLEF,
1, 2
Guntur
,
2,3,5,6
AP
-
52
2502,
1,4
Mal
la
r
e
dd
y
Instit
u
te
of
T
ec
hno
log
y
1,4
S
e
cund
e
rab
ad
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
J
un
25
, 2
019
Re
vised
A
ug
31
,
20
19
Accepte
d
Se
p
27
, 20
19
Te
st
arr
angi
ng
i
s
a
basic
issue
in
struct
ur
e
on
-
a
-
chi
p
(S
.
O.C)
expe
riment
m
ec
hani
z
at
ion
.
Capa
ble
inv
esti
gation
desi
gns
constra
in
the
gene
ra
l
orga
nizati
on
check
req
u
est
t
ime,
kee
p
awa
y
fro
m
ana
l
y
s
is
rese
r
ve
conf
li
c
ts
,
in
ad
di
ti
on
to
p
urpose
of
restr
i
ct
ion
cont
rol
d
i
ss
eminat
ing
in
t
he
m
idst
of
exa
m
ina
t
ion
m
anne
r.
In
thi
s broa
dshee
t
,
we
abse
nt
a
fused
m
et
hod
to
m
ana
ge
a
coupl
e
of
te
st
arr
angi
ng
issues.
W
e
first
pre
sent
a
sy
st
em
to
choose
per
fec
t
ti
m
et
ables
for
se
nsibl
y
eva
lu
at
ed
SO
C’s
among
n
ee
d
associ
at
ions,
i.
e
.
,
p
la
ns
tha
t
spare
a
ll
ur
ing
orde
rings
among
te
sts.
Thi
s
furthe
rm
ore
ac
quai
n
ts
a
ca
pab
le
heur
ist
ic
esti
m
at
ion
wit
h
pla
n
exa
m
ina
tions
designe
d
for
enor
m
ous
S.O.Cs
through
nee
d
nec
essit
i
es
in
pol
y
nom
ia
l
oc
ca
sion.
W
e
portray
a
nar
ra
ti
ve
figur
ing
with
the
pur
pose
of
uses
pre
-
emption
of
te
st
s
to
sec
ur
e
ca
pab
le
da
te
-
bo
oks
in
fav
our
of
SO
Cs.
Expl
ora
tor
y
m
ark
s
on
beha
lf
of
an
educ
a
ti
ona
l
S
-
O
-
C
plus
a
cu
tt
in
g
edge
SO
C
exh
ibi
t
wi
th
the
aim
of
ca
pab
l
e
inve
stigation
ti
m
et
ab
le
s
be
ab
le t
o
subs
ist
gai
n
ed in
sensibl
e
CP
U
occ
asion
.
Ke
yw
or
d
s
:
Core
base
dorg
a
nizat
ion
s
E
ntre
nch
e
d co
r
e test
ing
M
ixedinte
ger
l
inear
pro
gr
am
m
ing
So
C
Copyright
©
202
0
Instit
ute of
Ad
v
ance
d
Engi
ne
eri
ng
and
Sc
ie
n
ce
.
Al
l
rights re
serv
ed
.
Corres
pond
in
g
Aut
h
or
:
Sai K
um
ar K
a
yam
,
Dep
a
rtm
ent o
f El
ect
ro
nics
and C
omm
un
ic
ation
En
gin
ee
rin
g,
KL Un
i
versi
ty
,
Vij
ay
awa
da, G
un
t
ur,
I
ndia
.
Em
a
il
:
sai
ku
m
ark
ay
am
4@
gm
ai
l.com
1.
INTROD
U
CTION
This
work
co
ntracts
th
r
ough
the
ar
range
m
ent
of
chec
k
m
od
el
s
f
or
sp
eci
fic
S/O
/C
diff
ic
ult.
These
struct
ures
co
ntain
wra
pp
i
ng
al
ong
wi
th
A
TAMs
.
F
or
a
s
pecifie
d
S
-
O
-
C,
am
ong
s
howe
d
li
m
it
at
i
on
s
of
com
po
ne
nts
al
so
their
asse
ssm
ents,
we
structu
re
pla
ns
wh
ic
h
re
str
ai
n
the
neces
sary
A.
T.
E
ve
ct
or
rem
inisce
nce
sign
ific
a
nce
al
ong
am
id
te
st
r
equ
e
st
occasio
n.
I
n
this
broa
ds
he
et
,
we
fig
ur
e
the
iss
ues
of
te
st
bu
il
di
ng
pla
n
t
og
et
her
f
or
c
om
po
nen
ts
in
th
e
com
pan
y
of
set
tl
ed
with
ve
rsati
le
distance
en
d
to
e
nd
lo
ok
at
m
anacle
s.
Al
ong
these
li
ne
s,
we
decide
a
m
eanin
g
of
a
pla
n
sel
f
-
go
vernin
g
te
st
tim
e
cuts
dow
n
set
out
towa
r
d
SO
Cs
an
d
s
umm
ary
the
lowe
r
set
ou
t
c
har
a
ct
erist
ic
s
toward
the
'
ITC'
O3
SO
C
Test
Be
nc
hm
ark
s'
.
W
e
abse
nt
a
book
buil
di
ng
sel
f
-
suffici
ent
heurist
ic
count
t
o
fa
ci
li
t
at
e
su
ccess
f
ully
red
esi
gns
th
e
analy
sis
plan
f
or
a
pr
ea
rr
a
nged
S.O.C
[
1].
Th
e
com
pu
ta
ti
on
ben
e
fici
al
ly
c
hoos
es
t
he
am
ount
of
A_
T
A
MS
furthe
rm
or
e
thei
r
siz
es,
the er
rand o
f
el
em
ents
t
o
T
AMs,
m
or
eov
e
r
t
he
wr
a
pp
er
str
uctu
re p
er
sect
ion. We
dem
on
strat
e ho
w
th
us
fig
ur
in
g
co
ntainer
he
sec
ond
-
ha
nd
on
beh
a
lf
of
stream
li
n
ing
to
gethe
r
Test
m
oto
r
veh
i
cl
e
plu
s
Test
Ra
il
Ar
c
hitec
tures
a
m
on
g
co
ns
ec
utive
m
or
eo
ve
r
sim
i
la
r
exa
m
inati
on
de
sig
ns
.
Ex
plorat
or
y
m
ark
s
in
fa
vour
of
the
'
I.
T.C'
O3
SO
C
Test
Be
nc
hm
ark
s'
exh
ibit
to
facil
it
ate
stood
ou
t
f
r
om
i
n
adv
anc
e
disp
e
rse
d
fi
gurin
g’s
,
we
sho
w
sig
ns
of
im
pr
ov
em
ent
te
st
tim
es
at
insign
ific
a
nt
figure
tim
e
[2
,
3
]
.
T
he
usa
ge
of
te
st
trad
it
ion
s,
tree
-
creati
ng
figurin
g’s
f
or
po
wer
-
obli
ge
d
book
i
ng,
a
nd
co
m
po
sed
T
AM
plan
a
nd
te
st
a
rr
a
ng
i
ng
are
di
ff
e
ren
t
sta
rting
la
te
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
10
, No
.
1
,
Febr
uar
y
2020
:
71
9
-
727
720
M
odul
ar
test
de
vel
opmen
ts
Iso
la
te
d
te
st
enh
a
ncem
ent
is
log
ic
al
ly
us
ed
f
or
SO
C
s.
N
on
-
m
et
ho
d
of
reas
onin
g
m
od
ules,
for
in
sta
nce,
e
m
bed
ded
strai
gh
t
forw
a
r
d
e
quipm
ent
al
ong
with
rem
inisce
nces
necessit
at
e
sta
y
lon
e
t
axin
g
unpaid
to
t
heir
'
un
pre
dicta
ble'
r
ou
te
orga
niz
at
ion
.
Dim
boxed
outc
ast
j
ogs,
f
or
i
ns
ta
nc
e,
ha
r
d
(
plan
)
fo
c
us
e
s
and
m
ixed
fo
c
us
es,
f
or
w
hich
no
util
iz
at
ion
purposes
of
i
ntrigue
a
re
ide
ntifie
d,
ought
t
o
be
at
te
m
pted
beside
the
inv
est
i
gations
a
s
giv
e
n
t
hro
ugh
their
s
upplier,
in
a
ddit
ion
to
t
hu
sl
y
fu
rt
her
m
or
e
necessit
at
e
sta
y
lon
e
ta
xin
g.
Re
ga
r
dless,
des
pite
inten
ded
f
or
justi
ficat
ion
sect
ion
s
of
w
hich
t
he
util
i
zat
ion
unpret
entio
us
com
po
ne
nts
ar
e
recogn
iz
e
d,
sp
eci
fic
te
st
progressi
on
is
an
en
gag
i
ng
opti
on
[4
-
6].
At
this
po
i
nt,
a
sp
eci
fic
'
seg
m
ent
and
-
vanq
uish
'
chec
k
en
ha
ncem
ent
m
ov
e
to
ward
s
le
ssens
t
he
te
st
age
e
nrol
occasio
n
with
relat
ed
inf
or
m
at
ion
qu
antit
y.
Ultim
ately
,
a
delibe
ra
te
chec
k
c
om
e
nea
r
e
ng
a
ges
ordeal
r
ecy
cl
e,
w
hich
pa
rtic
ul
arly
fu
lfil
ls
if
a
m
i
dd
le
or
el
se
el
e
m
ent
is
worn
in
va
ri
ous
S
.
O.
C
str
uctu
res.
W
it
h
the
t
ru
e
obj
ect
ive
t
o
f
aci
li
ta
te
est
i
m
at
ed
exa
m
inati
on
e
nh
a
ncem
ent,
an
entre
nch
e
d
el
e
m
ent
ha
ve
t
o
be
i
naccess
ible
be
ginnin
g
it
’
s
inco
rpor
at
in
g
equ
i
pm
ent
al
s
o
el
ect
rical
ex
per
im
ent
get
towa
r
ds
ou
gh
t
to
ex
ist
giv
e
n.
Z
or
ia
n
et
al
.
[
1
]
introd
uce
a
tra
diti
on
al
c
onne
ct
ed
ex
per
im
e
nt
get
to
bu
il
din
g
pe
rm
i
tt
ing
confine
d
try
in
g
of
S
.
O
.C
.
s
in
cl
ud
i
ng
3
par
ts
f
or
eac
h
m
od
ule
-
un
de
r
-
te
st:
(
1
)a
ana
ly
sis
config
ur
a
ti
on
basis
al
on
g
with
desce
nd,
(
2)
a
in
vestig
at
io
n
get
the
op
portun
it
y
to
instr
um
ent
(T.A.M)
,
f
ur
the
rm
or
e
(
3)
a
c
ov
e
rin
g.
The
bi
nd
i
ng
con
ta
ine
r
with
dr
a
w
s
the
el
e
m
ent
beg
in
ning
it
s
conditi
on
m
or
eo
ve
r
giv
e
s
tradi
ng
help
fu
l
ness
fla
nk
e
d
by
util
it
a
rian
ri
gh
t
of
e
nt
ry
to
the
el
em
ent
plu
s
te
st
a
dm
iss
ion
from
side
to
side
the
T
.
A.
M.
The
exa
m
inati
o
n
c
onfigurati
on
hav
e
a
far
reachi
ng
c
ras
h
tog
et
he
r
on
th
e
necessa
ry
ve
ct
or
rem
inisce
nce
sig
nificanc
e
for
each
A.
T
.E
canal,
a
nd
unde
r
the
analy
sis
r
equ
e
st
instanc
e
of
t
he
S.
O.C
,
2
in
puts
pa
ram
et
er
in
the
gen
e
ral
S
O
C
exam
inati
on
costs.
In what
eve
r
is
le
ft o
f
this
ne
w
sp
a
per
,
w
e
un
r
eserv
e
dly i
ns
i
nuat
e these
II li
m
it
a
ti
on
s as
as
sessm
ent tim
e
'
.
This
broa
dshee
t
watches
out
f
or
the
s
ubj
ect
of
orga
nizing
c
onvin
ci
ng
as
w
el
l
as
capab
le
check
get
to
structu
res
incl
ud
i
ng
pac
kag
i
ng
furthe
rm
or
e
T.A
.M
’s.
T
he
broa
dsheet
po
i
nts
of
interest
the
offic
ia
l
issue
i
m
plica
ti
on
s
of
assessm
ent
pl
an
up
gr
a
de
co
ncernin
g
com
pulso
ry
A.
T
.E
ve
ct
or
rec
ollec
ti
on
si
gn
i
ficanc
e
al
so
analy
sis
subm
i
ssion
insta
nce,
tog
et
her
on
beh
al
f
of
sect
ion
s
th
rou
gh
s
et
tl
ed
extent
c
heck
ha
ndcu
ffs
a
nd
furthe
rm
or
e
int
end
e
d
f
or
At
la
st,
we
the
re
e
xplo
rator
y
m
arks
desi
gn
e
d
f
or
the
ITC'
O
2
S
-
O
-
C
Test
ya
r
dst
ic
ks
,
wh
ic
h
dem
on
s
trat
e
to
facil
itate
TRARC
H
I
TECT
giv
e
w
ay
centere
d
i
nvest
igati
on
po
i
nt
in
tim
e
gr
a
des
i
n
su
pe
r
flu
ou
s
en
l
ist
instance
[6
-
8]
.
a.
The
c
onti
nu
at
i
on of t
hi
s doc
um
ent is d
eal
t w
it
h
as see
ks
a
fter. Fra
gm
ent
b.
Re
views p
rev
i
ou
s
lab
our
i
n
t
his s
pace. Te
r
ritory
c.
Descr
i
be
the
issues
of
ex
pe
rim
ent
bu
il
din
g
struct
ur
e
m
utu
al
ly
on
be
half
of
el
em
e
nts
am
on
g
set
tl
ed
m
or
eov
er
ver
s
at
il
e
du
rati
on
channel
m
anacle
s,
tolerat
in
g
the
vital
lim
i
ta
ti
on
s
of
sect
ion
s
al
ong
with
a m
axi
m
a
l S
-
o
-
C T
-
A
-
M t
hic
kn
e
ss a
re
dem
on
st
rated.
d.
In
sect
or
un
derneath
we
decid
e
an
e
nhance
d
su
bo
rd
i
nate
set
out
to
ward
th
e
analy
sis
inst
ance
of
a
ag
re
e
d
S.o.C. Fi
nally
this
disp
la
ys
our bu
il
di
ng in
de
pende
nt h
e
uri
sti
c p
r
ogressi
on
fig
ur
in
g TR
-
A
RC
HI
TECT
.
e.
Sect
ion
3
pr
e
sents
exec
utio
n
pur
poses
of
enthusia
sm
o
f
buil
ding
sp
e
ci
fic
portion
s
m
eant
fo
r
T
R
-
ARCHI
TECT
pro
the
exam
inati
on
m
oto
r
veh
ic
le
al
on
g
with
Test
_Rail
Archit
ect
ur
es
.
At
end
e
ncloses
te
st
resu
lt
s
f
or
t
we
lve
ya
rd
sti
ck
SO
Cs.
We
ta
ke
a
gander
at
ordeal
instanc
e
m
ark
s
f
or
T
R
-
ARCH
ITEC
T
in
add
it
io
n
to
th
o
se
gaine
d
besi
de
va
rio
us
syst
e
m
s
to
the
theor
et
ic
al
infer
i
or
bounce.
Z
one
8
com
plete
s
t
his
m
anu
scri
pt.
2.
E
X
ISTE
D
W
ORK
Diff
e
re
nt
assessm
ent
m
od
el
s
con
ta
in
be
por
tray
ed
in
w
riti
ng.
Ae
rts
plu
s
Ma
rinissen
[
41
portray
e
d
the
3
swee
p
ba
se
analy
s
is
design
s
delineat
ed
in
Fig
ur
e
1
(a)
the
Mult
ip
le
xin
g
bu
il
di
ng,
(
b
)
the
Daisy
chain
Ar
c
hitec
ture
, a
lso (
c
)
t
he Dist
rib
ution A
rch
it
ect
ur
e.
Figure
1
.
(
a
)
M
ulti
plexing a
rc
hitec
ture
;(
b) Da
isy
_ch
ai
n arc
h
;(
b)
Distrib
ution arc
h
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
A test
a
rc
hitec
ture
desig
n
f
or
So
Cs
u
si
ng
atam
meth
od (D.
R. V. A.
Sh
ar
at
h
K
umar)
721
In
t
he
M
ulti
plexing
furthe
rm
or
e
Daisy
se
qu
ence
A
rc
hitec
tures,
e
ve
ry
e
lem
ents
gain
i
nductio
n
to
the
occ
upie
d
open
T.
A.
M
br
e
adth.
I
n
the
Mult
iplexin
g
plann
i
ng,
on
ly
a
so
li
ta
ry
unit
con
ta
ine
r
be
gott
en
t
o
without
a
m
ome
nt'
s
delay
.
This
reco
m
m
e
nd
s
the
total
check
m
o
m
ent
is
the
entire
of
the
perso
na
ge
com
po
ne
nt
ex
a
m
inati
on
pe
riod
s
,
ye
t,
ad
diti
on
al
ba
sic
al
ly,
in
li
ke
m
ann
er
with
the
i
ntentio
n
of
m
odule
-
exter
nal
diff
ic
ult
(i.e.,
ha
rd
the
utensils
plus
cabli
ng
am
id
st
the
el
e
m
ent)
is
un
gainly
be
fore
sti
ll
incr
edible
.
This
is
a
res
ult
of
the
way
with
the
ai
m
of
m
erely
a
s
olit
ary
el
e
m
ent
wr
ap
pe
r
co
nt
ai
ner
be
gott
en
to
on
the
do
ub
le
,
al
t
hough
desi
gn
e
d
f
or
m
od
ule
-
ou
tsi
de
di
ff
ic
ul
t
the
bindin
gs
of
so
m
ewh
e
re
arou
nd
tw
o
el
e
m
ents
require
to
he
got
to
in
the
m
e
antim
e.
In
view
of
it
s
bypass
fr
am
ewo
rk,
th
e
Daisy
chain
c
on
st
ru
ct
io
n
do
es
not
con
ta
in
this c
ontr
ol
[
9]. In
t
he
sh
a
rin
g plan
nin
g, the
su
m
o
pe
n
T
.A
-
M b
rea
dth
is
scat
te
red m
or
e than
t
he parts
.
This
em
po
wers
sect
ion
s
to
be
at
tem
pted
at
the
sam
e
t
i
m
e,
and
al
ong
the
se
li
nes
the
total
SO
C
te
st
pe
rio
d
is
the
m
os
t
extra
o
r
din
a
ry
of
th
e
sepa
rate
unit
te
st
per
i
od
s
.
W
it
h
the
ulti
m
at
e
obj
ect
ive
t
o
c
on
st
rain
t
he
S.C.
K
inv
est
igati
on
point
in
ti
m
e,
t
he
br
ea
dth
of
an
c
har
act
e
r
T
.A
.M
ha
ve
t
o
be
prom
otion
a
l
to
the
pro
por
ti
on
of
inv
est
igati
on
inf
or
m
at
ion
to
facil
it
at
e
ou
ght
to
be
el
a
te
d
to
in
ad
diti
on
to
sta
rting
a
un
it
relat
ed
with
the
T
-
A
-
M.
T
he
analy
sis
m
ea
ns
of
tran
sport
const
ru
ct
io
n
s
how
n
th
rou
gh
V
arm
a
m
or
eov
er
Bhati
a
is
a
m
i
x
of
the Mult
iplexi
ng also
A
ll
ocat
ion
A
rch
it
ect
ur
es.
A
sin
gula
r
an
al
ysi
s
transport
is
on
a
ver
y
basic
le
v
el
the
pr
opo
rtio
na
l
as
wh
at
is
portraye
d
by
the
Mult
iplexi
ng
st
ru
ct
ur
al
de
sign
:
el
em
ent
s
relat
ed
with
a
com
par
at
ive
te
st
transport
m
us
t
be
at
te
m
pted
su
ccessi
vely
.
The
Te
st
Bus
Ar
c
hitec
ture
co
ns
ide
rs
va
rio
us
te
st
tra
nsports
on
1
S
-
O
*C,
w
hich
w
ork
un
i
nh
i
bite
dly,
seei
ng
t
hat
in
the
Distrib
ution
B
uildin
g.
Elem
ent
li
nk
ed
t
o
a
pe
rio
di
c
te
st
sh
ipm
ent
m
eet
the
neg
at
i
ve
eff
ect
s
of
va
gue
dr
a
wb
ac
k
f
r
om
e
it
her
the
arch
it
ect
ure
of
m
ulti
plexing
,
i.e.
Diff
ic
ult
exter
nal
m
od
ule
is
trou
bling
or
un
us
ua
l.
The
desig
n
Test
Bus
Ar
c
hitec
ture
[10
-
14]
is
sh
own
in
Figure
2(a).
T
he
S
O
C
include
s
six
c
om
po
ne
nts,
ca
ll
ed
A
by
F.
This
Test
Bus
Ar
c
hitec
ture
sta
rting
po
i
nt
involve
s
3
e
xperim
ent
trans
portat
ion
s
.
Module
s
A
n
as
well
as
B
are
associat
ed
wi
th
ei
ther
the
th
ree
-
width
Te
st
Bus
1;
m
o
du
le
s
C,
D
al
so
E
are
as
s
ociat
ed
with
t
he
f
our
-
wi
dth
Test
Bus
2;
m
odule
F
is
ass
ociat
ed
with
t
wo
-
wi
dth
Te
st
Bus
3.
Figure
2(b)
de
m
on
strat
es
tha
t
the
te
st
plan
can
be
e
xam
i
ned.
T
he
th
ree
te
st
trans
port
s
be
ca
pa
ble
of
be
worked
un
i
nh
i
bitedly
.
The
m
od
ules
rel
at
ed
with
a
n
ave
r
age
te
st
transpor
t
are
at
te
m
pt
ed
in
a
sel
f
-
de
ci
sive
anyway
pro
gr
e
ssive
dem
and
.
We
cal
l
this
ti
m
et
able
su
cce
ssive,
i
n
li
ght
of
the
fact
tha
t
per
T
-
A
-
M
the
c
om
po
ne
nt
s
are
(a)
(
b)
Figure
2
Exam
ple
Test
m
oto
r
veh
ic
le
co
ns
tr
uction
(an)
as
well
as
a
po
ssi
ble
lookin
g
at
consecuti
ve
te
st
plan
(h).
The
a
naly
sis
Ra
il
Ar
c
hitec
ture
obta
ina
ble
by
Ma
rini
ssen
et
al
.
[6
]
is
a
m
ix
of
t
he
Daisy
c
hain
with
Allotm
ent
Ar
chite
ct
ur
es.
A
s
olit
ary
assessm
ent
rail
ing
is
for
the
m
os
t
par
t
the
e
qu
al
as
wh
at
is
por
tray
e
d
thr
ough
the D
a
isy
chain
A
rc
hitec
ture:
m
od
ules
relat
ed
with the
pro
portio
na
l
Test
Ra
il
ca
n
be
at
tem
pted
in
th
e
m
eantim
e
and
furthe
rm
or
e
pro
gr
e
ssively
.
A
Test
Ra
il
A
rch
it
ect
ure
thi
nk
s
ab
out
va
riou
s
Test
rail
s
on
one
SO
C
that
w
orks
without
res
erv
at
io
n,
as
in
the
Ar
c
hitec
ture
of
P
rod
uc
ti
on
.
The
ad
va
ntage
of
Test
Ra
il
Ar
c
hitec
ture
over
Test
B
us
Desig
n
is
that
it
si
m
ul
ta
neously
enab
le
s
a
ccess
to
dif
fere
nt
or
al
l
w
ra
pp
e
rs,
su
pp
or
ti
ng
s
ubsyst
e
m
-
external
te
sti
ng
.
Fi
gure
3
sho
w
s
a
m
od
el
Test
-
Ra
il
Ar
c
hitec
ture.
Tes
t
Ra
i
l
Ar
c
hitec
tures
s
upport v
ari
ou
s
so
rts of
te
st
de
sign
s
.
Fi
gure
2
(b)
as w
el
l
as (
c) d
em
on
strat
e
II
proba
ble
lookin
g
at
te
st
desig
ns
.
The
tim
et
able
in
Fig
ure
2(b)
is
a
co
ns
ec
utiv
e
tim
et
able;
the
el
em
ents
relat
ed
with
a
run
of
the
m
ill
Test
-
Ra
il
are
at
te
m
pted
in
a
sel
f
-
co
nf
i
den
t
a
nywa
y
pr
og
ressive
dem
and
[
15
-
17]
.
The
ti
m
et
a
ble
in
Figure
2(c)
is
a
par
al
le
l
date
-
bo
ok.
In
this
tim
e
ta
ble,
we
detai
l
to
te
st
a
ll
m
od
ules
relat
ed
with
a
co
m
m
on
Test
-
Ra
il
in
si
m
il
ar.
Figure
2
.
(a
)
E
xam
ple Test B
us
Ar
c
hitec
ture
an
d
(b)
a
possi
ble corre
spo
nding
serial
test
s
chedule
2.1.
Problems
i
n E
xisted Me
thod
To
plan
a
tria
l
buil
ding
i
n
s
upport
of
a
spe
ci
fied
plan
of
el
em
ent
al
on
g
with
a
kn
own
fi
gure
of
exp
e
rim
ent
stick
s,
a
S.
O
-
C
integrat
or
nee
ds
to
ch
oose
(
1
)
t
he
c
hec
k
desig
ning
m
ake,
(
2)
the
am
ou
nt
of
TAMs,
(3)
th
e
br
ea
dths
of
these
T
AMs,
(4)
the
unde
rtakin
g
of
m
od
ules
t
o
T.
A.M
s,
f
ur
the
rm
or
e
(
5)
the cove
rin
g
st
ru
ct
ur
e
f
or
ea
c
h
m
od
ule.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
10
, No
.
1
,
Febr
uar
y
2020
:
71
9
-
727
722
2.1.1.
Issue
1 [Fixe
d
-
dur
at
i
on
M
odul
e
-
inner
Sc
an
handcu
ff
s
]
A
gam
e
plan
of
el
em
ents
M,
furth
e
rm
or
e
f
or
e
ver
y
sect
io
n
m
E.M
the
am
ount
of
analy
si
s
str
uctur
e
s
p,
the
am
ount
of
pr
act
ic
a
l
da
ta
incur
a
ble
I
,
the
am
ou
nt
of
util
it
arian
yi
el
d
w
orkstat
io
n
om
.the
am
ou
nt
of
help
fu
l
bi
direc
ti
on
al
incu
rab
l
e
b,
,
the
am
ou
nt
of
ra
nge
m
a
nacles
s,,
al
on
g
with
in
fa
v
our
of
eve
ry
yi
e
ld
series
k,
the
exte
nt
of
the
yi
el
d
sequ
ence
in
flip
dis
appointm
ents
l
,
m
,
k
.
In
a
dd
it
ion
is
ag
reed
a
nu
m
eral
W
m
a
x
tha
t
addresses
the
m
os
t
extrao
r
di
nar
y
am
ount
of
S
.O.C
-
le
v
el
TAM
ropes
wi
th
the
as
pire
of
can
be
w
orn
.
Choose
a
T_
A_
M
desi
gn
i
ng
a
nd
a
w
rapper
pla
n
f
or
each
m
od
ule
with
the
tr
ue
obj
ect
iv
e
that
th
e
general
S
OC
-
le
vel
te
st t
i
m
e (in
clock cy
cl
es)
is
r
est
rict
ed
al
so
W
m
ax
isn'
t out
per
f
orm
ed.
2.1.2.
Issue
2 [Flexi
ble
-
dist
an
ce
e
nd
t
o
e
nd M
odul
e
-
d
omes
tic Scan
Chains]
All
lim
it
ation
s
as
dem
on
strat
ed
in
tro
uble
1,
anyway
as
oppose
d
to
the
a
m
ou
nt
of
co
m
pass
chains
entire
f
ur
t
herm
or
e
the
sp
a
n
l,
h
i
n
sup
port
of
e
ver
y
yi
el
d
series
k,
the
w
ho
le
in
te
ger
of
br
ea
dth
flip
disap
pointm
ents
is
kn
own
.
Choose
a
T
-
A
-
M
bu
il
ding
m
or
e
over
a
w
ra
pp
e
r
pla
n
f
or
each
one
el
em
ent
with
the
true
obj
ect
ive
to
facil
it
at
e
the
ge
ner
al
SO
C
-
le
vel
te
st
tim
e
(in
cl
ock
cy
cl
es)
is
const
rained
as
wel
l
as
W
m
ax
isn'
t ou
t
pe
rfor
m
ed
[
18
-
19]
.
3.
PROP
OSE
D MET
HO
D A
ND PR
OBL
EMS
SOL
UTION
3.1.
Problem
1
[
fi
xe
d
-
le
n
gt
h
m
odul
e
-
interi
or
sc
rutiniz
e
cuff
s
]
A
plan
of
sect
ion
s
M,
al
ong
with
ai
m
ed
at
ever
y
one
Mo
dule
m
E
-
M
the
am
ou
nt
of
c
he
ck
str
uctu
re
s
p.
the
am
oun
t
of
p
ra
gm
at
ic
d
at
a
incur
a
ble
I
,
the
am
ou
nt
of u
ti
li
ta
rian
yi
eld
m
or
ta
l
om
.
t
he
am
ou
nt
of va
luable
bid
irect
io
nal
f
at
al
s
b,
,
the
a
m
ou
nt
of
s
c
op
e
c
hains
s
,
m
or
eo
ver
for
al
l
yi
el
d
chain
k,
the
pi
ece
of
the
ra
ng
e
se
qu
ence
in
tu
r
n
ov
er
disa
ppoin
tm
ents
l
,
m
,k
.
Al
so
is
gi
ven
a
di
git
W
m
ax
tha
t
addresses
the
m
os
t
extra
ordina
ry
intege
r
of
SO
C
-
le
vel
T
AM
s
uppo
rts
to
facil
it
at
e
con
ta
ine
r
be
us
e.
Ch
oos
e
a
T
_A_M
buil
ding
and
a
w
ra
pp
e
r
structu
re
for
ea
ch
m
od
ule
with
the
t
ru
e
ob
j
e
ct
ive
that
the
ge
ner
al
S
OC
-
le
vel
te
st
tim
e
(i
n
cl
oc
k
cy
cl
es)
is restri
ct
ed
in a
ddit
io
n
to
W
m
ax
isn
'
t ou
tper
form
e
d.
3.2.
Problem
2 [
fle
xible
-
le
n
gt
h
m
od
ule
-
intern
al
sca
n
ch
ains
]
All
const
raints
as
decide
d
i
n
tro
ub
le
1,
ye
t
as
oppose
d
to
the
am
ou
nt
of
com
pass
chain
s
ad
d
up
to
al
ong
wit
h
the
extent
l,
h
on
be
half
of
e
ver
y
yi
el
d
chain
k,
t
he
w
hole
am
ou
nt
of
ra
nge
flip
disap
pointm
e
nts
fm
is
sp
eci
fied
.
Choose
a
T
.A.M
desig
ning
in
a
dd
it
io
n
t
o
a
wrapp
i
ng
plan
f
or
each
one
m
o
du
le
with
the
ul
tim
at
e
obj
ect
ive
that
the
ge
ner
al
SO
C
-
le
vel
te
s
t
tim
e
(in
cl
oc
k
cy
cl
es)
is
c
onstrai
ne
d
pl
us
W
m
ax
is
n'
t
ou
t
perform
ed
[20]
.
3.3.
Sequen
tial
ci
r
cuit t
e
stin
g
In
s
uccessi
ve
ci
rcu
it
s
the
un
der
ly
in
g
sta
te
(en
li
st'
s
qu
al
it
ie
s)
isn'
t
of
co
ur
se
know
n.
S
ub
s
eq
ue
ntly
,
the
re
finem
ent
of
flaws
an
d
the
prolife
rati
on
of
t
he
relat
ing
i
nc
orrect
r
eact
ion
s
m
ay
swing
t
o
be
a
ha
r
d
unde
rtakin
g.
An
a
ns
we
r
is
to
util
iz
e
strateg
ie
s
for
the
correct
instat
em
ent
of
the
ci
rcu
it
sta
te
to
kno
wn
qu
al
it
ie
s.
Use
of
ap
pro
pr
ia
te
te
st
vecto
r
s
uc
cessi
ons
as
well
as
the
uti
li
zat
ion
of
Set
/R
eset
sign
s
to
set
u
p
the
require
d
sta
te
.
I
m
pr
ovem
ent
of
pro
duct
ive
syst
e
m
s
to
set
the
un
de
rly
ing
sta
te
and
watch
the
co
nse
qu
e
nt
sta
te
after
the
re
act
ion
of the c
ircu
it.
The
m
e
m
or
y
com
po
ne
nts
(l
oc
ks
or
Fli
p
-
Flo
ps
)
in
a
st
ru
ct
ure
are
le
g
it
i
m
a
te
ly
associat
ed
with
fr
am
e
a
bro
ught
to
ge
ther
m
ov
e
e
nlist
(f
il
te
r
e
nlist
or
c
hai
n)
.
Al
ong
these
li
ne
s
the
insi
de
co
ndit
ion
of
the
ci
rcu
it
is
reso
l
ved
(cont
ro
ll
ed
)
by
m
ov
in
g
in
(e
xam
ine
in)
to
t
he
swee
p
enlist
the
re
qu
i
red
te
st
inform
at
ion
to
be
connecte
d
t
o
the
com
bin
at
io
nal
rati
on
al
e.
Be
sides,
the
c
urren
t
in
ward
sta
te
(p
ast
rati
on
al
e
reacti
on
)
can
be
seen
by
m
ov
in
g
out
(e
xam
ine
ou
t)
t
he
inf
orm
at
ion
pu
t
aw
ay
into
the
sw
eep
enlist
.
Fi
gure
3
is
te
sti
ng
relat
e
d
to sca
n
se
qu
e
nt
ia
ll
y. Figu
re
4
exp
al
in
s a
bout
scanin
g rega
r
din
g t
ing ge
ne
ral te
st.
Figure
5
e
xp
al
i
ns
a
bout
ou
t
put
way
of
te
sti
ng
her
e
d
-
flip
le
m
on
assum
es
an
im
per
at
ive
jo
b
s
o
we
ar
e
util
iz
ing
this
strat
egy
got
issu
e
1
arr
a
ng
em
ent
.
Fig
ur
e
6
ex
palins
that
outpu
t
ap
plica
ti
ons
in
the
te
st
pr
ocedu
re
her
e
we
got
hi
gh
preci
sio
n
con
t
rasted
wit
h
existe
d
strat
egies
A
cente
r
te
st
wr
app
e
r
cal
le
d
Test
Sh
e
ll
has
rem
ai
ned
pr
oj
ect
ed
thr
ough
Ma
rinissen
et
al
.
[3
]
si
m
ilarly
is
righ
t
n
ow
ex
plo
it
ed
pr
ivil
ege
d
P
hili
ps
.
The
Test
Shel
l
includes
of
the
com
ple
m
entary
segm
ents
.
Figu
re
6
sho
wn
s
that
ap
plica
ti
on
of
scan
ning
proce
dure
[21
-
22
]
.
A
m
or
ta
l
te
st
c
el
l
with
each
quar
ry.
The
te
st
cel
l
pr
ovides
bo
t
h
tran
sie
nt
r
esp
on
se
a
nd
di
scern
i
bili
ty
.
A
(d
isc
reti
onar
y)
bypass
rec
ord
al
lowi
ng
a
TAM
to
sideste
p
center
as
w
el
l
as
pack
agin
g,
thr
oughthe
ul
tim
at
e
go
al
of
te
sti
ng
an
ext
ra
center
co
nnect
ed
wi
th
a
com
par
a
ble
TAM.
A
blo
c
k
of
te
st
con
trols
(
TCB
)
.
The
TCB
has
a
bit
-
cut
envi
ronm
ent
al
on
g
with
enc
ompass
of
a
m
ov
e
in
ad
diti
on
to
a
rev
ive
enlist
.
The
TCB
is
pr
incipal
ly
fu
tu
r
e
to
m
anag
e
th
e
ta
sk
of
the
T
est
Sh
el
l,
th
rou
ghout
a
li
tt
le
ob
li
gato
ry
piece
c
uts.
Additi
on
al
c
ust
om
er
br
an
de
d
bit
cuts
ca
n
be
i
ncor
por
at
ed
f
or
m
anag
e
of
ce
ntre
inn
e
r
te
st
m
ann
e
r.
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
A test
a
rc
hitec
ture
desig
n
f
or
So
Cs
u
si
ng
atam
meth
od (D.
R. V. A.
Sh
ar
at
h
K
umar)
723
Figure
7
is
the
path
delay
m
od
el
reg
ard
i
ng
to
chai
n
of
flipflops
[23
-
24
]
.
Fig
ure
8
is
the
Fli
p
flop
chain
reord
i
ng w
it
h t
he help
of sca
n t
est
.
Figure
3
.
S
e
quentia
l
scan test
ing
Figure
4
.
G
e
ne
ral scan
test
Figure
5
.
S
ca
n path
desi
gn
Figure
6
.
S
ca
n app
li
cat
io
ns
Figure
7
.
De
la
y fault t
est
in
g
Figure
8
.
R
eo
r
der
i
ng of sca
n chain
Fli
p
Fl
op
We
hav
e
just e
xam
ined
va
rio
us
defo
rm
i
ti
es
that ca
n ca
us
e
po
st
pone
sho
rtcom
ing
s:
a.
GOS
deser
ts
b.
Re
sist
ive shor
t
ing
dese
rts am
ong h
ub
s
plu
s t
o
the
s
upply ra
il
s
c.
Scr
oungin
g
tra
ns
ist
or s
pill
ages,
ina
de
qu
at
e
on inte
rsecti
ons
as w
el
l as
w
rong o
t
herwise
m
ov
ed
lim
i
t
vo
lt
age
s
d.
Ce
rtai
n
ki
nds
of r
el
eases
e.
Pr
oc
ed
ur
e
v
a
ri
et
y con
ta
ine
r
si
m
il
arly
m
ake gadg
et
switc
h at
a s
peed le
sser t
han
t
he parti
c
ular.
Sw
ee
p
c
hain
r
eorderi
ng
is
a
proce
dure
util
i
zed
in
the
pla
n
a
nd
te
sti
ng
of
proces
sin
g
gadgets
th
at
e
m
po
we
rs
the
stream
li
ni
ng
of
set
ti
ng
and
s
ewin
g
flip
slu
m
p
reg
ist
ers
with
an
ou
tp
ut
chain.
It
is
util
i
zed
t
o
stream
li
ne
an
d reo
rd
e
r
t
he
s
w
eep c
hain p
ro
c
ess in t
he
e
ven
t
that it
g
et
s is
ol
at
es, ceased
or
congeste
d.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
10
, No
.
1
,
Febr
uar
y
2020
:
71
9
-
727
724
4.
HAZ
ARDS
2 vecto
r
s
ucces
sion i
s A
BC
=
(11
1)
,
(1
01)
a.
Gate G
1 prese
nts a
n
e
xtra post
ponem
ent o
f 1
un
it
.
b.
Pr
od
uctio
n
E
of e
ntryway
G 3
is h
ea
de
d
to
a
rati
on
al
e
1, on
ce com
po
ne
nt
fo
ll
owin
g D
-
> 0
.
c.
Pr
od
uce a
m
al
f
un
ct
io
n o
n
F
F
igure
9
is
t
he
gate
de
la
ys
w
hich
a
re
ci
cl
ed
in
F
ig
ur
e
10
exp
la
in
s
ab
out
tim
e
li
ne
ve
ct
or
a
he
n
ABC=
101
is
app
li
ed
.
Fi
gure
11
ex
palins
t
ha
t
dyanam
ic
hazard
s
relat
ed
to
te
st
path
w
hi
ch
is
sho
wed
in
F
ig
ur
e
12
s
hows
that
tim
e
loi
ne
sta
rting
w
he
n
AB=1
1
ap
plied
F
i
gure
13
i
s
the
sta
ti
c
ha
zard
s
pr
ocessi
ng
with
the
he
lp
of
dynam
ic
h
azar
ds
test
gen
e
rati
on.
Figure
9
.
G
at
e
delay
s ar
e ci
rcled
Figure
10
.
T
im
e li
ne
sta
rtin
g wh
e
n
vecto
r
ABC=
101 i
s a
pp
li
ed
Figure
11
.
Dyn
a
m
ic
H
azards
Figure
12
.
Tim
e Line
Starti
ng
Wh
en
V
ect
or
AB= (
11)
Is Appli
ed
Figure
13
.
Stat
ic
h
azar
ds
c
an
create
d
y
nam
ic
haza
rd
s
al
ong
te
ste
d pa
ths a
nd n
ee
d
t
o be c
onside
red
durin
g
te
st gen
e
rati
on
Tw
o vector
ar
r
ang
em
ent is A
B = (
01), (
11).
Entryway
G
2
has
a
de
fer
e
stim
ation
of
3
tim
e
un
it
s,
owin
g
wh
ic
hever
to
a
n
im
per
fecti
on
or
a
n
al
te
rn
at
e
physi
cal
execu
ti
on
of
the
N
A
ND
door.
No
te
,
i
n
c
on
t
rast
to
the
pa
st
pr
ece
den
t,
t
he
an
om
al
y
ha
pp
e
ns
pr
e
vious
to
the
plann
e
d
cha
nge
for
this
sit
ua
ti
on
,
al
s
o
be
able
to
nu
ll
ify
the
te
st
(e.g
.
bl
a
m
e
isn't
identifie
d)
.
Figure
14
e
xp
l
ai
ns
that c
riti
cal
p
at
hs at
6 tim
e uniuts s
how
n i
n belo
w
[25
-
26]
.
Figure
14
.
T
he
criti
cal
p
at
h(s)
of this ci
rcu
it
i
s 6
ti
m
e u
nits
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
A test
a
rc
hitec
ture
desig
n
f
or
So
Cs
u
si
ng
atam
meth
od (D.
R. V. A.
Sh
ar
at
h
K
umar)
725
Let
'
s set the clock
p
e
rio
d
T
=
7
Assu
me
onl
y
on
e
f
au
lt
y
p
at
h.
No d
el
ay
m
ist
a
ke
is
detect
ed
i
f path st
oppa
ge
the len
gth
of
P
3
is l
ess
tha
n 7 units.
This e
xp
e
rim
e
nt w
il
l n
ot
no
ti
ce sin
gle w
ai
t
f
aults t
he
le
ng
t
h o
f paths
P1 o
r P2.
Assu
me
ther
e
can be m
ultipl
e f
au
lt
y
p
aths.
Assum
e
P2
m
or
e
over
P
3
are
def
ect
ive
wit
h
P2
e
xten
d
th
e
"st
at
ic
gli
tc
h"
at
the
pr
oduc
ti
on
beyo
nd
7
unit
s,
after that
it
m
a
sk
P
3'
s d
el
ay
e
rror.
Thi
s t
est
is c
alled a no
n
-
r
obust
test f
or del
ay fault
P
3.
Alg
o
ri
th
m
1
[
TR
-
ARCHIT
ECT]
1
CR
E
ATE ST
ART S
OL
UT
I
ON
2 OPT
IMIZE
-
BOTT
OM UP;
3
OP
T
IM
IZE
-
TOP
D
O
WN
;
4
RES
H
UF
FL
E
Algori
th
m
2
[
CR
EATE STA
RT SO
LUT
I
O
N]
do
{
item =
10;
va
lue
=
va
lue
+ item;
} whil
e(value<
100);
Item =
10;
do
{
va
lue
=
va
lue
+ item;
} whil
e(value<
100);
Figure
.15
is
t
he
m
od
el
of
c
on
ce
ptu
al
te
st
sh
el
l
us
i
ng
thi
s
te
st
m
od
el
ezi
ly
and
fix
pro
blem
s
in
a
sim
ple
m
ann
er
[
27]
.
A
ll
m
et
hods
a
nd
te
c
hn
i
qu
e
s
a
re
re
du
ce
s
the
probl
e
m
s
1
a
nd
2
r
especti
vely
a
nd
th
ey
giv
e
bette
r q
ua
li
ty
d
esi
gn
.
Figure
15
.
C
onceptual
view
of Phil
ips’ Test
Sh
el
l
5.
RESU
LT
S
Fig
ure
16
(
a
-
e
)
is
the
blo
c
k
di
agr
am
of
te
st
cel
l
and
F
ig
ur
e
17
s
hows
the
internal
blo
c
k
diag
ram
of
te
st
cel
l
and
Figure
18
sho
ws
that
the
ou
tp
ut
wav
ef
or
m
s
of
te
st
cel
l.
Table
1
exp
la
ins
a
bout
com
par
ison
of
par
am
et
ers
at
TAM
an
d
AT
AM
m
od
el
s
he
re
eff
ic
ie
ncy
of
existe
d
is
80%
bu
t
pro
pose
d
m
e
tho
d
hav
e
99
.
8%
this is g
ood ac
hievem
ent. P
(
E)
is l
ess
co
m
par
ed
to
e
xisted
m
et
ho
d
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
10
, No
.
1
,
Febr
uar
y
2020
:
71
9
-
727
726
(a)
(
b
)
(
c
)
(
d
)
(e
)
Fig
ure
16
.
(a)
A
seq
_s
ca
n_te
s
t
,
(
b
).
Tec
hnol
og
ic
al
sc
hem
atic o
f
seq_sca
n
,
(
c
).
RTL
sch
e
m
at
ic
o
f
seq
_s
cani
ng_te
st
,
(
d
).
O
ut
pu
t
wav
e
f
or
m
s o
f seq
_s
can
_test
,
(e)
.
Test cel
l
Figure
17
.
I
nte
rn
al
blo
c
k of t
e
st cel
l
Figure
18
.
S
im
ulati
on
res
ult o
f
te
st cel
l
Table
1.
C
om
par
io
ns
with
AT
AM
PARAM
ET
ERS
TAM
ATA
M
Ef
f
icien
cy
80%
9
9
.8%
Prob
ab
ility
o
f
er
ro
r
0
.1
0
.01
6.
CONCL
US
I
O
N
At
lo
ng
la
st
ut
il
iz
ing
the
a
bo
ve
te
st
cel
ls
li
ke
powe
rful
pe
rils,
A
TAM
a
nd
re
orde
rin
g
te
sti
ng
a
nd
po
st
pone
blam
e
te
st
sq
ua
res
we
s
howe
d
sig
ns
of
im
pr
ove
m
ent
producti
vity
and
le
ss
li
kelihoo
d
of
bl
unde
r.
So
co
ntraste
d
with
TAM,
AT
AM
giv
es
the
bette
r
outc
om
e
s
.
Profici
ency
increm
ents
by
19.8%
an
d
li
kelihoo
d
of m
is
ta
ke
is 0
.
01 this is
great
accom
plish
m
e
nt contraste
d w
it
h
existe
d t
ech
niques.
REFERE
NCE
S
[1]
Y
.
Zori
an
,
e
t
a
l.,
“
Te
sting
Em
bed
ded
-
Core
Bas
e
d
S
y
stem
Chips
,”
Proceedi
ngs
I
EE
E
In
le
mation
ol
Test
Confe
ren
ce
(
ITC
)
,
Washington,
DC
,
1998.
[2]
E
.
J
.
Marini
ss
en,
et
al.
,
“
ITC'
02
SO
C
Te
st B
enc
h
m
ark
s W
eb
Site
,”
Available:
h
ttp
:
/
/www
.
ent
ra
.
res
ea
rch
.
philips
.
co
m
.
[3]
E
.
J.
Marin
issen
,
et
a
l.,“
A
Set
of
Benc
hm
a
rks
for
Madula
r
T
esti
ng
of
SO
Cs
,”
Proc
ee
dings
IE
EE
ln
rernationol
Test
Confe
renc
e
(
ITCJ
)
,
Baltimore,
M
D
,
2002.
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
A test
a
rc
hitec
ture
desig
n
f
or
So
Cs
u
si
ng
atam
meth
od (D.
R. V. A.
Sh
ar
at
h
K
umar)
727
[4]
E
.
J
.
Marin
issen,
et
al.,
“
W
rap
per
Design
for
Embedde
d
Core
Te
st
,”
Proc
e
edi
ngs
IEE
E
In
te
rnational
Tes
t
Confe
renc
e
(
ITC)
,
Atla
ntic
C
ity
,
NJ
,
pp.
911
-
920
,
2000.
[5]
V
.
I
y
enga
r
,
et
al.,
“
Co
-
Optimizat
ion
of
Te
stW
rap
per
and
T
est
Ac
ce
ss
Arch
it
e
ct
ur
e
for
Embedde
d
Cores
,”
Journal
of
E
le
c
tronic Te
sting:
Theory
an
d
Applications
,
v
ol.
18
,
pp
.
213
-
2
30,
2002
.
[6]
S
.
K
.
Goel
and
E
.
J
.
Marin
issen
,
“
Eff
ec
ti
v
e
and
Eff
icient
Te
st
A
rch
itect
ur
e
Desi
gn
for
SO
Cs
,”
Proce
ed
ings
IE
EE
Inte
rnational
Te
st Conf
ere
n
ce (
ITC)
,
Bal
ti
m
ore
,
MD
,
pp.
529
-
538
,
2002
.
[7]
S
.
K
.
Goel
and
E
.
J
.
Marini
ss
en
,
“
Lay
out
-
Driv
en
SO
C
Te
st
Archi
tectur
e
Design
for
Te
st
Ti
m
e
and
W
ire
Le
ngt
h
Minim
iz
at
ion
,”
Proce
ed
ings
De
sign,
Au
tomati
o
n,
and
Test
in
Europe
(
DATE)
,
Munich,
Germ
an
y
,
pp
.
738
-
74
3
,
2003.
[8]
S
.
K
.
Goe
l
and
E
.
J
.
Marin
issen
,
“
SO
C
Te
st
Archi
t
ec
tur
e
Design
fo
r
Eff
i
ci
en
t
Uti
li
z
at
ion
of
T
est
Ba
ndwidth
,”
AC
M
Tr
ansacti
ons on D
esign
Aut
omat
i
on
of El
e
ct
roni
c Syste
ms
,
vol
.
8
,
pp.
399
-
429
,
20
03.
[9]
S
.
K
.
Goel
and
E
.
J
.
Marin
issen
,
“
Control
-
Aw
are
Te
st
Archi
tectur
e
Design
for
Mo
dula
r
SO
C
Te
sti
ng
,”
Pro
ce
ed
ing
s
IEE
E
European
Test
Workshop (
ETW)
,
Maa
stricht,
The Net
he
rlands
,
pp.
57
-
62
,
2
003.
[10]
S
.
K
.
Goel
,
“
A
NovelWrappe
r
Cell
Design
f
or
Eff
i
cient
Tes
ti
ng
of
Hi
er
ar
chi
c
al
Cor
es
in
S
y
stem
Ch
ips
,”
Proce
ed
ings I
E
E
E
European
Test
Symposium (
ETS)
,
Ajac
c
io, Cor
sica
,
pp.
147
-
15
2
,
2004
.
[11]
L
.
Krundel
,
et
al
.
,
“
Us
er
-
Constrai
ned
T
est
Archi
tectur
e
Design
for
Modular
S
OC
Te
sting
,
”
P
roce
edi
ngs
IEEE
European
Test
S
ymposium (
ETS
)
,
Aja
ccio, Corsic
a,
pp
.
80
-
86
,
200
4.
[12]
A
.
Sehgal
,
et
a
l.,
“
P1500
-
Co
m
pli
ant
Te
st
W
rap
p
e
r
Design
for
Hier
arc
hi
ca
l
Cor
es
,”
Proce
ed
ings
IE
EE
Int
ernati
onal
Test
Confe
ren
ce
(
ITC
)
,
Charlott
e, NC
,
2004
.
[13]
S
.
K
.
Goel,
e
t
a
l.
,
“
Te
st
Infr
ast
ruc
ture
Design
for
the
Nexpe
ri
aT
M
Hom
e
Plat
form
PNX85
50
Sy
st
em
Chip
,”
Proce
ed
ings De
sign, Aut
omation
,
and
Test
in Europe
(
DATE)
, p
p.
108
-
113,
2004.
[14]
J
.
Aert
s
and
E
.
J
.
Marini
ss
en
,
“
Scan
Chai
n
Design
for
T
est
T
ime
R
educ
t
ion
in
Core
-
Based
ICs
,”
Pr
oce
ed
ings
IEEE
Inte
rnational
Te
st Conf
ere
n
ce (
ITC)
,
W
ashingt
on,
DC,
pp.
448
-
4
57
,
1998
.
[15]
P
.
Varm
a
and
S
.
Bhat
i
a
,
“
A
Struct
ure
d
T
est
Re
-
Us
e
Methodol
og
y
for
Core
-
B
a
sed
Sy
st
em
Chips
,”
Proceedi
ng
s
IEE
E
Inte
rnat
io
nal
Test
Con
fe
re
nce
(
ITC)
,
W
ashingt
on,
DC,
pp
.
294
-
302
,
1998
.
[16]
E
.
J
.
Marin
issen
,
et
al.
,
“
A
Struct
ure
d
a
nd
Sc
alable
Mec
h
ani
sm
for
Te
st
Ac
ce
s
s
to
Embedde
d
Reusabl
e
Cores
,”
Proce
ed
ings I
E
E
E
Int
ernati
onal
Test
Confe
ren
ce
(
ITC
)
,
W
ashingt
on,
DC,
pp.
284
-
293
,
1998
.
[17]
K
.
C
hakr
ab
arty
,
“
Design
of
S
y
stem
-
on
-
a
-
Chip
Te
st
Ac
ce
ss
Archi
t
ec
tur
es
U
sing
Inte
g
er
L
inea
r
Program
m
ing
,”
Proce
ed
ings I
E
E
E
VLSI
Test
Sym
posium (
VTS
)
,
Montrea
l
,
C
an
ad
a,
pp
.
127
-
1
34
,
2
000.
[18]
S.
K.
H.
Ahammad,
et
al.,
“
Fast
and
ac
cur
at
e
fea
tur
e
ext
r
ac
t
io
n
-
base
d
segm
ent
at
ion
fr
amework
for
spinal
cor
d
inj
ur
y
sev
eri
t
y
c
l
assific
a
ti
on
,”
I
E
EE
Ac
c
ess
,
vo
l.
7,
pp
.
46092
-
46
103
, 2
019
.
[19]
K.
Saikumar,
e
t
al
.
,
“
A
dee
p
learni
ng
proc
ess
for
spine
and
he
art
segm
ent
a
ti
on
using
pix
el
-
b
ase
d
convol
u
ti
ona
l
net
works
,”
Jour
nal
of
Inte
rnat
io
nal
Pharmace
u
tical Re
search
,
vo
l.
46
,
pp
.
278
-
28
2
,
2019
.
[20]
V.
Raj
esh,
e
t
a
l.
,
“
A
te
le
m
edicine
t
ec
hnolog
y
for
ca
rdiova
sc
ula
r
patient
s
di
agnosis
fea
ture
using
knn
-
m
p
m
al
gorit
hm
,”
Journal
of
Inte
rnat
io
nal
Pharmace
u
tical Re
search
,
vo
l.
46
,
pp
.
72
-
77
,
2019
.
[21]
Sengupta
,
A.
,
Dasgupta,
S.
,
Singh,
V.,
Sharm
a,
R.
,
&
Kum
ar
Vishvaka
rm
a,
S.
(Eds.
).
(2019).
VLSI
Design
and
Te
st.
Com
m
unic
at
ions
in
Com
pute
r and
Inform
ati
on
Scie
n
ce.
doi
:
10.
1007/978
-
98
1
-
32
-
9767
-
8
[22]
Acha
r
y
a
,
M.,
Pa
l,
C.
,
Maity
,
S.,
Chakra
bar
ti
,
A
.
,
“
Ine
xac
t
implem
ent
a
ti
on
of
wave
le
t
tr
ansform
and
it
s
per
form
ance
eva
lu
at
ion
throu
gh
bit
width
red
uct
ion
,
”
In
Chakra
bar
t
i,
A.,
Shar
m
a,
N.,
Bal
as,
V.E
.
(ed
s
.
)
Advanc
es
in
Com
puti
ng
Applic
a
ti
ons,
pp
.
227
–
242
.
Sprin
ger
,
Singapor
e
(
2016).
ht
tps:/
/do
i.
org/10
.
1007/97
8
-
981
-
10
-
2630
-
0_14
.
[23]
Venka
ta
r
aman
i,
S.,
Ro
y
,
K.
,
Raghuna
th
an,
A.,
“
Approxim
at
e
computin
g
,”
Tutor
ial
pr
ese
ntat
ion
a
t
2
9th
Inte
rnational
Co
nfe
renc
e
on
VLS
I
Design
,
Kolkata,
Ind
ia,
4
–
8
Jan
uar
y
2016
(2016
).
[24]
Gupta,
V.
,
Moh
apa
tr
a,
D
.
,
R
ag
hunat
han
,
A.
,
R
o
y
,
K.,
“
Low
-
power
digi
t
al
sig
nal
proc
essing
u
sing
appr
oximat
e
adde
rs,”
IE
EE Tr
ans.
Comput.
A
ide
d
Des.
Int
egr.
Circui
ts
Syst
.
3
2(1),
124
–
137
(2
013).
[25]
J
ia
ng,
H.
,
Han,
J.,
Lombardi,
F.
,
“
A
compara
ti
ve
rev
ie
w
an
d
e
v
aluati
on
of
appr
ox
imate
add
ers,
”
I
n:
GL
SVLSI
2015
Proce
ed
ings o
f
t
he
25th
Edition
on
Gr
eat
Lak
es
Symposium on
V
LSI
,
pp.
343
–
34
8
(2015).
[26]
Jiang,
H.,
Li
u
,
C.
,
Mahe
sh
wari
,
N.,
Lombardi,
F.,
Han,
J.,
“
A
Com
par
at
ive
stu
d
y
of
Approxim
at
e
Add
ers
and
Multi
pliers.
A
T
utori
al”
,
Univer
s
ity
of
Alber
ta
,
2
015.
[27]
Gogoi,
A.,
Kum
ar,
V.,
“
Design
of
low
power
,
are
a
eff
icient
and
high
spee
d
appr
oxima
te
a
dder
s
for
ine
xa
ct
computing,
”
In:
2016
Inte
rnatio
nal
Confe
renc
e
of
Signal
Proces
sing
and
Com
municat
ion
(
ICSC)
,
Noida,
India,
26
–
28
Dec
embe
r
2016.
IEEE
Xl
opre
(2016)
.
Evaluation Warning : The document was created with Spire.PDF for Python.