I
nd
o
ne
s
ia
n J
o
urna
l o
f
E
lect
rica
l En
g
ineering
a
nd
Co
m
pu
t
er
Science
Vo
l.
37
,
No
.
3
,
Ma
r
ch
20
25
,
p
p
.
1
7
8
5
~1
796
I
SS
N:
2
502
-
4
7
52
,
DOI
: 1
0
.
1
1
5
9
1
/ijee
cs
.v
37.
i
3
.
p
p
1
7
8
5
-
1
7
9
6
1785
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ee
cs
.
ia
esco
r
e.
co
m
Co
mm
unica
tion i
nduced
checkpo
i
nting ba
sed
fault
tolera
nce
mecha
nism
using
deep
-
lea
rni
ng
in
I
o
T
a
pplica
tions
So
wj
a
ny
a
L
a
k
s
hm
i A
1
,
Va
ni
priy
a
Ch
2
1
D
e
p
a
r
t
me
n
t
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
a
n
d
E
n
g
i
n
e
e
r
i
n
g
,
S
i
r
M
.
V
i
s
v
e
s
v
a
r
a
y
a
I
n
st
i
t
u
t
e
o
f
T
e
c
h
n
o
l
o
g
y
,
B
e
n
g
a
l
u
r
u
,
I
n
d
i
a
2
D
e
p
a
r
t
me
n
t
o
f
M
a
st
e
r
o
f
C
o
m
p
u
t
e
r
A
p
p
l
i
c
a
t
i
o
n
,
S
i
r
M
.
V
i
s
v
e
s
v
a
r
a
y
a
I
n
st
i
t
u
t
e
o
f
Te
c
h
n
o
l
o
g
y
,
B
e
n
g
a
l
u
r
u
,
I
n
d
i
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
u
n
11
,
2
0
2
4
R
ev
is
ed
Oct
5
,
202
4
Acc
ep
ted
Oct
9
,
2
0
2
4
In
tern
e
t
o
f
t
h
i
n
g
s
(I
o
T)
is
i
n
c
re
a
sin
g
l
y
u
se
d
in
d
iv
e
rse
e
n
v
iro
n
m
e
n
ts
su
c
h
a
s
h
e
a
lt
h
c
a
re
,
i
n
d
u
stry
a
n
d
a
g
r
icu
lt
u
re
.
T
h
e
y
c
a
rry
a
risk
o
f
a
d
v
e
rs
e
e
ffe
c
ts
if
th
e
y
m
a
k
e
d
e
c
isio
n
s
b
a
se
d
o
n
fa
u
lt
y
i
n
fo
rm
a
ti
o
n
.
S
o
f
twa
re
fa
u
lt
s,
e
sp
e
c
ially
tran
sie
n
t
fa
u
lt
s
a
re
a
p
r
ima
ry
c
o
n
tri
b
u
to
r
t
o
d
e
ficie
n
t
d
e
c
isio
n
-
m
a
k
in
g
.
T
h
e
e
x
isti
n
g
fa
u
lt
t
o
lera
n
t
m
e
c
h
a
n
ism
s o
ften
su
ffe
r
fr
o
m
c
h
e
c
k
p
o
in
t
o
v
e
rh
e
a
d
s a
s
c
h
e
c
k
p
o
i
n
ts
a
re
p
lac
e
d
in
a
ll
t
h
e
n
o
d
e
s.
Th
is
p
a
p
e
r
d
e
sc
rib
e
s
a
n
o
v
e
l
c
o
m
m
u
n
ica
ti
o
n
in
d
u
c
e
d
c
h
e
c
k
p
o
in
ti
n
g
b
a
se
d
fa
u
lt
t
o
lera
n
c
e
m
e
c
h
a
n
ism
(CIC
-
F
TM
)
d
e
sig
n
e
d
to
e
fficie
n
tl
y
re
c
o
v
e
r
fro
m
tran
sie
n
t
fa
u
lt
s,
wh
i
le
m
in
imiz
in
g
u
se
les
s
a
n
d
f
o
rc
e
d
c
h
e
c
k
p
o
in
ts.
L
o
n
g
sh
o
rt
-
term
m
e
m
o
ry
(LS
TM
)
b
a
se
d
d
e
e
p
lea
rn
i
n
g
a
l
g
o
rit
h
m
is
u
se
d
i
n
o
u
r
a
p
p
ro
a
c
h
t
o
p
r
e
d
ict
fa
u
l
t
o
c
c
u
rre
n
c
e
s
a
n
d
stra
teg
ica
ll
y
p
lac
e
c
h
e
c
k
p
o
i
n
ts.
T
h
e
p
r
o
p
o
se
d
m
e
th
o
d
a
lso
in
tu
rn
imp
r
o
v
e
s
y
ste
m
re
li
a
b
il
it
y
a
n
d
p
e
rf
o
rm
a
n
c
e
.
Ex
p
e
rime
n
t
a
l
re
su
lt
s
d
e
m
o
n
stra
te
th
e
e
ffe
c
ti
v
e
n
e
ss
o
f
p
ro
p
o
se
d
CIC
-
F
TM
in
Io
T
e
n
v
ir
o
n
m
e
n
t
b
y
m
in
imiz
in
g
t
h
e
p
ra
c
ti
c
a
b
le
o
p
e
ra
ti
n
g
ti
m
e
f
o
r
c
h
e
c
k
p
o
in
ti
n
g
a
n
d
b
a
c
k
p
ro
p
a
g
a
ti
o
n
,
c
o
m
p
a
re
d
to
trad
it
io
n
a
l
fa
u
lt
-
t
o
lera
n
c
e
m
e
c
h
a
n
ism
s.
K
ey
w
o
r
d
s
:
C
h
ec
k
p
o
in
t a
t in
ter
m
e
d
iate
node
C
o
m
m
u
n
icatio
n
in
d
u
ce
d
ch
ec
k
p
o
i
n
tin
g
b
ased
f
au
lt
to
ler
an
ce
m
ec
h
a
n
is
m
L
o
n
g
s
h
o
r
t
-
ter
m
m
e
m
o
r
y
alg
o
r
ith
m
T
r
an
s
ien
t f
au
lts
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
:
So
wjan
y
a
L
ak
s
h
m
i A
Dep
ar
tm
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
an
d
E
n
g
in
ee
r
in
g
,
Sir.
M
.
Vis
v
esv
ar
ay
a
I
n
s
titu
te
o
f
T
ec
h
n
o
lo
g
y
B
en
g
alu
r
u
-
5
6
2
1
5
7
,
I
n
d
ia
E
m
ail:
s
o
wjan
y
a.
en
g
g
@
g
m
ail
.
co
m
1.
I
NT
RO
D
UCT
I
O
N
T
h
e
in
ter
n
et
o
f
t
h
in
g
s
(
I
o
T
)
ap
p
licatio
n
s
ar
e
p
er
v
asiv
e
in
v
ar
io
u
s
d
o
m
ain
s
s
u
ch
as
s
m
ar
t
h
o
m
es,
in
d
u
s
tr
ies,
ag
r
icu
ltu
r
e
an
d
h
e
alth
ca
r
e.
T
h
ese
a
p
p
licatio
n
s
r
ely
h
ea
v
ily
o
n
th
e
in
teg
r
ity
o
f
a
d
ata
tr
an
s
ac
tio
n
s
am
o
n
g
in
ter
c
o
n
n
ec
te
d
d
ev
ices
to
f
u
n
ctio
n
co
r
r
ec
tly
[
1
]
.
I
n
te
r
ac
tio
n
s
o
cc
u
r
th
r
o
u
g
h
m
ess
ag
e
ex
ch
an
g
es,
with
o
p
er
atio
n
s
an
d
d
ec
is
io
n
s
r
elia
n
t
o
n
d
ata
tr
an
s
m
itted
an
d
r
e
ce
iv
ed
ac
r
o
s
s
I
o
T
n
o
d
es
[
2
]
,
[
3
]
.
Sen
s
o
r
s
c
o
llect
u
n
s
tr
u
ctu
r
ed
d
ata
,
later
co
n
v
er
ted
to
s
tr
u
ctu
r
ed
in
p
u
t
f
o
r
I
o
T
p
r
o
ce
s
s
in
g
u
n
its
.
Pro
ce
s
s
o
r
s
,
l
in
k
ed
v
ia
s
tan
d
a
r
d
n
etwo
r
k
s
,
ex
ch
a
n
g
e
m
ess
ag
es
th
r
o
u
g
h
d
ed
icate
d
c
h
an
n
els
t
o
m
in
im
ize
co
m
m
u
n
icatio
n
d
elay
s
[
4
]
-
[
7
]
.
Fo
g
co
m
p
u
tin
g
,
p
o
s
itio
n
in
g
co
m
p
u
tatio
n
al
d
ev
ices
clo
s
er
to
I
o
T
d
ev
ices,
aid
s
tim
e
-
cr
itical
ap
p
licatio
n
s
.
Data
p
r
o
ce
s
s
in
g
o
cc
u
r
s
in
f
o
g
co
m
p
u
tin
g
i
n
s
tead
o
f
clo
u
d
co
m
p
u
ti
n
g
f
o
r
s
wif
t
d
ec
is
io
n
-
m
ak
i
n
g
.
T
h
e
ar
c
h
itectu
r
e
d
iag
r
am
,
Fig
u
r
e
1
f
o
r
ag
r
icu
lt
u
r
e
I
o
T
d
ep
icts
th
is
s
etu
p
.
I
o
T
s
y
s
tem
s
ar
e
p
r
o
n
e
to
tr
an
s
ien
t
f
au
lts
d
u
r
in
g
d
ata
tr
an
s
ac
tio
n
s
,
wh
ich
ca
n
lead
to
in
co
r
r
ec
t
o
p
e
r
atio
n
s
an
d
p
o
ten
tial
s
y
s
tem
d
am
ag
e.
So
f
tw
ar
e
tr
an
s
i
en
t
f
au
lts
ar
is
e
f
r
o
m
is
s
u
es
lik
e
d
ata
tr
a
n
s
ac
tio
n
s
,
s
er
v
er
f
ailu
r
es
an
d
b
r
ea
ch
es
[
8
]
-
[
1
0
]
ca
n
d
is
r
u
p
t
I
o
T
s
y
s
tem
s
b
u
t
ar
e
ty
p
ically
r
ec
o
v
er
ab
le
b
y
r
estar
tin
g
task
s
.
T
r
an
s
ien
t
f
a
u
lts
lik
e
s
eq
u
en
ce
n
u
m
b
er
,
c
h
ec
k
s
u
m
,
n
u
ll
c
h
ar
ac
ter
a
n
d
out
-
of
-
r
an
g
e
er
r
o
r
s
th
at
co
n
tr
i
b
u
te
to
r
is
k
s
[
7
]
,
[
10
]
-
[
1
3
]
ar
e
d
etec
ted
in
th
is
p
ap
e
r
’
s
f
ir
s
t c
o
n
tr
ib
u
tio
n
.
Var
io
u
s
ch
ec
k
p
o
in
tin
g
b
ased
f
au
lt
to
ler
an
ce
m
ec
h
an
is
m
s
(
FTM
)
ar
e
ef
f
ec
tiv
ely
r
ec
o
v
er
i
n
g
f
au
lts
,
b
u
t
n
o
t
ad
d
r
ess
in
g
tr
an
s
ien
t
f
au
lt
s
an
d
also
r
ec
o
v
er
y
is
at
th
e
co
s
t
o
f
ch
ec
k
p
o
in
tin
g
o
v
er
h
ea
d
s
an
d
s
to
r
ag
e
o
v
er
h
ea
d
s
.
T
h
e
f
o
llo
win
g
o
b
s
er
v
atio
n
s
ar
e
f
o
u
n
d
in
th
e
liter
atu
r
e
r
ev
iew
-
v
ar
i
o
u
s
s
u
cc
ess
f
u
l
FTM
s
u
tili
ze
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
37
,
No
.
3
,
Ma
r
ch
20
25
:
1
785
-
1
7
9
6
1786
r
o
llb
ac
k
v
ia
ch
ec
k
p
o
in
tin
g
,
ca
teg
o
r
ized
as
u
n
-
c
o
o
r
d
in
ated
,
co
o
r
d
in
ated
an
d
co
m
m
u
n
icatio
n
i
n
d
u
ce
d
ch
ec
k
p
o
i
n
tin
g
(
C
I
C
)
b
ased
FTM
s
[
6
]
,
[
10
]
,
[
14
]
-
[
2
1
]
.
C
h
ec
k
p
o
in
tin
g
s
av
es
cr
u
cial
in
f
o
r
m
atio
n
tem
p
o
r
a
r
ily
,
in
clu
d
in
g
co
d
e
,
d
ata,
s
tatu
s
,
r
e
g
is
ter
co
n
ten
ts
,
e
n
v
ir
o
n
m
en
t
c
o
n
d
itio
n
s
,
ac
ce
s
s
co
u
n
ts
o
f
th
e
f
iles
,
f
ile
p
o
in
ter
s
an
d
f
ile
r
elate
d
d
etails,
en
s
u
r
i
n
g
co
n
s
is
ten
t
s
y
s
tem
s
tate
[
1
0
]
,
[
2
2
]
.
I
n
u
n
-
co
o
r
d
in
ate
d
ch
ec
k
p
o
in
tin
g
FTM
s
,
wh
er
e
ch
ec
k
p
o
in
ts
ar
e
s
et
at
f
ix
ed
in
ter
v
als
o
r
s
tr
ateg
ically
,
in
d
ep
en
d
en
t
o
f
p
r
o
ce
s
s
o
r
o
p
er
atio
n
s
,
ca
u
s
in
g
in
ef
f
icien
cy
d
u
e
to
in
co
n
s
is
ten
t
n
o
d
es
an
d
th
e
Do
m
in
o
ef
f
ec
t
[
8
]
,
[
10
]
,
[
11
]
,
[
23
]
-
[
2
5
]
.
C
o
o
r
d
in
ated
ch
ec
k
p
o
i
n
tin
g
s
y
n
ch
r
o
n
izes
p
r
o
ce
s
s
es
th
r
o
u
g
h
s
y
s
tem
m
ess
a
g
es
to
en
s
u
r
e
a
co
n
s
is
ten
t
g
lo
b
al
s
tate,
alb
eit
w
ith
s
to
r
ag
e
o
v
er
h
ea
d
s
[
8
]
,
[
1
0
]
.
C
I
C
co
m
b
in
es
asp
ec
ts
o
f
b
o
th
tech
n
iq
u
es,
p
lacin
g
ch
ec
k
p
o
in
ts
b
ased
o
n
ap
p
licatio
n
m
ess
ag
es
to
av
o
id
u
n
n
ec
ess
ar
y
o
n
es
b
u
t
r
is
k
in
g
m
is
s
in
g
u
s
ef
u
l
o
n
es,
n
ec
ess
itatin
g
f
o
r
ce
d
ch
ec
k
p
o
in
ts
p
lace
m
en
ts
f
o
r
f
a
u
lt
r
ec
o
v
er
y
[
5
]
,
[
7
]
,
[
9
]
,
[
2
5
]
.
C
I
C
allo
ws
d
esig
n
er
s
to
p
lace
ch
ec
k
p
o
in
t
s
b
ased
o
n
s
p
ec
if
ic
co
n
d
itio
n
s
c
o
n
v
e
y
ed
b
y
ex
p
l
icit
m
ess
ag
es
[
8
]
,
[
1
8
]
,
[
20
]
,
[
23
]
-
[
29
]
.
W
id
ely
u
s
ed
in
v
ar
io
u
s
ap
p
licatio
n
s
in
clu
d
in
g
p
ar
allel
an
d
d
is
tr
ib
u
ted
co
m
p
u
tin
g
,
C
I
C
-
FTM
s
ar
e
u
n
d
er
co
n
s
tan
t r
esear
ch
,
esp
e
cially
in
I
o
T
.
Fig
u
r
e
1
.
Fo
g
e
n
ab
le
d
clo
u
d
in
f
r
astru
ctu
r
e
E
les
et
a
l
.
[
2
4
]
in
teg
r
ates
er
r
o
r
d
etec
tio
n
an
d
e
q
u
id
is
tan
t
ch
ec
k
p
o
in
tin
g
with
r
o
llb
ac
k
r
ec
o
v
er
y
a
n
d
ac
tiv
e
r
ep
licatio
n
to
h
an
d
le
f
a
u
lts
in
to
s
o
f
twar
e
ar
ch
itectu
r
e,
aim
in
g
to
m
ee
t
p
er
f
o
r
m
an
ce
an
d
co
s
t
co
n
s
tr
ain
ts
in
s
af
ety
-
cr
itical
ap
p
licatio
n
s
.
T
h
is
ap
p
r
o
ac
h
ass
u
m
es
a
lim
ited
n
u
m
b
er
o
f
f
au
lts
an
d
lea
d
s
to
p
o
ten
tial
tim
e
o
v
er
h
ea
d
s
ev
en
th
o
u
g
h
it
p
r
i
m
ar
ily
h
an
d
les
tr
an
s
ien
t
f
au
lts
.
Hela
r
y
et
a
l.
[
2
5
]
f
o
c
u
s
es
o
n
o
p
tim
izin
g
th
e
ch
ec
k
p
o
i
n
tin
g
m
ec
h
an
is
m
in
d
is
tr
ib
u
ted
s
y
s
tem
s
b
y
id
en
tify
i
n
g
an
d
elim
in
atin
g
r
e
d
u
n
d
an
t
c
h
ec
k
p
o
i
n
ts
th
r
o
u
g
h
a
co
m
m
u
n
icatio
n
-
b
ased
FTM
ap
p
r
o
ac
h
.
T
h
is
m
eth
o
d
h
as
i
n
cr
ea
s
ed
co
m
p
le
x
ity
an
d
d
if
f
i
cu
lties
in
d
y
n
am
ic
en
v
ir
o
n
m
en
ts
.
T
h
e
co
n
tin
u
ed
r
esear
ch
[
2
6
]
f
o
cu
s
o
n
im
p
lem
e
n
tatio
n
o
f
C
I
C
tech
n
iq
u
es
to
d
eter
m
in
e
co
n
s
is
ten
t
s
n
ap
s
h
o
ts
ef
f
icien
tly
b
y
in
iti
atin
g
ch
ec
k
p
o
in
ts
th
r
o
u
g
h
m
ar
k
er
m
ess
ag
es,
lo
g
g
in
g
m
es
s
ag
es
in
tr
an
s
it
an
d
en
s
u
r
in
g
co
n
s
is
ten
cy
th
r
o
u
g
h
r
eg
u
lar
m
o
n
ito
r
in
g
.
Hen
ce
ch
ec
k
p
o
in
t
o
v
er
h
ea
d
s
ar
e
r
ed
u
ce
d
an
d
th
e
ap
p
r
o
ac
h
is
ad
ap
tab
le
to
b
o
th
ce
n
tr
alize
d
an
d
d
ec
e
n
tr
alize
d
s
y
s
tem
s
b
u
t,
s
till
f
ac
es c
o
m
p
lex
ity
is
s
u
e
s
in
im
p
lem
en
tatio
n
to
f
u
r
n
is
h
th
e
co
n
s
tr
ain
ts
an
d
n
o
t
ef
f
ec
tiv
ely
h
an
d
lin
g
c
o
m
p
lex
f
ailu
r
e
s
ce
n
ar
i
o
s
.
Gar
cia
e
t
a
l.
[
2
2
]
,
[
30
]
,
an
d
Vieir
a
et
a
l.
[
31
]
ex
p
l
o
r
es
a
n
ad
v
a
n
ce
d
c
h
ec
k
p
o
in
tin
g
te
ch
n
iq
u
e
d
esig
n
ed
to
e
n
h
an
ce
f
au
lt
to
ler
a
n
ce
in
d
is
tr
ib
u
ted
s
y
s
tem
s
u
s
in
g
lo
ca
l
ch
ec
k
p
o
in
ts
an
d
r
ec
o
v
er
y
lin
es,
wh
er
e
ea
ch
p
r
o
ce
s
s
in
d
ep
en
d
en
tly
m
an
ag
es
its
ch
ec
k
p
o
i
n
ts
b
ased
o
n
lo
ca
l
k
n
o
wled
g
e
an
d
co
o
r
d
in
atio
n
is
u
s
ed
to
estab
lis
h
co
n
s
is
ten
t
g
lo
b
al
r
ec
o
v
e
r
y
lin
es.
T
h
is
ap
p
r
o
ac
h
r
ed
u
ce
c
h
ec
k
p
o
in
tin
g
r
elate
d
o
v
e
r
h
ea
d
s
a
n
d
c
o
m
p
lex
ity
co
m
p
ar
ed
to
tr
ad
itio
n
al
g
lo
b
al
ch
ec
k
p
o
i
n
tin
g
m
eth
o
d
s
,
b
u
t
f
ac
es
co
o
r
d
in
atio
n
co
m
p
le
x
ity
an
d
ch
allen
g
es
in
h
an
d
lin
g
p
ar
tial
f
ailu
r
es.
Simo
n
et
a
l.
[
2
1
]
in
tr
o
d
u
ce
s
d
elay
ed
C
I
C
(
D
C
I
C
)
b
ased
FT
M
to
im
p
r
o
v
e
p
er
f
o
r
m
an
ce
in
d
is
tr
ib
u
ted
s
y
s
tem
s
b
y
s
tr
ateg
ically
p
lacin
g
ch
ec
k
p
o
in
ts
b
ased
o
n
co
m
m
u
n
icatio
n
d
elay
s
r
ath
er
th
an
f
ix
ed
in
ter
v
als.
T
h
is
ap
p
r
o
ac
h
en
h
an
ce
s
f
a
u
lt
to
ler
a
n
ce
an
d
s
e
lf
-
h
ea
lin
g
ca
p
ab
i
liti
es
an
d
m
in
im
izes
th
e
p
er
f
o
r
m
a
n
ce
o
v
er
h
e
ad
s
b
u
t
en
co
u
n
ter
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
C
o
mmu
n
ica
tio
n
in
d
u
ce
d
c
h
ec
kp
o
in
tin
g
b
a
s
ed
fa
u
lt to
lera
n
c
e
mec
h
a
n
is
m
u
s
in
g
…
(
S
o
w
ja
n
ya
La
ksh
mi
A
)
1787
u
s
eless
ch
ec
k
p
o
in
ts
.
T
h
is
ap
p
r
o
ac
h
is
also
n
o
t
ad
d
r
ess
in
g
o
v
e
r
all
f
au
lt
s
ce
n
ar
io
s
i
n
d
is
tr
ib
u
te
d
s
y
s
tem
s
.
R
o
b
er
to
B
ald
o
n
i
et
a
l.
[
5
]
p
r
o
p
o
s
e
an
i
n
d
ex
-
b
ased
C
I
C
alg
o
r
ith
m
t
o
r
e
d
u
ce
ch
ec
k
p
o
in
tin
g
o
v
e
r
h
ea
d
s
,
en
h
an
ce
ef
f
icien
cy
in
d
is
tr
ib
u
ted
s
y
s
tem
s
with
o
u
t
ce
n
tr
alize
d
co
n
t
r
o
l.
Desp
i
te
its
in
n
o
v
ativ
e
ap
p
r
o
ac
h
,
t
h
e
alg
o
r
ith
m
f
ac
es
lim
itatio
n
s
s
u
ch
as
p
o
ten
tial
o
v
er
h
ea
d
,
im
p
lem
en
tatio
n
co
m
p
lex
ities
,
s
ca
lab
ilit
y
co
n
ce
r
n
s
,
lim
ited
f
au
lt
m
o
d
el
co
v
er
ag
e,
s
ig
n
if
ican
t r
ec
o
v
er
y
tim
e
an
d
r
eso
u
r
ce
u
tili
za
tio
n
i
s
s
u
es.
L
u
o
an
d
Ma
n
i
v
an
n
a
n
[
2
7
]
in
tr
o
d
u
ce
s
a
f
u
lly
i
n
f
o
r
m
ed
a
n
d
e
f
f
icien
t
(
FIN
E
)
C
I
C
p
r
o
to
co
l
d
esig
n
ed
to
m
in
im
ize
ch
ec
k
p
o
in
tin
g
o
v
er
h
ea
d
b
y
r
ed
u
cin
g
f
o
r
ce
d
ch
ec
k
p
o
in
ts
an
d
a
v
o
id
in
g
th
e
d
o
m
in
o
ef
f
ec
t,
en
s
u
r
in
g
co
o
r
d
in
ate
d
r
ec
o
v
er
y
th
r
o
u
g
h
a
f
u
lly
in
f
o
r
m
e
d
g
lo
b
al
s
tate.
T
h
is
ap
p
r
o
ac
h
s
till
f
ac
es
im
p
lem
en
tatio
n
co
m
p
lex
ity
is
s
u
es
a
lo
n
g
with
co
m
m
u
n
icatio
n
o
v
er
h
ea
d
an
d
n
etwo
r
k
d
ep
en
d
en
cy
,
in
ad
d
itio
n
.
Ah
n
[
9
]
p
r
o
p
o
s
es
a
FTM
th
at
co
m
b
i
n
es
C
I
C
wit
h
m
ess
ag
e
lo
g
g
i
n
g
to
o
v
e
r
co
m
e
th
e
lim
itatio
n
s
in
ex
is
tin
g
FTM
s
in
d
is
tr
ib
u
ted
s
y
s
tem
s
.
Hi
s
m
eth
o
d
o
lo
g
y
in
v
o
lv
es
f
o
r
cin
g
ch
ec
k
p
o
in
ts
b
ased
o
n
c
o
m
m
u
n
icatio
n
p
atter
n
s
an
d
lo
g
g
in
g
all
m
ess
ag
es
f
o
r
r
ep
lay
d
u
r
in
g
r
e
co
v
er
y
a
n
d
ad
v
an
ce
d
m
ec
h
an
i
s
m
s
to
h
an
d
le
d
ep
e
n
d
en
cies.
Desp
ite
its
b
en
ef
its
,
th
e
ap
p
r
o
ac
h
f
ac
es
u
s
eless
ch
ec
k
p
o
i
n
t
o
v
er
h
ea
d
s
,
r
ec
o
v
er
y
laten
cy
an
d
d
e
p
en
d
e
n
cy
m
an
ag
em
en
t
is
s
u
es.
Ma
lh
o
tr
a
an
d
B
ala
[
1
7
]
p
r
o
p
o
s
e
a
C
I
C
FTM
tailo
r
ed
f
o
r
I
o
T
s
y
s
tem
s
.
T
h
e
p
r
o
to
c
o
l
is
a
co
m
b
i
n
atio
n
o
f
s
p
o
n
tan
eo
u
s
c
h
ec
k
p
o
in
tin
g
wh
er
e
ea
ch
n
o
d
e
tak
es
s
p
o
n
tan
eo
u
s
ch
ec
k
p
o
in
ts
b
ased
o
n
a
l
o
g
is
tic
f
u
n
ctio
n
th
at
esti
m
ates
th
e
tim
e
in
ter
v
al
b
et
wee
n
ch
ec
k
p
o
in
ts
an
d
co
o
r
d
in
ate
d
ch
ec
k
p
o
in
tin
g
wh
er
e
n
o
d
es tak
e
co
o
r
d
in
ated
ch
ec
k
p
o
i
n
ts
u
s
in
g
th
e
T
ak
a
g
i
-
Su
g
en
o
(
T
-
S)
f
u
zz
y
s
y
s
tem
,
wh
ich
g
en
e
r
ates
r
esu
lts
b
ased
o
n
d
ef
in
ite
-
3
9
r
u
les.
T
h
is
s
y
s
tem
u
s
es
p
ar
am
eter
s
s
u
ch
as
en
e
r
g
y
,
f
ailu
r
e
r
ate
a
n
d
r
ec
eiv
ed
s
ig
n
al
s
tr
en
g
th
i
n
d
ica
to
r
(
R
SS
I
)
to
a
v
o
id
u
n
n
ec
ess
ar
y
c
h
ec
k
p
o
in
ts
.
T
h
is
ap
p
r
o
ac
h
m
in
im
izes th
e
n
u
m
b
er
o
f
ch
ec
k
p
o
in
ts
,
s
y
s
tem
o
v
er
h
ea
d
a
n
d
en
s
u
r
in
g
non
-
b
lo
ck
in
g
p
r
o
ce
s
s
es
d
u
r
in
g
ch
ec
k
p
o
in
tin
g
.
I
n
th
is
p
er
m
a
n
en
t
ch
ec
k
p
o
in
ts
ar
e
s
to
r
ed
o
n
I
o
T
d
ev
ices
wh
ic
h
also
in
d
ir
ec
tly
lead
s
to
co
n
s
id
er
ab
le
u
s
eless
ch
ec
k
p
o
in
ts
af
ter
f
au
lt
r
ec
o
v
er
y
an
d
also
f
au
l
t
r
ec
o
v
er
y
b
ased
o
n
s
p
ec
if
ic
p
ar
am
eter
s
d
ep
en
d
e
n
c
y
an
d
s
in
ce
,
th
is
p
r
o
to
co
l
is
d
esig
n
ed
s
p
ec
if
ically
f
o
r
m
o
b
ile
d
is
tr
ib
u
ted
s
y
s
tem
s
,
h
as
its
p
er
f
o
r
m
an
ce
is
h
i
g
h
l
y
d
y
n
am
ic
o
r
lar
g
e
-
s
ca
le
n
et
wo
r
k
s
with
s
ig
n
if
ican
t
n
u
m
b
er
o
f
n
o
d
es
is
n
o
t
th
o
r
o
u
g
h
ly
e
v
alu
ated
.
J
ag
g
i
a
n
d
Sin
g
h
[
1
3
]
in
t
r
o
d
u
ce
an
a
d
ap
tiv
e
ch
ec
k
p
o
in
tin
g
tech
n
iq
u
e
to
en
h
an
ce
f
au
lt
to
ler
an
ce
in
m
o
b
ile
co
m
p
u
tin
g
g
r
id
(
Mo
G)
.
T
h
is
ap
p
r
o
ac
h
u
s
es
co
o
p
er
ativ
e
ch
ec
k
p
o
in
tin
g
,
wh
er
e
n
o
d
es
in
t
h
e
s
y
s
tem
co
o
p
er
ate
to
s
to
r
e
ch
ec
k
p
o
in
t
d
ata.
I
f
a
n
o
d
e
lac
k
s
s
tab
le
s
to
r
ag
e,
it
u
s
es
th
e
s
to
r
ag
e
o
f
o
th
e
r
n
o
d
es.
T
h
e
ch
ec
k
p
o
i
n
t
d
ata
is
r
e
p
licated
a
cr
o
s
s
m
u
ltip
le
n
o
d
es
t
o
e
n
s
u
r
e
h
ig
h
er
ch
an
ce
s
o
f
r
ec
o
v
er
y
d
e
p
en
d
in
g
o
n
r
eso
u
r
c
e
av
ailab
ilit
y
.
T
h
e
ch
ec
k
p
o
in
tin
g
s
ch
em
e
a
d
ap
ts
b
ased
o
n
t
h
e
av
ailab
ilit
y
o
f
r
eso
u
r
ce
s
in
t
h
e
Mo
G
m
a
k
in
g
it
f
lex
ib
le
an
d
ef
f
icien
t.
B
u
t,
th
is
ap
p
r
o
ac
h
h
ig
h
ly
r
elies
o
n
th
e
a
v
ailab
ilit
y
o
f
s
tab
le
s
to
r
ag
e
an
d
r
eso
u
r
ce
s
in
o
th
er
n
o
d
es,
wh
ich
m
ig
h
t
n
o
t
alwa
y
s
b
e
g
u
ar
an
teed
in
h
ig
h
ly
d
y
n
a
m
ic
o
r
r
eso
u
r
ce
-
co
n
s
tr
a
in
ed
en
v
ir
o
n
m
e
n
ts
.
R
ep
licatin
g
ch
ec
k
p
o
in
tin
g
d
ata
at
m
u
l
t
i
p
le
n
o
d
e
s
c
a
n
l
e
a
d
t
o
i
n
c
r
e
a
s
e
d
n
e
t
w
o
r
k
t
r
a
f
f
i
c
a
n
d
o
v
e
r
h
e
a
d
a
n
d
a
s
t
h
e
n
u
m
b
e
r
o
f
n
o
d
e
s
i
n
c
r
ea
s
es
,
m
an
a
g
i
n
g
a
n
d
c
o
o
r
d
i
n
a
t
i
n
g
c
h
ec
k
p
o
i
n
t
s
ac
r
o
s
s
a
l
a
r
g
e
n
u
m
b
e
r
o
f
n
o
d
e
s
m
a
y
b
e
c
o
m
e
c
o
m
p
l
e
x
a
n
d
l
e
s
s
e
f
f
i
c
ie
n
t
.
T
an
et
a
l.
[
3
2
]
u
s
es
lo
n
g
s
h
o
r
t
-
ter
m
m
e
m
o
r
y
(
L
STM
)
n
et
wo
r
k
s
f
o
r
d
etec
tin
g
f
au
lts
in
n
o
n
-
lin
ea
r
d
y
n
am
ical
s
y
s
tem
s
.
T
h
e
m
et
h
o
d
o
lo
g
y
in
v
o
lv
es
tr
ain
in
g
L
S
T
M
m
o
d
el
o
n
n
o
r
m
al
o
p
er
ati
n
g
d
ata,
u
s
in
g
it
to
p
r
ed
ict
f
u
t
u
r
e
s
tates
an
d
d
etec
tin
g
p
r
e
d
ictiv
e
f
au
lts
.
T
h
is
ap
p
r
o
ac
h
ca
n
b
e
r
ed
esig
n
e
d
f
o
r
d
if
f
er
en
t
s
y
s
tem
s
o
n
r
eq
u
ir
em
e
n
t
b
asis
ev
en
th
o
u
g
h
it
is
ch
allen
g
in
g
.
Ab
s
ar
et
a
l.
[
3
3
]
ass
ess
th
e
u
s
e
o
f
L
STM
m
o
d
els
f
o
r
f
o
r
ec
asti
n
g
co
n
tag
io
u
s
d
is
ea
s
e
o
u
tb
r
ea
k
s
,
f
o
cu
s
in
g
o
n
u
tili
zin
g
h
is
to
r
i
ca
l
d
ata
f
o
r
ea
r
ly
d
etec
tio
n
.
T
h
e
m
eth
o
d
o
lo
g
y
in
v
o
lv
es
tr
ain
in
g
L
STM
m
o
d
e
ls
o
n
d
is
ea
s
e
d
ata,
m
a
k
in
g
p
r
e
d
ictio
n
s
an
d
e
v
alu
atin
g
th
e
m
o
d
el’
s
p
er
f
o
r
m
a
n
c
e
u
s
in
g
ac
cu
r
ac
y
m
etr
ics.
Z
h
a
n
g
et
a
l.
[
3
4
]
in
tr
o
d
u
ce
an
L
STM
-
b
ased
au
to
en
co
d
e
r
f
o
r
n
etwo
r
k
an
o
m
al
y
d
etec
tio
n
aim
in
g
to
id
en
tify
u
n
u
s
u
al
tr
af
f
ic
p
atter
n
s
with
h
ig
h
r
ec
o
n
s
tr
u
ctio
n
er
r
o
r
s
an
d
r
ec
o
n
s
tr
u
ct
n
o
r
m
a
l
n
etwo
r
k
tr
af
f
ic.
T
h
is
ap
p
r
o
ac
h
n
o
t
f
ea
s
ib
le
wit
h
r
esp
ec
t
to
p
r
ac
tical
ap
p
licab
ilit
y
a
n
d
i
n
teg
r
atio
n
in
r
ea
l
-
wo
r
l
d
s
ce
n
ar
io
s
.
W
an
g
et
a
l.
[
3
5
]
f
o
cu
s
to
im
p
r
o
v
e
p
r
ed
ictio
n
ac
c
u
r
ac
y
an
d
ef
f
icien
cy
o
f
an
L
S
T
M
m
o
d
el
u
s
ed
f
o
r
p
r
ed
ictin
g
tem
p
er
atu
r
e
i
n
d
ata
ce
n
ter
s
b
y
o
p
tim
izin
g
its
h
y
p
er
p
ar
am
eter
s
.
T
h
is
h
y
p
e
r
p
ar
a
m
eter
s
o
p
tim
izatio
n
tech
n
iq
u
e
ca
n
b
e
u
tili
ze
d
b
ased
o
n
d
o
m
ain
an
d
r
eq
u
ir
e
m
en
t
s
p
ec
if
ic
s
in
ce
it a
d
d
s
o
n
co
m
p
u
tatio
n
al
co
s
ts
.
T
h
e
d
etailed
s
u
r
v
ey
o
u
tlin
e
s
th
at
co
o
r
d
in
ated
an
d
u
n
-
co
o
r
d
in
ate
d
ch
ec
k
p
o
in
tin
g
e
n
co
u
n
ter
s
ch
ec
k
p
o
i
n
tin
g
o
v
er
h
ea
d
s
an
d
i
n
ef
f
ici
en
cies.
W
h
ile
ad
d
r
ess
in
g
th
e
c
h
allen
g
es
in
o
p
tim
izin
g
th
e
ch
ec
k
p
o
i
n
ti
n
g
b
ased
FTM
s
,
th
e
ex
is
tin
g
tech
n
iq
u
es
lik
e
DC
I
C
en
co
u
n
ter
s
u
s
eless
ch
ec
k
p
o
in
ts
,
in
d
ex
-
b
a
s
ed
C
I
C
alg
o
r
ith
m
s
f
ac
e
is
s
u
es lik
e
ch
ec
k
p
o
in
tin
g
o
v
er
h
ea
d
s
an
d
im
p
lem
e
n
tatio
n
co
m
p
le
x
ities
.
Ad
ap
t
iv
e
c
h
ec
k
p
o
in
tin
g
in
m
o
b
ile
co
m
p
u
tin
g
g
r
id
s
is
r
eso
u
r
ce
-
d
ep
en
d
en
t
a
n
d
m
ay
n
o
t
alwa
y
s
b
e
f
ea
s
ib
le
in
d
y
n
a
m
ic
en
v
ir
o
n
m
en
ts
lik
e
I
o
T
.
Pro
to
co
ls
tailo
r
ed
f
o
r
I
o
T
s
y
s
tem
s
n
ee
d
to
m
in
im
ize
n
u
m
b
er
o
f
ch
ec
k
p
o
in
ts
,
u
s
eless
ch
ec
k
p
o
in
ts
an
d
ef
f
icien
tly
r
ec
o
v
er
tr
an
s
i
en
t
f
a
u
lts
with
m
in
im
al
r
eso
u
r
ce
u
tili
za
tio
n
.
T
h
e
m
ai
n
co
n
tr
ib
u
tio
n
s
o
f
th
is
r
esear
ch
ar
e
as
f
o
llo
ws:
−
T
h
e
r
esear
ch
in
tr
o
d
u
ce
s
a
n
o
v
el
ch
ec
k
p
o
in
t
at
in
ter
m
ed
iat
e
n
o
d
e
(
C
I
N
)
C
I
C
-
FTM
f
o
r
I
o
T
en
v
ir
o
n
m
en
ts
th
at
s
tr
ateg
ically
p
lace
ch
ec
k
p
o
in
ts
o
n
ly
j
u
s
t
b
ef
o
r
e
p
r
e
d
icted
f
au
lty
n
o
d
es
b
y
in
teg
r
at
in
g
L
STM
-
d
ee
p
lear
n
in
g
alg
o
r
ith
m
s
.
T
h
is
m
in
i
m
iz
es
u
s
eless
an
d
f
o
r
ce
d
ch
ec
k
p
o
in
ts
,
u
n
lik
e
t
r
ad
itio
n
al
m
et
h
o
d
s
th
at
p
lace
ch
ec
k
p
o
i
n
ts
in
d
is
cr
im
in
ately
a
cr
o
s
s
all/m
an
y
n
o
d
es
,
a
n
d
also
ad
d
r
ess
es th
e
d
ef
icien
cies o
f
ex
is
tin
g
FTM
s.
−
Descr
ib
in
g
tr
an
s
ien
t f
au
lt o
cc
u
r
r
en
ce
s
ce
n
a
r
io
s
an
d
im
p
lem
en
tatio
n
o
f
t
h
e
f
au
lt
d
etec
tio
n
alg
o
r
ith
m
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
37
,
No
.
3
,
Ma
r
ch
20
25
:
1
785
-
1
7
9
6
1788
−
Ou
r
ap
p
r
o
ac
h
u
s
es
L
STM
-
d
ee
p
lear
n
in
g
f
o
r
tr
ain
i
n
g
u
s
in
g
h
is
to
r
y
o
f
f
au
lt
o
cc
u
r
r
e
n
ce
s
an
d
n
ew
m
ess
ag
e
lo
g
s
an
d
p
r
ed
ictin
g
n
ew
f
au
lt
o
cc
u
r
r
e
n
ce
.
T
h
e
r
em
ai
n
in
g
s
ec
tio
n
s
ar
e
ar
r
an
g
ed
as
f
o
llo
ws:
s
ec
tio
n
2
d
escr
ib
es
th
e
p
r
o
p
o
s
ed
m
eth
o
d
o
lo
g
y
th
at
in
clu
d
es
tr
an
s
ien
t
f
au
lt
d
etec
tio
n
,
L
STM
alg
o
r
ith
m
f
o
r
p
r
ed
ictin
g
f
au
lt
o
cc
u
r
r
en
ce
a
n
d
C
I
N
C
I
C
-
FT
M
o
p
er
atio
n
al
m
ec
h
an
is
m
f
o
r
p
lacin
g
ch
ec
k
p
o
in
ts
a
n
d
f
a
u
lt
r
ec
o
v
er
y
.
Sectio
n
3
ex
p
lain
s
e
x
p
er
im
en
tal
s
etu
p
,
s
ec
tio
n
4
ab
o
u
t r
esu
lts
an
aly
s
is
an
d
f
in
ally
s
ec
tio
n
5
co
n
cl
u
d
es th
e
r
esear
ch
wo
r
k
.
2.
M
E
T
H
O
D
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
o
lo
g
y
in
teg
r
ates
n
o
v
el
C
I
N
C
I
C
-
FTM
t
h
at
s
tr
ateg
ically
p
lace
ch
ec
k
p
o
in
ts
b
ased
o
n
L
STM
p
r
ed
ictio
n
s
th
at
m
in
im
izes c
h
ec
k
p
o
in
tin
g
o
v
e
r
h
ea
d
s
an
d
en
h
an
ce
s
s
y
s
tem
p
er
f
o
r
m
an
ce
b
y
tr
an
s
ien
t
f
au
lt
r
ec
o
v
er
y
.
T
h
e
s
tep
-
by
-
s
tep
d
escr
ip
tio
n
s
tar
ts
with
s
y
s
t
em
ar
ch
itectu
r
e
d
iag
r
am
,
Fig
u
r
e
2
s
tar
ts
f
r
o
m
d
ata
co
llectio
n
,
p
r
e
-
p
r
o
ce
s
s
in
g
th
at
in
clu
d
e
h
an
d
lin
g
m
is
s
in
g
v
al
u
es,
r
em
o
v
in
g
o
u
tlier
s
a
n
d
e
n
s
u
r
in
g
co
n
s
is
ten
cy
,
tr
an
s
ien
t
f
au
lt
d
etec
tio
n
,
p
r
ed
ictio
n
o
f
f
a
u
lt
o
cc
u
r
r
en
ce
s
u
s
in
g
L
STM
an
d
ch
ec
k
p
o
in
t
p
l
ac
em
en
t
an
d
f
au
lt
r
ec
o
v
er
y
u
s
in
g
C
I
N
C
I
C
-
FTM
.
Fig
u
r
e
2
.
B
lo
ck
d
iag
r
am
o
f
C
I
N
C
I
C
-
FTM
u
s
in
g
L
STM
Data
co
llectio
n
a
n
d
p
r
o
ce
s
s
in
g
:
in
it
ially
,
th
e
s
en
s
o
r
s
r
ea
d
i
n
g
s
lik
e
tem
p
er
atu
r
e,
h
u
m
id
ity
,
w
ater
lev
el,
lig
h
t
an
d
m
o
is
tu
r
e
in
I
o
T
ag
r
i
cu
ltu
r
e
s
ce
n
ar
io
,
ar
e
co
n
f
ig
u
r
e
d
an
d
co
llected
ev
er
y
m
in
u
te
(
ad
ap
tab
le
b
ased
o
n
ap
p
licatio
n
r
eq
u
ir
e
m
en
t)
.
W
e
s
im
u
late
th
e
p
r
o
p
o
s
ed
m
o
d
el
o
n
Go
o
g
le
C
o
lab
u
s
in
g
Py
t
h
o
n
3
.
7
,
u
tili
zin
g
Py
th
o
n
lib
r
ar
ies
in
clu
d
in
g
Pan
d
as,
OS
,
Pick
le,
So
c
k
et,
Nu
m
p
y
,
C
r
c
m
o
d
e,
T
r
ac
em
allo
c,
Ps
u
til,
M
atp
lo
tlib
,
Ker
as
a
n
d
T
en
s
o
r
Flo
w.
Me
s
s
ag
es
b
etwe
en
p
r
o
ce
s
s
es
wer
e
g
en
e
r
ated
u
s
in
g
ex
p
o
n
en
tial
d
is
tr
ib
u
tio
n
s
p
an
n
in
g
b
etwe
en
4
to
1
0
0
p
r
o
ce
s
s
es.
Per
io
d
ically
s
av
e
th
e
s
tate
o
f
ea
ch
s
en
s
o
r
/e
d
g
e
n
o
d
e
an
d
co
m
m
u
n
icatio
n
s
tate
u
s
in
g
‘
Pick
le’
lib
r
ar
y
to
s
er
ialize
an
d
s
to
r
e
th
e
s
tate.
Me
s
s
ag
es e
x
ch
an
g
ed
u
s
in
g
‘
So
ck
et’
lib
r
a
r
y
b
etwe
en
n
o
d
es we
r
e
lo
g
g
ed
in
lo
ca
l a
n
d
ce
n
tr
al
d
atab
ase
t
o
en
ab
le
r
ec
o
v
er
y
.
A
co
o
r
d
in
a
to
r
n
o
d
e
is
d
esig
n
ated
to
m
an
a
g
e
ch
ec
k
p
o
in
tin
g
.
T
r
an
s
ien
t
f
au
lt
d
etec
tio
n
:
f
a
u
lt
d
etec
tio
n
alg
o
r
ith
m
s
(
r
e
f
er
Alg
o
r
ith
m
1
to
4
)
ar
e
im
p
le
m
en
ted
to
id
en
tify
tr
a
n
s
ien
t
f
au
lts
.
T
r
an
s
ien
t
f
au
lts
ar
e
s
o
f
twar
e
f
a
u
lts
an
d
tem
p
o
r
ar
y
th
at
o
cc
u
r
d
u
r
in
g
r
u
n
tim
e
o
f
t
h
e
s
y
s
tem
.
T
r
an
s
ien
t
f
au
lts
ap
p
ea
r
to
b
e
u
n
p
r
ed
ictab
le
f
o
r
a
v
er
y
s
h
o
r
t
p
er
io
d
o
f
tim
e
wh
o
s
e
im
p
ac
t
is
an
o
m
alo
u
s
ly
lar
g
e
o
n
I
o
T
s
y
s
tem
wh
ile
d
ec
is
i
on
-
m
ak
in
g
an
d
ac
tu
atin
g
.
A
ls
o
,
f
r
eq
u
en
c
y
o
f
th
ese
tr
an
s
ien
t
f
au
lt
o
cc
u
r
r
en
ce
s
is
ex
p
o
n
en
tially
p
r
o
p
o
r
tio
n
al
t
o
ab
n
o
r
m
ality
i
n
s
y
s
tem
ac
tio
n
s
[
4
]
,
[
8
]
-
[
10
]
,
[
12
]
,
[
36
]
.
I
n
I
o
T
ap
p
licatio
n
s
th
e
m
o
s
t
f
r
eq
u
en
tly
ap
p
ea
r
in
g
tr
a
n
s
ien
t
f
au
lts
ar
e
i
)
if
th
er
e
is
in
co
r
r
ec
t/m
is
s
in
g
s
eq
u
e
n
ce
n
u
m
b
er
in
p
ac
k
et
o
f
in
f
o
r
m
atio
n
f
lo
w,
th
e
s
y
s
tem
r
esp
o
n
d
s
d
if
f
e
r
en
tly
,
an
d
th
en
th
e
s
eq
u
en
ce
n
u
m
b
er
f
a
u
lt
–
F1
is
d
etec
ted
.
T
h
e
in
f
o
r
m
atio
n
is
o
f
two
ty
p
es
–
p
ay
lo
ad
a
n
d
co
m
m
an
d
s
eq
u
e
n
ce
n
u
m
b
er
.
ii)
I
f
in
teg
r
ity
o
f
th
e
in
f
o
r
m
ati
o
n
is
af
f
ec
ted
,
ch
ec
k
s
u
m
f
au
lt
–
F2
is
d
etec
ted
an
d
ch
ec
k
s
u
m
b
as
ed
f
au
lt
d
etec
tio
n
tech
n
i
q
u
e
is
u
s
ed
.
iii)
I
n
teg
r
ity
o
f
th
e
in
f
o
r
m
atio
n
is
also
v
er
if
ied
b
y
ch
ec
k
in
g
th
e
len
g
th
o
f
t
h
e
m
ess
ag
e.
I
f
th
e
len
g
th
o
f
th
e
m
ess
ag
e
r
ec
eiv
ed
o
r
tr
an
s
m
itte
d
is
n
o
t
as
ex
p
ec
t
ed
o
r
r
ec
eiv
e
d
d
ata
is
ir
r
elev
a
n
t
with
its
p
r
ev
io
u
s
ly
g
en
er
ate
d
p
atter
n
,
th
e
n
s
u
ch
f
au
lt
is
r
ep
r
esen
ted
as
n
u
ll
ch
ar
ac
ter
f
au
lt
–
F3
.
(
iv
)
W
h
en
th
e
in
f
o
r
m
atio
n
r
ec
eiv
ed
o
r
t
r
a
n
s
m
itted
is
b
ey
o
n
d
th
e
p
ay
lo
a
d
r
an
g
e,
th
e
o
u
t
o
f
r
an
g
e
f
a
u
lt
–
F4
is
d
etec
ted
.
T
h
e
tr
an
s
ien
t
f
a
u
lts
d
etec
ted
u
s
in
g
th
e
f
au
lt
d
etec
tio
n
alg
o
r
ith
m
s
af
f
ec
t
th
e
f
au
lt
o
cc
u
r
r
en
ce
s
ce
n
ar
io
s
b
y
d
esig
n
in
g
o
cc
u
r
r
en
ce
s
em
an
tics
in
m
y
wo
r
k
.
T
h
e
d
etec
ted
f
au
lts
,
th
eir
o
cc
u
r
r
en
ce
a
n
d
f
a
u
lt st
ates a
r
e
lo
g
g
ed
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
C
o
mmu
n
ica
tio
n
in
d
u
ce
d
c
h
ec
kp
o
in
tin
g
b
a
s
ed
fa
u
lt to
lera
n
c
e
mec
h
a
n
is
m
u
s
in
g
…
(
S
o
w
ja
n
ya
La
ksh
mi
A
)
1789
Alg
o
r
ith
m
1
.
T
o
d
etec
t
f
au
lt F1
Initialize: expected_seq_num = 0
tolerance_threshold = 5 // Adjust as needed
Function
detect_seq_num_fault(received_seq_num):
if received_seq_num = = expected_seq_num:
// Received sequence number is as
expected
expected_seq_num += 1
return NO_FAULT
else if received_seq_num < expected_seq_num:
// Received a duplicate sequence number
return "Duplicate Sequence Number Detected"
else if received_seq_num
-
expected_seq_num > tolerance_threshold:
// Sequence number gap exceeds the
tole
rance threshold
return "Sequence Number Gap Detected"
else:
// Sequence number gap is within the
tolerance threshold
expected_seq_num =
received_seq_num + 1
return NO_FAULT
Alg
o
r
ith
m
2
.
T
o
d
etec
t f
au
lt F2
Alg
o
r
ith
m
3
.
T
o
d
etec
t f
au
lt F3
Function calculate_chksum(data):
//Calculate a checksum value for the given
data
chksum = 0
for each byte in data:
chksum += byte
// Add the value of each byte to the
checksum
return checksum
Function
detect_chksum_fault(received_data,
received_chksum):
expected_chksum =
calculate_chksum(received_data)
if received_chksum == expected_chksum:
//Checksum matches, data is intact
return NO_FAULT
else:
//Checksum mismatch, data is corrupted
return "Checksum Mismatch Detected"
fault =
detect_chksum_fault(received_data,
received_chksum)
if fault != NO_FAULT: print(fault)
while true:
// Wait for incoming data
trg_data = receiveData()
src_data = sentData()
// Check for null character
if hasNullCharacter(src_data,trg_data):
logError("Null character detected in the
transmitted data")
else:
// Null character not detected, pro
cess the
data
processData(data)
// Function to check for null character
function
hasNullCharacter(src_data,trg_data):
if length(src_data) !=
length(trg_data):
return true
for character in trg_data:
if character
== '
\
0'
return true
return false
Alg
o
r
ith
m
4
.
T
o
d
etec
t f
au
lt F4
//For different types of sensors
it is
essential to configure range values. Based
on configuration, error is detected
// Initialize variables
expectedMinValuea = 0, expectedMaxValue =
100
receivedMinValue = 0, receivedMaxValue =
0
// Function to check if the received data is
within the
expected range
function isDataInRange(data):
if data >= expectedMinValue and data
<= expectedMaxValue:
return true
else:
return false
// Function to calculate the checksum
function calculateChecksum(data):
checksum = 0
for value in data:
checksum += value
return checksum % 256
// Main program loop
while true:
// Wait for incoming message
message = receiveMessage()
// Extract the data, received minimum value,
received maximum value, a
nd checksum from
the message
data = extractData(message)
receivedMinValue =
extractMinValue(message)
receivedMaxValue =
extractMaxValue(message)
receivedChecksum =
extractChecksum(message)
// Verify the checksum
calculatedChecksum =
ca
lculateChecksum(data)
if calculatedChecksum ==
receivedChecksum:
// Checksum is valid, now check for data
range
if receivedMinValue >=
expectedMinValue and receivedMaxValue <=
expectedMaxValue:
// Data range is within the expected range
if isDataInRange(data):
// Data is within the expected range,
process the data
processData(data)
else:
// Data is out of range
logError("Received data
is out of range")
else:
// Received range is out of the expected
range
logError("Received data range
is out of the expected range")
else:
// Checksum is invalid
logError("Invalid checksum
detected")
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
37
,
No
.
3
,
Ma
r
ch
20
25
:
1
785
-
1
7
9
6
1790
Pre
d
ictio
n
o
f
f
a
u
lt
o
cc
u
r
r
en
ce
:
th
e
L
STM
n
etwo
r
k
tak
es
co
lle
cted
s
eq
u
en
tial
in
p
u
t
d
ata
at
v
a
r
io
u
s
tim
e
s
tam
p
s
an
d
h
is
to
r
ical
d
ata
o
f
f
a
u
lt
o
cc
u
r
r
en
ce
s
f
o
r
tr
ain
in
g
p
u
r
p
o
s
es
an
d
tr
ai
n
ed
with
m
u
ltip
le
ep
o
ch
s
to
p
r
ed
ict
th
e
n
ex
t
f
a
u
lty
n
o
d
e.
T
h
is
p
r
e
d
icted
o
u
tp
u
t
is
lo
g
g
ed
as
a
h
i
s
to
r
ical
d
atas
et
o
n
ev
er
y
n
ew
p
r
ed
ictio
n
.
Data
s
et
also
in
clu
d
es
tim
estam
p
s
a
n
d
lab
els
in
d
icatin
g
w
h
eth
er
a
f
au
lt
o
cc
u
r
r
ed
.
Sp
lit
th
e
d
ataset
in
to
tr
ain
in
g
,
v
alid
atio
n
an
d
test
s
ets.
T
h
e
tr
ain
in
g
s
et
is
u
s
ed
to
tr
ain
th
e
L
STM
m
o
d
el,
th
e
v
alid
atio
n
s
et
is
u
s
ed
to
tu
n
e
hyp
er
p
ar
am
eter
s
an
d
t
h
e
test
s
et
is
u
s
ed
to
ev
alu
ate
th
e
m
o
d
el.
T
h
e
L
STM
m
o
d
el
h
as
an
i
n
p
u
t
lay
er
,
lay
er
s
o
f
L
STM
an
d
a
d
e
n
s
e
o
u
tp
u
t
lay
e
r
.
T
r
ai
n
th
e
L
STM
m
o
d
el
o
n
th
e
tr
ain
in
g
d
ata
an
d
v
alid
ate
it
u
s
in
g
th
e
v
alid
atio
n
d
ata.
T
h
e
m
o
d
el
is
ev
alu
ated
o
n
th
e
test
d
ata
u
s
in
g
r
elev
an
t
m
etr
ics
as
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all
an
d
F1
-
s
co
r
e.
L
STM
,
an
ad
v
a
n
ce
d
r
ec
u
r
r
en
t
n
eu
r
al
n
etwo
r
k
(
R
NN)
v
er
s
io
n
,
r
etain
s
v
ital
d
ata,
d
is
ca
r
d
in
g
l
ess
u
s
ef
u
l.
L
STM
in
clu
d
es
in
p
u
t,
f
o
r
g
et
an
d
o
u
tp
u
t
g
ates
[
3
4
]
.
T
h
e
f
o
r
g
et
g
ate,
r
ep
r
esen
ted
b
y
“f
t
”
,
d
is
ca
r
d
s
ir
r
elev
an
t
d
ata.
I
t’
s
d
eter
m
in
e
d
b
y
a
s
ig
m
o
id
f
u
n
ctio
n
b
ased
o
n
weig
h
ts
an
d
b
ias.
I
n
p
u
t “
X
t
”
f
ee
d
s
in
to
th
e
n
etwo
r
k
,
with
th
e
f
o
r
g
et
g
ate
d
ec
id
in
g
if
“h
t−
1
”
o
u
tp
u
t
is
r
el
ev
an
t
to
cu
r
r
en
t
d
ata.
Activ
atio
n
f
u
n
ctio
n
“(
h
t
-
1
+X
t
)
”
d
eter
m
in
es
r
elev
an
ce
,
clo
s
er
to
ze
r
o
im
p
lies
ir
r
el
ev
an
ce
,
o
th
e
r
wis
e
r
elev
an
ce
.
Fig
u
r
e
3
s
h
o
ws
th
e
L
STM
u
n
it with
its
co
m
p
o
n
e
n
ts
.
Fig
u
r
e
3
.
B
asic L
STM
c
ell
T
h
e
“
f
t
”
is
a
f
o
r
g
et
g
ate
f
u
n
ctio
n
,
th
at
f
o
r
g
ets,
o
r
ig
n
o
r
es
th
e
ir
r
elev
an
t
in
f
o
r
m
atio
n
.
“
Wt
f
”
,
ar
e
th
e
weig
h
ts
an
d
“
bs
f
”
th
e
b
ias o
f
t
h
e
f
o
r
g
et
g
ate,
σ
is
a
s
ig
m
o
id
cu
r
v
e,
an
d
".
"
im
p
lies
m
u
ltip
li
ca
tio
n
o
f
th
e
m
atr
ix
.
T
h
e
n
e
x
t
s
tep
is
t
o
d
ete
r
m
in
e
,
u
s
in
g
two
p
r
o
ce
d
u
r
es,
wh
ic
h
n
ew
k
n
o
wled
g
e
is
p
r
esen
t
in
th
e
ce
ll
s
tate.
T
h
e
in
p
u
t
g
ate
f
ir
s
t
d
eter
m
in
es
wh
i
ch
s
tates
ar
e
u
p
d
ate
d
,
an
d
th
en
th
e
“
ta
n
h
”
ac
tiv
atio
n
f
u
n
ctio
n
is
u
s
ed
to
g
e
n
er
ate
a
v
ec
to
r
o
f
p
o
te
n
tial n
ew
v
alu
es.
T
h
e
cu
r
r
en
t c
ell
s
tate
“
C
t
”
is
th
en
ch
an
g
ed
u
s
in
g
o
u
tp
u
ts
f
r
o
m
th
e
in
p
u
t g
ate,
th
e
f
o
r
g
et
g
ate
,
an
d
th
e
“
ta
n
h
”
lay
er
.
Fin
ally
,
th
e
o
u
tp
u
t
g
a
te
an
d
a
“
ta
n
h
”
f
u
n
ctio
n
ar
e
u
s
ed
to
ca
lcu
late
th
e
o
u
tp
u
t
o
f
n
etwo
r
k
“h
t
”
,
W
h
er
e
“
Wt
o
”,
is
th
e
in
p
u
t w
eig
h
t a
n
d
“
bs
o
”
th
e
b
ias o
f
th
e
o
u
tp
u
t
g
ate,
r
esp
ec
tiv
ely
.
=
(
.
[
ℎ
−
1
,
]
+
(
1
)
=
(
.
[
ℎ
−
1
,
]
+
(
2
)
~
=
ℎ
(
.
[
ℎ
−
1
,
]
+
(
3
)
=
∗
−
1
+
∗
~
(
4
)
=
(
.
[
ℎ
−
1
,
]
+
)
(
5
)
ℎ
=
∗
ℎ
(
)
(
6
)
+
=
(
,
+
1
,
+
2
…
.
+
−
1
)
(
7
)
Dr
o
p
o
u
t
r
eg
u
la
r
izatio
n
co
m
b
at
s
o
v
er
-
f
itti
n
g
b
y
r
em
o
v
in
g
ce
r
t
ain
n
eu
r
o
n
s
.
E
ac
h
h
id
d
en
u
n
it
i
n
a
n
e
u
r
al
n
etwo
r
k
is
t
r
ain
ed
u
s
in
g
a
r
a
n
d
o
m
s
am
p
le
o
f
o
t
h
er
s
.
L
STM
n
etwo
r
k
s
p
r
ed
ict
‘
n
’
tim
e
s
tep
s
ah
ea
d
,
co
n
tr
asti
n
g
p
r
o
jecte
d
with
ac
tu
al
v
al
u
es.
I
n
p
u
t
d
ata“
X
t
”at
tim
e
s
tam
p
“t
”is
p
r
o
ce
s
s
ed
s
eq
u
en
tially
with
a
s
lid
in
g
win
d
o
w
m
eth
o
d
.
L
STM
p
r
e
d
icts
th
e
n
e
x
t
f
au
lty
n
o
d
e
with
in
th
e
e
(
s
iz
e
-
15)
.
T
h
e
f
ir
s
t
s
am
p
le
o
f
in
p
u
t
is
x
1
,
s
ec
o
n
d
is
x
2
an
d
s
o
X
t
={x
t
-
1
,
x
t
,
x
t+
1
}.
T
h
is
i
n
p
u
t
tak
e
n
at
“W
L
”
win
d
o
w
le
n
g
th
is
{x
t−
w
,
x
t+
1
-
w
...x
t−
1
}
u
s
ed
f
o
r
p
r
ed
ictio
n
.
T
h
e
L
STM
n
etwo
r
k
is
tr
ain
ed
with
“n
”
ep
o
c
h
s
an
d
“
Y
t
”
is
o
u
tp
u
t
–
p
r
ed
icted
f
a
u
lty
n
o
d
e
a
n
d
“
O
n
”
is
L
STM
n
etwo
r
k
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
C
o
mmu
n
ica
tio
n
in
d
u
ce
d
c
h
ec
kp
o
in
tin
g
b
a
s
ed
fa
u
lt to
lera
n
c
e
mec
h
a
n
is
m
u
s
in
g
…
(
S
o
w
ja
n
ya
La
ksh
mi
A
)
1791
o
u
tp
u
t.
L
STM
’
s
ad
v
an
ta
g
e
lies
in
p
r
ed
ictin
g
f
u
tu
r
e
v
alu
es
wh
ile
r
em
em
b
er
in
g
p
ast
o
n
e
s
,
u
n
til
th
e
r
eq
u
ir
e
d
m
u
lt
is
tep
p
r
ed
ictio
n
s
ar
e
m
a
d
e.
C
h
ec
k
p
o
in
t
p
lace
m
en
t
a
n
d
f
a
u
lt
r
ec
o
v
er
y
u
s
in
g
C
I
N
C
I
C
-
FT
M:
t
he
co
r
e
o
f
th
e
C
I
N
C
I
C
-
FTM
is
th
e
p
r
ed
ictio
n
o
f
p
o
ten
tial f
au
lt o
c
cu
r
r
en
ce
s
u
s
in
g
L
STM
n
etwo
r
k
.
T
h
e
p
r
e
d
icted
f
au
lts
o
cc
u
r
r
e
n
ce
s
ar
e
lo
g
g
ed
as
f
au
lt
h
is
to
r
y
.
I
n
itially
,
lo
ca
l
c
h
ec
k
p
o
in
ts
ar
e
p
lace
d
at
n
o
d
es
af
ter
s
ig
n
if
ican
t
s
y
s
tem
ev
e
n
ts
.
L
ater
af
ter
tr
ain
in
g
L
STM
m
o
d
el,
f
o
r
ce
d
ch
ec
k
p
o
in
ts
ar
e
p
lace
d
at
latest
co
n
s
is
ten
t
n
o
d
e
p
r
io
r
to
p
r
ed
icted
f
a
u
lty
n
o
d
e
.
T
h
is
is
d
o
n
e
d
y
n
am
ically
,
with
th
e
m
o
d
el
co
n
tin
u
o
u
s
ly
u
p
d
atin
g
its
p
r
ed
ictio
n
s
as
n
ew
d
ata
is
p
r
o
ce
s
s
ed
.
T
h
e
ch
ec
k
p
o
i
n
tin
g
p
r
o
ce
s
s
in
v
o
lv
es
s
av
in
g
th
e
cu
r
r
en
t
s
tate
o
f
th
e
s
y
s
tem
,
in
clu
d
in
g
th
e
p
r
o
ce
s
s
d
ata
an
d
cu
r
r
en
t
s
tatu
s
o
f
th
e
m
ess
ag
es
b
ein
g
p
r
o
ce
s
s
ed
.
T
h
e
s
y
s
tem
u
p
o
n
f
au
lt
d
etec
tio
n
ch
ec
k
s
wh
eth
er
th
e
cu
r
r
en
t
n
o
d
e
o
r
its
p
r
ec
ed
in
g
n
o
d
e
h
as
v
alid
ch
ec
k
p
o
in
ts
.
T
h
e
s
y
s
tem
in
itiates
a
r
o
llb
ac
k
to
th
e
m
o
s
t
r
ec
en
t
ch
ec
k
p
o
in
t,
allo
win
g
th
e
p
r
o
ce
s
s
to
r
estar
t
f
r
o
m
th
e
k
n
o
w
n
g
o
o
d
(
f
a
u
lt
f
r
ee
)
s
tate
an
d
p
r
e
v
en
tin
g
th
e
s
p
r
ea
d
o
f
f
au
lts
th
r
o
u
g
h
th
e
s
y
s
tem
.
Af
ter
s
u
cc
ess
f
u
l
r
ec
o
v
er
y
an
d
th
e
s
tab
ilizatio
n
o
f
t
h
e
p
r
o
ce
s
s
,
th
e
f
o
r
ce
d
c
h
ec
k
p
o
in
ts
th
at
ar
e
alr
ea
d
y
lo
g
g
ed
as
f
au
lt
h
is
to
r
y
ar
e
e
r
a
s
ed
f
r
o
m
th
e
lo
ca
l
m
em
o
r
y
an
d
th
e
s
y
s
tem
co
n
tin
u
es
to
o
p
er
ate
with
th
e
L
STM
m
o
d
el
ad
ju
s
tin
g
its
p
r
e
d
ictio
n
s
b
ased
o
n
th
e
f
au
lt
h
is
to
r
y
.
T
h
is
ap
p
r
o
ac
h
s
p
ec
if
ically
r
ed
u
ce
th
e
s
to
r
a
g
e
o
v
er
h
ea
d
s
b
ec
a
u
s
e
o
f
ch
ec
k
p
o
in
ts
.
T
h
e
alg
o
r
ith
m
is
g
iv
en
in
Alg
o
r
ith
m
5
an
d
w
o
r
k
in
g
p
r
i
n
cip
le
is
ex
p
lain
ed
as sh
o
wn
in
F
ig
u
r
e
4
.
Alg
o
r
ith
m
5
.
C
I
N
C
I
C
-
FTM
Import libraries: numpy, tensorflow
,
sequential, LSTM, dense, pickle, os
# Load historical data for fault
prediction
def load_dta(file_path):
load historical data related to
transient faults
data = np.load(file_path)
return data
# Build and train LSTM model
def build_lstm_model(input_
set): model =
Sequential()
# predict fault occurrences
def predict_faults(model, x_input):
predictions = model.predict(x_input)
return predictions
# Place checkpoints based on predictions
def place_checkpoints(nodes,predictions,
thresho
ld=0.65):
checkpoints = [ ]
for i, predictions in
enumerate(predictions):
if predictions > threshold:
checkpoints.append(nodes[i])
return checkpoints
# Save system state as checkpoint
def save_checkpoint(node, state,
checkpoint_dir):
checkpoint_path=os.path.join(checkpoint_di
r, f’checkpoint_node_{node}.pk1’)
with open(checkpoint_path, ‘wb’) as
file:
pickle.dump(state, file)
# Fault detection
def
detect_faults(nodes,predictions,current_st
ate):
faulty_nodes = [ ]
for i, node in enumerate(nodes):
if predictions[i] > 0.65 and
current_state[node]== ‘faulty’:
faulty_node.append(node)
return faulty_nodes
# Fault recovery using checkpoints
def recover_from_fault(node,
checkpoint_dir):
checkpoint_path=os.pa
th.join(checkpoint_di
r, f’checkpoint_node_{node}.pk1’)
with open(checkpoint_path, ‘rb’)as file:
state = pickle.load(file)
return state
model.add(LSTM(50,activation=’relu’,
input_set=input_set))
model.add(Dense(1))
model.compile(optimizer = ‘adam’,
loss =
‘mse’)
return model
def train_lstm_model(model,x_train,y_train,
epochs=1000):
model.fit(x_train, y_train, epochs =
epochs, erbose =1)
return model
# Main process to run CIN CIC
-
FTM
def
cin_cic_ftm_process(
nodes,fault_data_file,
checkpoint_dir, current_state):
data = load_data(fault_data_file)
x_train, y_train = data[‘x_train],
data[‘y_train’]
input_set = (x_train.set[1],
x_train.set[2])
# Build and train the LSTM model
model = build_lstm_model(i
nput_set)
model = train_lstm_model(model,
x_train, y_train)
# Predict faults
predictions = predict_faults(model,
x_train)
# Place checkpoints based on predictions
checkpoints= place_checkpoints(node,
predictions)
for node in checkpoints:
state = current_state[node]
save_checkpoint(node, state,
checkpoint_dir)
# Detect faults
faulty_nodes=detect_faults(nodes,prediction
s, current_state)
# Recover from faults
for node in faulty_nodes:
recovered_State=recover_from_fault(node,
che
ckpoint_dir)
current_state[node] = recovered_state
return current_state
# Example usage
nodes = [‘node1’, ‘node2’, ‘node3’,
‘node4’]
fault_data_file = ‘fault_data.npz’
checkpoint_dir = ‘checkpoints’
current_state
= {‘node1’: ‘normal’,
‘node2’: ‘normal’, ‘node3’: ‘normal’,
‘node4’: ‘normal’}
final_state=cin_cic_ftm_process(nodes,
fault_data_file, checkpoint_dir,
current_state)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
37
,
No
.
3
,
Ma
r
ch
20
25
:
1
785
-
1
7
9
6
1792
T
h
is
alg
o
r
ith
m
p
r
o
v
id
es
a
d
etailed
f
r
am
ewo
r
k
f
o
r
im
p
le
m
en
tin
g
th
e
C
I
N
C
I
C
-
FTM
in
Py
th
o
n
,
en
ab
lin
g
ef
f
icien
t
f
au
lt
p
r
ed
ict
io
n
,
s
tr
ateg
ic
c
h
ec
k
p
o
in
t
p
lace
m
en
t
an
d
f
a
u
lt
r
ec
o
v
er
y
in
I
o
T
ap
p
licatio
n
s
.
T
h
e
‘
lo
ad
_
d
ata
’
f
u
n
ctio
n
lo
ad
s
h
is
to
r
ical
f
au
lt
d
ata
f
o
r
tr
ain
in
g
th
e
L
STM
m
o
d
el.
T
h
e
d
ata
is
ex
p
ec
ted
to
b
e
in
a
‘
.
n
p
z’
f
ile
co
n
tain
in
g
tr
ain
in
g
an
d
test
in
g
d
atasets
.
T
h
e
L
ST
M
m
o
d
el
is
b
u
ilt
u
s
in
g
T
en
s
o
r
Flo
w’
s
Ker
as
API
,
with
an
in
p
u
t
la
y
er
,
L
STM
lay
er
an
d
o
u
tp
u
t d
en
s
e
lay
er
.
T
h
e
m
o
d
el
is
tr
ain
ed
o
n
th
e
h
is
to
r
ical
f
au
lt
d
ata.
T
h
e
ch
ec
k
p
o
i
n
ts
ar
e
p
lace
d
o
n
n
o
d
es
wh
er
e
th
e
p
r
ed
icted
f
au
lt
p
r
o
b
ab
ilit
y
ex
ce
ed
s
a
ce
r
tain
th
r
esh
o
ld
(
0
.
6
5
)
.
T
h
e
s
tate
o
f
th
ese
n
o
d
es is
s
av
ed
u
s
in
g
th
e
‘
p
ick
le’
lib
r
ar
y
.
T
h
e
m
ain
f
u
n
ctio
n
‘
ci
n
_
cic_
f
tm
_
p
r
o
c
ess
’
in
teg
r
ates a
ll
s
tep
s
an
d
p
r
o
ce
s
s
es th
e
n
o
d
es
in
th
e
I
o
T
n
etwo
r
k
to
p
lace
c
h
ec
k
p
o
in
ts
an
d
p
er
f
o
r
m
f
au
lt r
e
co
v
er
y
.
C
o
n
s
id
er
Fig
u
r
e
4
as
an
e
x
am
p
le
f
o
r
u
n
d
er
s
tan
d
in
g
C
I
N
C
I
C
-
FTM
.
Firstl
y
,
C
I
N
b
ased
C
I
C
in
d
u
ce
s
ch
ec
k
p
o
i
n
ts
C
1
an
d
C
2
at
n
o
d
e
N
2,
1
an
d
N
3,
2
as
th
ese
n
o
d
es
ar
e
p
r
ed
icted
t
o
h
av
e
f
au
lt
s
b
ased
o
n
L
STM
alg
o
r
ith
m
.
T
h
e
s
eq
u
en
ce
o
f
m
ess
ag
e
tr
an
s
m
is
s
io
n
is
-
m
ess
ag
e
M1
is
s
en
t
f
r
o
m
p
r
o
ce
s
s
P4
to
P3
,
m
ess
ag
e
M2
f
r
o
m
P1
to
P3
,
a
n
d
s
o
o
n
.
I
f
a
t
M5
,
a
f
au
lt
is
d
etec
ted
,
C
I
C
-
FTM
in
itiate
s
th
e
r
o
llb
ac
k
p
r
o
p
a
g
atio
n
[
3
7
]
a
n
d
will
r
o
llb
ac
k
to
th
e
latest
p
lac
ed
ch
ec
k
p
o
in
t,
C
1
an
d
r
estar
ts
f
r
o
m
C
1
.
Ag
ain
,
if
f
au
lt
is
n
o
t
r
ec
o
v
er
ed
at
M5
,
th
e
s
y
s
tem
will
r
o
llb
ac
k
to
n
ex
t
latest
ch
ec
k
p
o
in
t
C
2
an
d
r
est
ar
ts
f
r
o
m
th
er
e.
Sti
ll
if
f
au
lt
is
n
o
t
r
ec
o
v
er
e
d
,
th
e
n
FTM
r
etu
r
n
s
to
in
itial
s
tates
an
d
r
estar
ts
th
e
p
r
o
ce
s
s
f
r
o
m
in
itial
s
tates.
S
till
if
f
au
lt
is
n
o
t
r
ec
o
v
er
ed
,
th
e
FTM
co
n
clu
d
es its
h
ar
d
war
e
f
au
lt.
Fig
u
r
e
4
.
E
x
am
p
le
f
o
r
c
h
ec
k
p
o
in
t a
t in
ter
m
ed
iate
n
o
d
e
b
ased
C
I
C
-
FTM
3.
RE
SU
L
T
AND
DI
SCUS
S
I
O
N
I
n
th
is
s
ec
tio
n
,
we
s
im
u
late
t
h
e
p
r
o
p
o
s
ed
m
o
d
el
o
n
th
e
G
o
o
g
le
C
o
lab
p
latf
o
r
m
u
s
in
g
P
y
th
o
n
3
.
7
.
Dis
cr
ete
ev
en
t
s
im
u
latio
n
s
p
a
n
s
4
-
1
0
0
p
r
o
ce
s
s
es,
g
en
er
atin
g
m
ess
ag
es
v
ia
ex
p
o
n
en
tial
d
is
tr
ib
u
tio
n
.
Sen
s
o
r
f
r
eq
u
e
n
cy
,
s
et
to
o
n
e
m
in
u
te,
a
d
ap
tab
le
b
ased
o
n
ap
p
lic
atio
n
r
eq
u
ir
em
e
n
ts
.
T
h
e
s
en
s
o
r
s
u
s
ed
in
I
o
T
ag
r
icu
ltu
r
e
in
clu
d
e
tem
p
er
at
u
r
e,
h
u
m
id
ity
,
wate
r
lev
el,
lig
h
t,
an
d
m
o
is
tu
r
e
s
en
s
o
r
s
.
T
r
ain
in
g
in
v
o
lv
es
7
0
0
0
+
d
ataset,
1
0
0
0
+
f
o
r
test
in
g
u
n
d
er
d
iv
e
r
s
e
s
ce
n
ar
io
s
.
E
ac
h
s
ce
n
ar
io
u
n
d
er
g
o
e
s
1
-
1
5
test
s
,
av
er
ag
in
g
p
ar
am
e
ter
s
f
o
r
ass
es
s
m
en
t.
R
ec
o
n
f
ig
u
r
in
g
th
e
s
en
s
o
r
f
r
eq
u
en
cy
to
s
u
it th
e
n
ee
d
s
o
f
th
e
ap
p
licatio
n
is
p
o
s
s
ib
le.
T
h
e
p
e
r
f
o
r
m
an
ce
p
ar
am
eter
s
f
o
r
p
r
ed
ictio
n
o
f
f
au
lt
n
o
d
es u
s
in
g
L
STM
m
o
d
el
a
r
e
i)
p
r
ec
is
i
o
n
,
“P”,
is
th
e
p
er
ce
n
ta
g
e
o
f
v
alu
es
i
d
en
t
if
ied
p
r
ec
is
ely
;
ii
)
r
ec
all,
“R”,
wh
ich
is
th
e
n
u
m
b
e
r
o
f
p
o
s
itiv
e
class
p
r
ed
ictio
n
o
u
t
o
f
all
p
o
s
itiv
e
ex
am
p
les
;
an
d
iii)
F1
-
s
co
r
e,
“F1
”,
wh
ich
is
th
e
h
ar
m
o
n
ic
m
ea
n
o
f
p
r
ec
is
io
n
an
d
r
ec
all,
p
r
o
v
id
es
a
m
o
r
e
ac
cu
r
ate
p
ic
tu
r
e
o
f
ca
s
es
th
at
wer
e
i
n
co
r
r
ec
tly
class
if
ied
th
an
th
e
a
cc
u
r
ac
y
m
etr
ic.
T
h
e
ac
cu
r
ac
y
is
d
ef
in
ed
as
in
ex
p
r
ess
io
n
(
1
1
)
,
wh
er
e
tr
u
e
n
eg
ativ
e
=“
TN
”,
tr
u
e
p
o
s
iti
v
e
=“
TP
”
an
d
f
alse
p
o
s
itiv
e
=“
F
P
”,
f
alse
n
eg
ativ
e
=“
F
N
”.
Am
o
n
g
th
e
v
ar
io
u
s
lo
s
s
f
u
n
ctio
n
s
,
r
o
o
t
m
ea
n
s
q
u
ar
ed
er
r
o
r
(
R
MSE
)
is
th
e
ap
p
r
o
p
r
iate
m
e
tr
ic
to
ev
alu
ate
tim
e
s
er
ies
[
3
8
]
p
r
ed
ictio
n
m
o
d
els
an
d
it
is
d
ef
in
ed
as
in
ex
p
r
ess
io
n
(
1
2
)
,
wh
er
e,
“n
”
is
n
u
m
b
er
o
f
n
o
d
es
,
“i”
is
n
u
m
b
er
o
f
iter
atio
n
s
,
“y
i
”
is
r
ea
l
o
u
tp
u
t,
“
y
p
i
”
is
d
esire
d
o
u
tp
u
t.
T
h
e
lar
g
e
r
R
MSE
in
d
icate
s
les
s
ac
cu
r
ac
y
an
d
least R
MSE
in
d
icate
s
h
ig
h
e
s
t a
cc
u
r
ac
y
[
3
9
]
:
=
+
(
8
)
=
+
(
9
)
1
=
2
∗
(
∗
)
+
(
1
0
)
(
,
)
=
(
(
+
+
+
+
),
(
+
+
+
+
)
)
(
1
1
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
C
o
mmu
n
ica
tio
n
in
d
u
ce
d
c
h
ec
kp
o
in
tin
g
b
a
s
ed
fa
u
lt to
lera
n
c
e
mec
h
a
n
is
m
u
s
in
g
…
(
S
o
w
ja
n
ya
La
ksh
mi
A
)
1793
=
√
1
∑
(
−
)
2
=
1
(
1
2
)
Hy
p
er
p
ar
a
m
eter
s
ar
e
u
s
ed
to
tu
n
e
t
h
e
L
STM
b
ased
p
r
ed
ictio
n
m
o
d
els.
T
h
e
m
o
s
t
im
p
o
r
tan
t
h
y
p
er
p
ar
am
eter
c
o
m
b
in
atio
n
t
h
at
h
av
e
d
ir
ec
t
im
p
ac
t
o
n
m
o
d
el
p
er
f
o
r
m
an
ce
ar
e
as
m
e
n
tio
n
ed
in
T
ab
le
1
with
th
e
th
r
esh
o
ld
v
alu
e
o
f
0
.
6
5
f
o
r
wh
ich
th
e
p
r
e
d
ictio
n
ac
cu
r
a
cy
o
b
t
ain
ed
is
9
8
.
5
%
as
tab
u
lated
in
T
ab
le
2
an
d
s
h
o
wn
in
Fig
u
r
e
5
.
T
h
e
in
teg
r
atio
n
o
f
th
e
L
STM
-
b
ased
d
e
ep
lear
n
in
g
al
g
o
r
ith
m
s
h
o
wed
h
ig
h
ac
c
u
r
ac
y
in
p
r
ed
ictin
g
f
a
u
lt
o
cc
u
r
r
en
ce
s
as
s
h
o
wn
in
Fig
u
r
e
5
.
T
h
e
L
ST
M
m
o
d
el
was
tr
ain
ed
an
d
v
alid
ated
u
s
in
g
h
is
to
r
ical
f
au
lt
d
ata
an
d
it
s
u
cc
ess
f
u
lly
id
en
tifie
d
tr
an
s
ien
t
f
au
lt
o
cc
u
r
r
en
ce
s
with
an
ac
c
u
r
ac
y
o
f
9
8
.
5
%
ac
cu
r
ac
y
,
as
m
ea
s
u
r
ed
b
y
s
tan
d
ar
d
m
etr
ics in
clu
d
in
g
p
r
ec
is
io
n
,
r
ec
all
a
n
d
F1
-
s
co
r
e
as tab
u
lated
in
T
ab
l
e
2
.
T
ab
le
1
.
Hy
p
er
p
ar
a
m
eter
v
alu
es f
o
r
h
ig
h
est ac
cu
r
ac
y
H
y
p
e
r
p
a
r
a
me
t
e
r
s
V
a
l
u
e
s
W
i
n
d
o
w
-
si
z
e
15
D
r
o
p
o
u
t
-
r
a
t
e
0
.
2
R
e
g
u
l
a
r
i
z
e
r
L2
R
e
g
u
l
a
r
i
z
e
r
-
r
a
t
e
0
.
0
1
4
Ep
o
c
h
s
1
0
0
0
A
c
t
i
v
a
t
i
o
n
-
f
u
n
c
t
i
o
n
R
e
L
u
Le
a
r
n
i
n
g
-
r
a
t
e
0
.
0
1
2
T
ab
le
2
.
Av
e
r
ag
e
p
e
r
f
o
r
m
an
ce
m
etr
ics o
f
L
STM
m
o
d
el
Ep
o
c
h
s
%
A
c
c
u
r
a
c
y
R
e
c
a
l
l
F1
-
s
c
o
r
e
50
9
8
.
4
9
8
.
2
98
1
0
0
9
8
.
5
9
8
.
2
3
9
8
.
1
2
2
0
0
9
8
.
6
9
8
.
4
9
8
.
2
5
0
0
9
8
.
5
9
8
.
2
9
8
.
1
1
0
0
0
9
8
.
6
9
8
.
3
98
A
v
e
r
a
g
e
(
f
r
o
m
5
0
t
o
1
0
0
0
e
p
o
c
h
s)
9
8
.
5
2
9
8
.
2
3
9
8
.
1
An
av
er
a
g
e
r
ec
all
o
f
9
8
.
2
3
%
in
d
icate
d
th
at,
o
u
t
o
f
all
th
e
ac
tu
al
f
au
lt
o
cc
u
r
r
en
ce
s
,
t
h
e
L
S
T
M
m
o
d
e
l
s
u
cc
ess
f
u
lly
p
r
ed
icted
9
8
.
2
3
%
o
f
th
em
an
d
th
is
h
ig
h
r
ec
all
v
alu
e
in
d
icate
s
th
at
th
e
m
o
d
el
is
v
er
y
ef
f
ec
tiv
e
at
d
etec
tin
g
m
o
s
t
o
f
th
e
f
a
u
lts
th
at
o
cc
u
r
,
m
in
im
izin
g
th
e
r
is
k
o
f
u
n
p
r
e
d
icted
f
au
lts
wh
ich
lead
s
to
s
y
s
tem
f
ailu
r
es.
An
F1
-
s
co
r
e
o
f
9
8
.
1
%
in
d
icate
s
th
at
th
e
L
STM
m
o
d
el
h
as
a
well
-
b
alan
ce
d
p
er
f
o
r
m
a
n
ce
,
with
b
o
th
h
ig
h
r
ec
al
l
an
d
h
i
g
h
p
r
ec
is
io
n
.
T
h
is
in
d
ic
ates
th
e
m
o
d
el
is
r
eliab
le
in
p
r
ed
ictin
g
f
au
lts
with
m
i
n
im
al
f
a
ls
e
alar
m
s
,
wh
ich
is
cr
itical
f
o
r
ef
f
icien
t
f
au
lt
m
an
ag
em
en
t
i
n
I
o
T
s
y
s
tem
s
.
No
w,
th
e
p
r
o
p
o
s
ed
C
I
N
C
I
C
-
FTM
p
lace
s
th
e
ch
ec
k
p
o
i
n
ts
at
th
e
id
en
tifie
d
f
a
u
lty
n
o
d
es.
T
h
e
Fig
u
r
e
6
s
h
o
w
s
av
er
ag
e
n
u
m
b
er
o
f
c
h
ec
k
p
o
in
ts
p
lace
d
,
av
er
a
g
e
n
u
m
b
er
o
f
u
s
eless
ch
ec
k
p
o
in
t
s
,
m
em
o
r
y
co
n
s
u
m
p
tio
n
a
n
d
C
PU
co
n
s
u
m
p
tio
n
f
o
r
ch
ec
k
p
o
in
t
in
g
in
th
e
n
etwo
r
k
,
an
d
a
v
er
ag
e
b
ac
k
p
r
o
p
a
g
atio
n
tim
e
an
d
f
au
lt
r
ec
o
v
er
y
tim
e,
wh
en
th
e
n
etwo
r
k
is
o
f
d
if
f
e
r
e
n
t n
o
d
e
s
izes lik
e,
4
n
o
d
es,
1
0
n
o
d
es,
4
0
n
o
d
es a
n
d
1
0
0
n
o
d
es.
Fig
u
r
e
5
.
L
STM
f
a
u
lt p
r
ed
icti
o
n
ac
cu
r
ac
y
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
37
,
No
.
3
,
Ma
r
ch
20
25
:
1
785
-
1
7
9
6
1794
Fig
u
r
e
6
.
C
h
ec
k
p
o
in
ts
co
u
n
t a
n
d
r
eso
u
r
ce
co
n
s
u
m
p
tio
n
b
y
C
I
N
C
I
C
-
FTM
T
h
e
o
b
s
er
v
atio
n
s
ar
e:
−
T
h
e
C
I
N
C
I
C
-
FTM
u
s
in
g
L
STM
p
lace
s
ch
ec
k
p
o
in
ts
at
p
r
e
d
icted
f
au
lty
n
o
d
es
th
at
ar
e
1
5
%
o
f
th
e
to
tal
n
u
m
b
er
o
f
n
o
d
es i
n
th
e
n
etwo
r
k
,
h
e
n
ce
s
ig
n
i
f
ican
tly
r
ed
u
ce
s
th
e
n
u
m
b
er
o
f
f
o
r
ce
d
an
d
u
s
el
ess
ch
ec
k
p
o
in
ts
co
m
p
ar
ed
to
e
x
is
tin
g
m
eth
o
d
s
[
9
]
,
[
17
]
,
[
21
]
,
[
22
]
,
[
39
]
t
h
at
o
f
ten
p
lace
c
h
ec
k
p
o
in
ts
in
d
is
cr
im
in
ately
ac
r
o
s
s
m
u
ltip
le
n
o
d
es.
−
T
h
e
C
I
N
C
I
C
-
FTM
m
ak
es
tar
g
eted
ch
ec
k
p
o
in
t
p
lace
m
en
t
j
u
s
t
b
ef
o
r
e
n
o
d
es
p
r
ed
icted
to
b
e
f
au
lty
.
T
h
ese
ch
ec
k
p
o
i
n
ts
b
ec
o
m
e
u
s
eless
o
n
ly
if
f
au
lt
r
ec
o
v
er
y
is
s
u
cc
ess
f
u
l
b
y
r
estar
tin
g
f
r
o
m
th
e
m
o
s
t
r
ec
en
tly
p
lace
d
ch
ec
k
p
o
i
n
t,
p
r
e
v
en
tin
g
t
h
e
n
ee
d
to
r
o
llb
ac
k
to
s
u
b
s
eq
u
e
n
t
o
n
es.
As
a
r
esu
lt,
th
e
n
u
m
b
er
o
f
u
s
eless
ch
ec
k
p
o
i
n
ts
is
m
in
im
ized
-
less
th
an
1
0
% o
f
th
e
t
o
tal
n
u
m
b
er
o
f
ch
ec
k
p
o
in
ts
.
−
E
x
is
tin
g
ch
ec
k
p
o
in
tin
g
m
et
h
o
d
s
in
v
o
lv
e
f
r
eq
u
e
n
t,
p
er
io
d
ic
an
d
in
d
is
cr
im
in
ate
ch
ec
k
p
o
i
n
t
p
lace
m
en
ts
[
5
]
,
[
9
]
,
[
13
]
,
[
17
]
,
[
21
]
,
[
22
]
,
[
24
]
,
[
25
]
,
[
27
]
,
[
32
]
,
[
39
]
,
lea
d
in
g
to
u
n
n
ec
ess
ar
y
m
em
o
r
y
co
n
s
u
m
p
tio
n
d
u
e
to
th
e
s
to
r
ag
e
o
f
m
u
ltip
le
an
d
u
s
eless
ch
ec
k
p
o
in
ts
.
T
h
e
C
PU
is
lo
ad
ed
with
th
e
co
n
tin
u
o
u
s
m
o
n
ito
r
in
g
an
d
p
r
o
ce
s
s
in
g
o
f
th
ese
ch
ec
k
p
o
in
ts
,
r
ed
u
cin
g
s
y
s
tem
ef
f
icien
cy
[
5
]
,
[
9
]
,
[
13
]
,
[
17
]
,
[
39
]
.
I
n
c
o
n
tr
ast,
th
e
C
I
N
C
I
C
-
FTM
,
s
in
ce
th
e
ch
ec
k
p
o
in
ts
ar
e
p
lace
d
at
tar
g
eted
n
o
d
es
b
ased
o
n
h
ig
h
ac
cu
r
ately
p
r
ed
icted
n
o
d
es,
m
in
im
izes
th
e
m
em
o
r
y
c
o
n
s
u
m
p
tio
n
-
b
y
r
e
d
u
cin
g
th
e
n
u
m
b
er
o
f
ch
ec
k
p
o
in
ts
s
to
r
ed
an
d
th
e
C
PU
co
n
s
u
m
p
tio
n
-
as
th
e
s
y
s
tem
i
s
n
o
t
r
eq
u
ir
ed
to
co
n
s
tan
tly
p
r
o
ce
s
s
an
d
m
an
ag
e
an
ex
ce
s
s
iv
e
n
u
m
b
er
o
f
ch
ec
k
p
o
i
n
ts
.
−
I
o
T
n
etwo
r
k
s
f
ac
e
ch
allen
g
es
with
r
esp
ec
t
to
its
d
y
n
am
ic
b
eh
av
io
r
an
d
r
eso
u
r
ce
c
o
n
s
tr
ain
t
s
[
5
]
,
[
9
]
,
[
13
]
,
[
22
]
,
[
25
]
,
[
27
]
,
[
39
]
.
T
h
e
r
e
d
u
ctio
n
in
m
em
o
r
y
an
d
C
P
U
co
n
s
u
m
p
tio
n
n
o
t
o
n
ly
im
p
r
ess
th
e
o
v
er
all
ef
f
icien
cy
o
f
th
e
s
y
s
tem
b
u
t a
ls
o
en
s
u
r
es th
at
th
e
I
o
T
ap
p
licatio
n
s
ca
n
co
n
tin
u
e
to
o
p
er
ate
e
f
f
ec
tiv
ely
ev
en
in
r
eso
u
r
ce
-
lim
ited
s
ce
n
a
r
io
s
.
−
B
ac
k
p
r
o
p
a
g
atio
n
tim
e
r
e
f
er
s
to
th
e
tim
e
r
eq
u
ir
ed
to
r
o
llb
ac
k
to
t
h
e
r
ec
e
n
tly
p
lace
d
c
h
ec
k
p
o
in
t
f
r
o
m
th
e
f
au
lt
d
etec
ted
n
o
d
e.
Fau
lt
r
ec
o
v
er
y
tim
e
r
e
f
er
s
to
th
e
tim
e
tak
en
to
r
ec
o
v
er
f
r
o
m
a
f
au
lt,
b
y
lo
ad
in
g
th
e
p
r
ev
io
u
s
c
o
n
s
is
ten
t c
h
ec
k
p
o
in
t a
n
d
r
esu
m
i
n
g
e
x
ec
u
tio
n
f
r
o
m
th
at
ch
ec
k
p
o
in
t
o
n
war
d
s
.
I
t in
clu
d
es th
e
tim
e
r
eq
u
ir
ed
to
tr
a
n
s
f
er
d
ata
f
r
o
m
ch
ec
k
p
o
i
n
t sto
r
ag
e
to
th
e
p
r
o
g
r
am
’
s
m
em
o
r
y
,
r
ein
itialize
th
e
ex
ec
u
tio
n
s
tate
an
d
co
n
tin
u
e
its
ex
ec
u
tio
n
.
T
h
e
r
esu
lts
in
d
icate
th
at
th
e
C
I
N
C
I
C
-
FTM
r
ed
u
ce
s
th
e
b
ac
k
p
r
o
p
ag
atio
n
tim
e
an
d
f
a
u
lt
r
ec
o
v
er
y
tim
e
b
y
c
o
n
s
id
er
ab
le
m
a
r
g
i
n
c
o
m
p
a
r
ed
to
ex
is
tin
g
c
h
ec
k
p
o
in
tin
g
te
ch
n
iq
u
es,
wh
ic
h
s
u
f
f
er
f
r
o
m
i
n
ef
f
icien
cies
d
u
e
to
in
co
n
s
is
ten
t
n
o
d
e
s
tates
an
d
th
e
d
o
m
in
o
ef
f
ec
t
[
5
]
,
[
9
]
,
[
13
]
,
[
24
]
,
[
39
]
,
[
40
]
.
B
ec
au
s
e
o
f
p
r
ed
icted
a
n
d
tar
g
ete
d
ch
ec
k
p
o
i
n
tin
g
ap
p
r
o
ac
h
,
t
h
e
o
v
er
all
n
u
m
b
e
r
o
f
c
h
ec
k
p
o
i
n
ts
ar
e
r
ed
u
ce
d
an
d
ar
e
p
lace
d
ju
s
t
b
ef
o
r
e
n
o
d
es
p
r
e
d
icted
to
b
e
f
a
u
lty
,
th
e
b
ac
k
p
r
o
p
a
g
atio
n
tim
e
a
n
d
f
au
lt
r
ec
o
v
er
y
tim
e
ar
e
s
ig
n
if
ican
tly
r
ed
u
ce
d
in
tu
r
n
allo
win
g
f
o
r
q
u
ick
e
r
s
y
s
tem
r
esp
o
n
s
e,
ev
en
in
p
r
ese
n
ce
o
f
tr
an
s
ien
t
f
au
lts
.
4.
CO
NCLU
SI
O
N
T
h
e
in
cr
e
asin
g
r
elian
ce
o
n
I
o
T
ap
p
licatio
n
s
ac
r
o
s
s
v
a
r
io
u
s
s
ec
to
r
s
s
u
ch
as
h
ea
lth
ca
r
e,
in
d
u
s
tr
y
a
n
d
ag
r
icu
ltu
r
e
f
o
cu
s
o
n
th
e
cr
itic
al
n
ee
d
f
o
r
r
o
b
u
s
t
FTM
s
.
T
r
an
s
ien
t
s
o
f
twar
e
f
au
lts
p
o
s
e
s
ig
n
if
ican
t
r
is
k
s
to
th
e
p
er
f
o
r
m
an
ce
a
n
d
r
eliab
ilit
y
o
f
th
ese
s
y
s
tem
s
.
T
h
e
ex
is
tin
g
FTM
s
s
tr
u
g
g
le
with
in
e
f
f
icien
cies
r
elate
d
to
ch
ec
k
p
o
i
n
tin
g
o
v
er
h
ea
d
s
,
as
ch
ec
k
p
o
i
n
ts
ar
e
ty
p
ically
p
la
ce
d
f
r
eq
u
en
tly
,
p
e
r
io
d
ically
a
n
d
in
d
is
cr
im
in
ately
ac
r
o
s
s
all
n
o
d
es,
lead
in
g
to
u
s
eless
an
d
f
o
r
ce
d
ch
ec
k
p
o
in
ts
,
f
o
llo
wed
b
y
d
o
m
in
o
ef
f
ec
t
an
d
u
n
n
ec
ess
ar
y
r
eso
u
r
ce
co
n
s
u
m
p
tio
n
.
T
h
e
r
e
s
ea
r
ch
in
tr
o
d
u
ce
s
a
n
o
v
el
C
I
N
b
ased
C
I
C
-
FTM
th
at
in
co
r
p
o
r
ates
L
STM
d
ee
p
lear
n
in
g
alg
o
r
ith
m
s
to
p
r
ed
ict
f
au
lt o
cc
u
r
r
en
ce
s
an
d
s
tr
ateg
ic
ally
p
lace
c
h
ec
k
p
o
in
ts
b
e
f
o
r
e
t
h
e
p
r
ed
icted
f
au
lty
n
o
d
es.
T
h
e
p
r
ed
ictio
n
ac
cu
r
ac
y
o
b
tain
ed
is
9
8
.
5
%.
T
h
is
r
ed
u
ce
s
th
e
n
u
m
b
e
r
o
f
c
h
ec
k
p
o
in
ts
p
lace
d
,
th
at
is
at
<1
5
%
n
o
d
es
o
f
th
e
o
v
er
all
n
o
d
es
an
d
<1
0
%
o
f
th
em
g
o
u
s
e
less
,
wh
ich
p
r
o
p
o
r
tio
n
ally
r
ed
u
ce
s
r
o
llb
ac
k
tim
e
Evaluation Warning : The document was created with Spire.PDF for Python.