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.
4
,
A
ugus
t
2020
,
pp.
4416
~
44
25
IS
S
N:
20
88
-
8708
,
DOI: 10
.11
591/
ijece
.
v
10
i
4
.
pp
4416
-
44
25
4416
Journ
al h
om
e
page
:
http:
//
ij
ece.i
aes
c
or
e.c
om/i
nd
ex
.ph
p/IJ
ECE
Self
-
ch
eckin
g method for f
ault tole
ra
n
ce solu
tion
in wir
eless senso
r netw
ork
Muayad
S
ad
i
k C
r
oock,
S
aja Dh
yaa Khu
der
,
Z
ah
r
aa Abbas H
as
s
an
Com
pute
r
Engi
n
ee
ring
Depa
r
tment,
Univ
ersity
of
Technol
og
y
,
Ir
a
q
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
Ja
n 6
, 2020
Re
vised
Ma
r
2
2
,
2020
Accepte
d
Ma
r
28
, 202
0
Rec
en
tly
,
th
e
wire
le
ss
sensor
net
work
(W
SN
)
has
bee
n
c
onsidere
d
in
diffe
ren
t
appl
i
c
at
ion
,
pa
rticula
r
l
y
in
emerge
n
c
y
s
y
stems
.
Therefore,
i
t
is
important
to
k
ee
p
t
hese
n
etw
orks
in
high
rel
i
ability
usi
ng
software
engi
ne
eri
ng
t
ec
h
nique
s
in
the
fi
el
d
of
fau
lt
tol
e
ran
ce.
Thi
s
pap
er
proposed
a
fau
lt
nod
e
detec
t
ion
m
et
hod
i
n
W
SN
using
the
self
-
ch
ec
king
te
chni
qu
e
ac
cor
d
ing
to
the
rule
s
of
softwa
re
engi
n
ee
r
ing.
The
n,
the
d
etec
t
ed
fau
l
ted
node
is
cove
r
ed
emplo
y
ing
the
r
ea
ding
of
n
ea
res
t
nei
ghbor
node
s
(sensors
).
In
addi
ti
on
,
the
proposed
m
et
hod
sends
a
m
essa
ge
for
m
ai
nte
na
nce
to
solv
e
the
f
aul
t
.
The
p
roposed
m
et
hod
ca
n
red
u
ce
the
ti
m
e
b
et
we
en
t
he
de
te
c
ti
o
n
and
re
cove
r
y
of
a
fau
lt
to
pre
v
en
t
the
conf
usion
of
ado
pt
ing
wro
ng
rea
d
ings,
in
which
the
d
et
e
ct
ion
is
m
aki
ng
with
m
ista
ke.
Moreove
r
,
it
guar
ant
e
es
the
re
li
ab
il
i
t
y
o
f
the
W
SN
,
in
te
rm
s
of
oper
at
i
on
and
data
trans
m
ission.
The
proposed
m
et
hod
has
b
e
en
t
este
d
over
diffe
r
ent
sce
n
ari
os
and
the
obta
in
ed
re
sults
s
how
the
superior
eff
ic
i
ency
in
te
rm
s
of
rec
ove
r
y
,
rel
i
abi
l
ity
,
and
c
onti
nuous data tr
ansm
ission.
Ke
yw
or
d
s
:
Fault t
olera
nce
Self
-
c
heck
i
ng t
echn
i
qu
e
So
ft
war
e
engin
eerin
g
WSN
Copyright
©
202
0
Instit
ut
e
o
f Ad
vanc
ed
Engi
n
ee
r
ing
and
S
cienc
e
.
Al
l
rights
reserv
ed
.
Corres
pond
in
g
Aut
h
or
:
Muay
ad Sadi
k C
roock,
Dep
a
rtm
ent o
f C
om
pu
te
Engi
neer
i
ng,
Un
i
ver
sit
y o
f Te
ch
no
l
og
y
,
Al
-
sinaa
Street
, Bag
hdad
, Ira
q
.
Em
a
il
:
1
Muay
a
d.
S
.Cr
oo
c
k@u
otech
no
l
og
y.
ed
u.
iq
;
2
1200
99@uotec
hn
olog
y.edu.i
q
;
3
120052@
uote
chnolo
gy.e
du.iq
1.
INTROD
U
CTION
It
is
well
kn
own
that
the
WSN
has
been
c
on
si
der
e
d
in
num
ero
us
fiel
ds
of
the
m
od
ern
li
fe,
su
ch
a
s
sm
art
bu
il
din
gs
an
d
em
erg
en
cy
syst
e
m
s
[1
]
.
These
a
ppli
cat
ion
s
prov
i
de
the
W
S
N
with
high
im
po
rtan
ce
and
sh
oul
d
be
m
or
e
reli
able,
fl
exible
an
d
e
xt
end
a
ble.
T
hes
e
prop
e
rtie
s
ha
ve
bee
n
a
ddr
essed
by
desi
gn
i
ng
the
so
ft
war
e
al
gorithm
s
accord
in
g
to
the
sof
tware
en
gin
ee
r
ing
te
ch
niques
[2
]
.
The
fa
ult
toleran
ce
m
et
ho
d
is
on
e
of
the
m
os
t
i
m
po
rtant
pa
r
t
in
the
softwa
re
en
gin
ee
rin
g
that
can
gu
a
ra
ntee
the
op
e
rat
ion
of
a
syst
em
with
high
reli
abili
ty.
T
o
be
m
or
e
preci
se,
t
he
sel
f
-
chec
king
te
c
hniq
ue
ca
n
be
c
on
si
der
e
d
f
or
c
heck
i
ng
the
sta
tus
of
su
c
h
syst
em
based
on
th
e
rea
di
ng
s
with
the
ti
m
e.
The
c
once
pts
of
this
te
c
hniq
ue
a
re
bu
il
t
on
al
locat
in
g
uppe
r
and
l
ow
e
r
th
re
sh
ol
d
values
f
or
rig
ht
operat
ion
to
any
par
t
of
a
syst
em
.
These
values
a
re
us
e
d
for
c
he
ckin
g
the v
al
idit
y
of
t
he
syst
em
, in
w
hic
h
the
f
a
ult can
be dete
ct
e
d
e
ff
ic
ie
ntly
[3
]
.
In
ge
ner
al
,
t
he
fau
lt
in
WSN
are
occ
urre
d
m
ai
nly
in
the
nodes
(se
ns
ors
)
[
1].
T
hese
fa
ults
can
be
cl
assifi
ed
into
hardw
a
re
an
d
so
ftwa
re.
T
he
se
two
cl
assifi
cat
ion
s
are
al
s
o
bei
ng
unde
r
the
per
m
anent
and
tem
po
rar
y faul
ts.
As
m
entione
d
ab
ov
e,
t
he
s
el
f
-
chec
ki
ng
a
ppr
oach
is use
d
f
or
d
et
ect
in
g
the
fa
ult
of
no
des
a
nd
the
m
on
it
or
in
g
syst
e
m
send
s
a
m
essage
f
or
m
ai
ntenan
ce
f
or
s
olv
i
ng
t
he
occurre
d
pro
ble
m
s
[2
]
.
Th
r
oughout
the
m
ai
ntenance
pr
oc
ed
ur
e
,
the
syst
e
m
can
cov
er
this
er
r
or
in
di
ff
e
ren
t
ways
dep
e
ndin
g
on
the
desig
n
an
d
sta
tus of t
he
rel
at
ed
co
ndit
ion
s
[3].
In
this
pa
per
,
a
m
on
it
or
i
ng
sy
stem
fo
r
WSN
is
pr
op
os
e
d
ba
sed
on
s
of
t
wa
re
en
gin
ee
rin
g
te
chn
iq
ues
.
The
pr
opos
e
d
m
et
ho
d
a
dopt
s
the
fau
lt
to
le
ran
ce
a
ppro
a
ch
f
or
m
anag
i
ng
t
he
operati
on
of
node
s
inside
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
Self
-
checki
ng
meth
od for
fa
ul
t t
oleran
ce
s
olu
ti
on in
wi
rel
e
ss sensor
n
et
w
or
k
(
M
ua
y
ad
Sadik Cr
oock
)
4417
the
W
S
N.
I
n
the
fa
ult
detect
ion
pa
rt
of
t
he
pro
po
se
d
m
et
ho
d,
the
sel
f
-
c
he
ckin
g
te
ch
nique
is
co
ns
ide
re
d
with
uppe
r
an
d
lo
w
er th
re
shold
for
each
node
i
nc
lud
e
d
in a clu
ste
r.
T
he
de
sig
ne
d WSN
c
onta
ins num
ero
us
cl
us
te
rs
and
eac
h
of
w
hich
incl
ud
es
diff
e
r
e
nt
node
s.
At
a
tim
e
t
hat
the
fau
lt
is
occurre
d
in
a
node,
the
propose
d
m
et
ho
d
ta
ke
s
the
nei
ghbor
no
des
rea
ding
to
ta
ckle
the
pro
bl
e
m
in
an
eff
ic
ie
nt
way.
T
he
r
ecov
e
ry
operat
ion
is
perform
ed
by
cal
culat
ing
the
aver
a
ge
value
of
near
e
st
nei
ghbor
nodes
unti
l
the
fau
lt
is
so
lve
d.
T
he
propose
d
syst
e
m
is
te
st
ed
unde
r
dif
fe
ren
t
co
ndit
ions
and
the
ach
ie
ved
res
ults
pro
ve
the
sat
isfact
ory
in
te
rm
s
of
accuracy a
nd e
ff
ic
ie
ncy.
2.
RE
LATE
D
W
ORK
The
ap
plica
ti
on
s
of
W
S
N
and
relat
ed
s
upporte
d
syst
e
m
s
at
tract
the
at
te
ntion
s
o
f
num
ero
us
researc
hers
in
disti
nct
fiel
ds
.
Each
gr
oup
of
them
ta
ckled
a
side
of
ap
pe
ared
pro
blem
s,
su
ch
as
fa
ults,
data
transm
issi
on
.
All
these
rese
arch
es
f
ocu
se
d
on
i
ntr
oducin
g
r
obus
t
syst
e
m
s
with
hig
h
accuracy,
reli
abili
ty
,
flexibili
ty
an
d
exten
dib
il
it
y.
Ther
e
f
or
e,
the
con
ce
pts a
nd a
ppr
oach
es
of s
of
t
war
e e
ngine
erin
g were
a
do
pted.
In
[
4],
the
au
thors
pr
ese
nte
d
a
de
velo
pe
d
m
anag
em
ent
m
et
ho
d
f
or
W
S
N
base
d
on
s
of
t
ware
eng
i
neer
i
ng
r
ul
es.
They
cl
ass
ifie
d
the
sit
uat
ion
s
of
WSN
in
te
rm
s
of
inf
rastr
uctur
e
an
d
fou
nd
t
he
s
uitable
adap
ti
ng
softw
are.
Wh
il
e
the
aut
hors
of
[5
]
pro
pose
d
a
re
al
tim
e
oil
an
d
gas
m
on
it
ori
ng
syst
em
based
on
WSN
us
i
ng
th
e
so
ftw
are
de
ve
lop
m
ent
as
a
par
t
of
softwa
r
e
eng
i
neer
in
g.
The
prese
nted
syst
e
m
too
k
care
of
diff
e
re
nt
equ
i
pm
ent,
su
ch
as
pu
m
ps
,
sens
or
s
and
m
et
ering
infr
ast
r
uctu
re.
In
ad
diti
on
,
it
m
anag
ed
the
fl
ow
of
oil
and
gas
se
qu
e
ntial
ly
.
In
[6
]
,
a
ne
w
f
ra
m
ewo
r
k
f
or
WSN
was
pr
e
sented
.
The
in
tro
du
ce
d
f
ram
ewor
k
was
bu
il
t
ba
s
ed
on
softwa
r
e
en
gin
ee
rin
g
chall
en
ges
i
nc
lud
in
g
flexi
bi
li
t
y
and
sc
a
la
bili
ty
.
It
wor
ke
d
as
a
m
idd
le
war
e
that
cou
ld
m
a
nag
e
the
data
m
ining
,
ene
r
gy
con
su
m
ption
and
sel
f
-
orga
nizat
ion
.
In
a
ddit
ion
,
the
auth
ors
of
[7
]
pe
rfor
m
ed
a
m
od
el
ing
fr
am
ewo
r
k
f
or
analy
zi
ng
the
featur
es
i
n
te
rm
s
of
var
ia
ti
on
a
nd
com
m
inati
on
.
These
feat
ur
es
pro
vid
e
d
a
co
m
ple
te
view
for
the
W
S
N
bas
ed
ag
ricult
ure
syst
e
m
.
All
pr
opos
e
d
al
gorithm
s
and
so
ft
war
e
we
re
desig
ne
d
un
de
r
the
ru
le
s
of
s
of
t
war
e
e
ngine
erin
g.
In
[8
]
,
a
searc
hing
ba
s
ed
on
so
ft
war
e
e
ngineeri
ng
was
i
m
ple
m
ented
in
WSN.
Th
e
i
m
ple
m
ented
m
et
ho
d
a
dopt
ed
the
m
ulti
-
ob
j
ect
iv
e
al
gorithm
to
be
so
lved
us
in
g
gen
et
ic
m
e
thod.
It
al
so
co
ns
i
der
e
d
the
fixe
d
ob
sta
cl
es
in
th
e
searching
pr
ocess.
The
aut
hors
of
[
9]
introd
uce
d
ha
rdwa
re
an
d
softwa
re
in
f
rastr
uctur
e
f
or
W
S
N
base
d
on
t
he
re
gu
la
ti
on
of
so
ft
war
e
e
ng
i
neer
i
ng,
pa
rtic
ularly
the
software
requirem
ents.
T
he
f
oc
us
e
d
on
the
r
ecei
ver
a
nd
how
it
descr
i
bed
it
sel
f
to
the
oth
e
rs
nodes
of
W
S
N.
The
wor
k
was
done
in
physi
cal
la
ye
r
i
n
real
tim
e
s
itu
at
ion.
It
is
i
m
po
rtant
to
no
te
that
the
te
st
resu
lt
s
of
the
li
te
ratur
e
r
e
searc
h
works
we
re
cl
ear
en
ough
to
prov
e
the cla
im
o
f
au
thors.
At
the
ot
her
ha
nd,
the
fa
ult
to
le
ran
ce
of
s
of
t
war
e
en
gin
ee
ri
ng
wa
s
a
dopte
d
in
s
olv
in
g
t
he
pro
blem
of
reli
abili
ty
.
In
[
10
]
,
a
s
urvey
on
fa
ult
detect
ion
a
nd
the
sol
ution
s
we
re
presente
d.
It
c
ons
ide
red
the
re
la
ti
on
betwee
n
dif
fere
nt
ty
pes
of
fa
ults
in
W
S
N
node
s
an
d
the
c
hoos
i
ng
of
s
uitable
fau
lt
dete
ct
or
as
well
as
fau
lt
tolerance
m
et
h
od.
T
his
relat
ion
co
uld
le
a
d
to
the
root
of
f
aulte
d
node,
in
w
hich
the
so
l
ution
was
pres
ented
.
The
aut
hors
of
[11]
propos
ed
a
m
anag
e
m
ent
arch
it
ect
ur
e
of
W
S
N
base
d
on
f
aul
t
tolerance
ap
proac
h.
The
pro
posed
arc
hitec
ture
ta
ckled
tw
o
s
ides
of
c
halle
ng
e
s
insi
de
WSN.
T
he
fir
st
one
is
the
pow
e
r
consum
ption
,
wh
il
e
the
ot
he
r
side
is
the
f
ault
tolerance
to
co
ver
the
s
el
f
-
or
gan
iz
e
d
netw
ork.
T
his
was
to
increase
the
r
el
ia
bili
ty
of
t
he
WSN
a
nd
beco
m
e
fau
lt
tolerance
.
I
n
[
12
]
,
a
new
t
r
end
of
guara
nt
eei
ng
the
reli
abili
ty
of
WSN
was
pr
ese
nted
ba
se
d
on
fa
ult
tolerance
m
et
ho
d.
Both
t
heory
a
nd
ap
plica
ti
ons
we
re
consi
der
e
d
in
the
ori
ent
e
d
r
esearch
.
In
[
13]
,
a
cl
us
te
r
he
ad
base
d
faul
t
tolerance
syst
e
m
was
pro
po
s
ed
.
The
ob
j
ect
ives
of
the
pro
po
s
ed
syst
em
were
to
keep
ey
e
on
t
he
W
S
N
t
hro
ughout
the
op
e
rati
on
in
te
rm
s
of
data tra
ns
m
issio
n, m
ob
il
it
y and
fau
lt
occur.
The
so
l
utions
wer
e
al
so
pr
ovide
d
to
guara
ntee
the
reli
ab
il
ity.
The
auth
or
s
of
[
14]
us
ed
an
age
nt
factor
t
o
pro
pose
a
fa
ult
tolerance
m
et
ho
d
for
W
S
N.
T
he
platfo
rm
of
the
pro
pose
d
m
et
hod
w
orke
d
acro
s
s
diff
e
re
nt
le
vels
including
se
nsors
,
cl
us
te
r
hea
ds
an
d
base
sta
ti
on
s.
I
n
[
15]
,
the
r
esea
rch
e
rs
e
m
plo
ye
d
the
f
ault
tolerance
m
et
h
od
f
or
re
du
ci
ng
the
re
dunda
ncy
an
d
pac
ke
t
loss
ov
e
r
the
bro
ken
li
nks
(p
at
hs
)
of
WSN
a
nd
fin
ding
the
sol
ution
s
to
kee
p
the
W
S
N
in
op
erati
on.
More
over,
they
su
pp
or
te
d
the
secur
it
y
of
the
data
transm
issi
on
usi
ng
cry
ptogra
ph
y
m
et
ho
ds
.
The
aut
hors
of
[1
6]
desi
gn
e
d
a
so
ftwar
e
e
nginee
rin
g
base
d
fau
lt
detect
ion
m
et
ho
d
f
or
op
ti
m
izati
on
m
et
ho
d
t
hat
was
pro
posed.
In
[
17
]
,
a
li
st
of
the
rec
ent
m
et
ho
ds
of
fa
ult
detect
ion
a
nd
diag
nosin
g
wa
s
pr
ese
nted
.
It
include
d
a
real
su
r
vey
on
fau
l
t
detect
ion
in
WSN.
T
he
res
earche
r
s
of
[18]
pr
opose
d
a
fa
ult
detec
ti
on
based
cl
ust
ering
netw
ork
.
The
pro
po
se
d
m
et
ho
d
i
nvolve
d
tw
o
side
s;
fi
rstly
,
the
cl
us
te
r
c
onfi
gurati
on,
w
hile
the
oth
e
r
is
fau
lt
detec
ti
on
an
d
re
co
ver
y.
In
[19],
a
m
et
ho
d
of
so
lvin
g
the
pro
blem
of
node
is
olati
on
i
n
A
GR
-
M
AC
prot
oco
l
was
prese
nted
.
The
pr
opos
e
d
m
et
ho
d
use
d
fa
ult
tolerance
te
c
hniq
ue
f
or
m
ain
ta
inin
g
the
i
so
la
ti
on
pro
bl
e
m
.
The
aut
hors
of
[
20
]
propose
d
a
gu
a
r
anteed
al
gorithm
of
bi
oin
f
or
m
at
ic
connecti
vity
in
WSN
base
d
on
fau
lt
tole
rance
te
ch
nique
f
or
am
bu
la
nce
syst
e
m
.
In
[
21]
,
a
WSN
bas
ed
wildfire
detect
ion
syst
e
m
was
pro
po
se
d
ba
se
d
on
sel
f
-
org
anizat
ion
a
nd
fau
lt
tolerance
app
r
oach
e
s.
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
.
4
,
A
ugus
t
2020
:
4416
-
4425
4418
At
the
ot
her
ha
nd,
a
syst
em
based
virt
ual
bio
lo
gical
co
m
m
un
it
y
of
na
ture
li
ving
st
yl
e
has
been
pro
po
se
d.
The
evo
l
ution
a
ry
gam
e
m
et
ho
d
w
as
ad
op
te
d
to
intr
oduce
a
fa
ult
tolerance
pro
cedure
for
m
anag
in
g
the
pro
blem
s
o
f
W
S
N
[
22
]
.
I
n
[23],
a
fau
lt
tolerance
m
et
ho
d
was
pr
ese
nte
d
f
or
s
olv
in
g
the
pro
blem
of
cl
us
te
r
head
fail
ures.
This
wa
s
do
ne
by
p
la
nn
i
ng
t
he
s
olu
ti
on
ba
sed
on
virt
ual
heads
an
d
th
e
colle
ct
ed
in
for
m
at
ion
.
The
resea
rc
her
of
[
24
]
pr
opose
d
a
data
tran
sm
issi
on
proto
col
in
save
sit
uation
us
in
g
fa
ult
tolerance
m
et
ho
d.
This
protoc
ol
us
e
d
the
resi
dual
inf
or
m
at
ion
reg
a
rd
i
ng
the
li
nk
s
of
WSN
t
o
bu
il
d
a
plan
for
tole
ran
t
the
fa
ults.
In
[
25
]
,
t
he
cl
ust
er
hea
d
sel
ec
ti
on
m
et
ho
d
w
as
pro
posed
ba
sed
on
fa
ult
tolerance
an
d
fu
z
zy
-
log
ic
a
ppr
oa
ches
.
The
s
of
t
c
om
pu
ti
ng
m
et
ho
d
pro
vid
e
d
the
s
yst
e
m
with
optim
a
li
ty
in
al
locat
ing
the
cl
us
te
r
hea
d,
in
wh
ic
h
the
fail
ur
e
is
c
ov
e
re
d
w
el
l.
I
n
[
26,
27]
,
dif
fer
e
nt
m
et
ho
ds
for
WSN
st
ruct
ur
e
a
nd
cl
us
t
erin
g
m
et
ho
ds
ha
ve
been
pro
po
se
d
to
so
l
ve
num
ero
us
pr
oblem
.
They
us
ed
softwa
re
eng
i
neer
i
ng
re
gu
la
ti
ons
in
bu
il
di
ng
the pr
opos
e
d m
et
ho
ds.
3.
THE
PROPO
SED SYSTE
M
As
m
entione
d
earli
er,
the
pro
po
se
d
syst
e
m
includes
t
wo
sides
of
work
based
on
s
of
t
war
e
eng
i
neer
i
ng
m
et
hods
that
are u
se
d
f
or
g
ua
ra
ntee
the
reli
abi
li
ty
of
the
ad
opte
d
W
S
N.
T
he
first
one
is
rela
te
d
to
detect
the o
cc
urre
d
fau
lt
i
n
a
n
ode at
the
ad
opte
d WSN
us
i
ng
t
he
sel
f
-
c
hec
king
m
et
ho
d. Wh
il
e
t
he
sec
ond
one
ta
kes
care
of
f
ind
in
g
t
he
s
olut
ion
of
t
he
detect
ed
fa
ult
us
i
ng
the
a
vaila
bl
e
tolera
nce
in
fin
ding
the
rec
ov
e
ry
values
w
it
hi
n
t
he neig
hborh
ood area
.
3.1.
Sy
s
tem
s
truc
t
ure
In
this
sect
io
n,
the
gen
e
ral
structu
re
of
th
e
pro
posed
syst
e
m
is
exp
la
ine
d
as
a
blo
ck
dia
gr
am
sh
own
in
Fig
ur
e
1
.
It
is
sh
own
t
hat
the
colle
ct
ed
read
i
ng
data
from
the
un
de
rly
ing
sen
sors
of
al
locat
ed
WSN
i
s
entere
d
to
the
fau
lt
detect
io
n
un
it
.
T
his
unit
is
res
pons
i
ble
on
detect
ing
the
fau
lt
t
ha
t
can
b
e
occ
urred
in
a
node
us
in
g
t
he
sel
f
-
chec
ki
ng
a
ppr
oac
h.
This
a
ppro
ac
h
is
base
d
on
uppe
r
a
nd
l
owe
r
thr
esh
old
val
ues
f
or
each
sen
sor
as
a
tolerance
range.
T
he
recei
ve
d
val
ues
that
f
al
l
within
the
outa
ge
tole
ran
ce
area
is
cl
assifi
ed
as
a fa
ult. If the
re
is no
detect
ed fault
, t
he data
is sent t
o
the
sy
stem
saf
el
y.
Figure
1
.
The
pro
po
se
d
syst
e
m
stru
ct
ur
e
The
detect
ed
fau
lt
s
are
pas
sed
to
the
fa
ul
t
tolerance
unit
.
This
un
it
is
responsi
bl
e
on
fin
ding
the
s
olu
ti
on
to
the
pro
blem
of
fa
ult
us
i
ng
the
avail
a
bl
e
tolera
nces.
The
so
l
ution
ad
op
te
d
in
the
pro
po
s
e
d
m
et
ho
d
us
e
s
th
e
val
ues
of
nei
ghbor
no
des
w
it
h
co
rr
ect
v
al
ue
s
f
or
guessi
ng
the
e
xpect
ed
r
eadin
g
of
the
f
aulte
d
sens
or
no
de.
Wh
il
st
the
pro
cessi
ng
of
fin
di
ng
t
he
s
olu
ti
ons,
a
c
omm
un
ic
at
ion
betwee
n
fau
lt
dete
ct
ion
unit
and fa
ult t
olera
nce
un
it
. T
he r
easo
n
be
hind t
his co
m
m
un
ic
at
ion
is e
xch
a
ng
ing
t
he
in
form
at
ion
a
nd corre
ct
an
y
error
ca
n
be
ha
pp
e
ne
d
in
fin
di
ng
the
tole
ran
t
so
luti
on.
The
resu
lt
s
of
fa
ult
detect
ion
an
d
the
tolera
nt
s
olu
ti
on
is
app
li
e
d
to
sen
d
the
c
orre
ct
ed
rea
dings
as
well
as
the
est
i
m
at
ed
read
in
g
to
the
sy
stem
fo
r
c
onti
nui
ng
the pr
ocess.
3.2.
Prop
os
ed
al
gori
th
m
It
is
well
known
t
hat
the
s
oft
war
e
al
gorith
m
is
necessar
y
to
be
pro
posed
f
or
m
anagi
ng
t
he
fa
ult
detect
ion
an
d
fin
ding
the
s
ol
ution
.
I
n
t
his
sect
ion,
the
pro
po
se
d
al
gorithm
is
pr
ese
nted
a
s
a
flo
wch
a
rt,
sh
ow
n
in
Fi
gur
e
2.
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
Self
-
checki
ng
meth
od for
fa
ul
t t
oleran
ce
s
olu
ti
on in
wi
rel
e
ss sensor
n
et
w
or
k
(
M
ua
y
ad
Sadik Cr
oock
)
4419
Figure
2. Flo
w
char
t
of the
pro
po
s
ed
alg
or
it
hm
Ba
sed
on
the
flo
wch
a
rt
of
t
he
pro
posed
al
gorithm
,
the
work
i
ng
proce
dure
can
be
su
m
m
arized
in
the foll
owin
g:
a.
Coll
ect
i
ng
the
read
i
ng
s
of t
he
sen
s
or no
des
i
n
the
ado
pted
WSN.
b.
Checki
ng
the
validit
y
of
the
received
read
i
ng
s
by
the
cl
ust
er
hea
d.
I
n
c
ase
the
read
i
ngs
are
no
t
va
li
d,
a re
qu
est
is
sent
b
ack
to
t
he n
od
e
s fo
r prov
i
di
ng
a
nothe
r
c
opy o
f readi
ngs.
c.
The
valid
readi
n
gs
are
c
hec
ke
d
f
or
fa
ult
de
te
ct
ion
us
in
g
the
sel
f
-
c
hec
kin
g
process
bas
ed
on
the
uppe
r
and lo
wer t
hr
e
sh
ol
d
as
explai
ned m
at
he
m
ati
cal
ly
in
(1):
≥
(
,
)
≥
(1)
Wh
e
re
an
d
ar
e
the
uppe
r
a
nd
l
ower
th
res
hold
values.
In
case
of
fau
lt
oc
cur,
t
he
pro
po
sed
al
gorithm
sends
the
inf
orm
ati
on
t
o
the
ne
xt
ste
p.
Otherwis
e,
the
rea
dings
can
be
pas
sed
to
the
syst
em
.
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
.
4
,
A
ugus
t
2020
:
4416
-
4425
4420
It
is
i
m
po
rtant
to
highli
ght
th
at
the
and
val
ues
are
sel
ect
e
d
de
pe
nd
i
ng
on
the
ty
pe
of
use
d
sens
or
as
well
as the
s
urr
ound
ing
c
onditi
ons.
d.
The
fa
ulted
re
adin
gs
are
pas
sed
to
the
fa
ul
t
tolerance
unit
.
This
unit
adopts
the
val
id
read
i
ng
s
of
the
cl
os
e
neig
hbour
no
des.
The
m
os
t
i
m
p
or
ta
nt
conditi
on
is
the
validit
y
of
these
rea
ding
as
cou
l
d
be
happe
ned the
e
ven the cl
os
e
ne
ighbou
r
rea
di
ng
s
have
f
a
ults as we
ll
accord
ing
t
o
the
foll
owin
g
e
qu
at
io
n:
(
,
)
=
,
{
(
±
,
±
)
,
≥
(
±
,
±
)
≥
(
,
)
,
}
(2)
Wh
e
re
(
,
)
is
the
instant
rea
ding
of
a
no
de
with
locat
io
n
(i,
j
)
i
n
t
he
c
ov
ere
d
area.
(
±
,
±
)
is
the n
e
xt
neig
hbor
read
i
ngs,
i
ncr
ease
d by
∈
to
ta
l WSN dim
ensio
n, a
nd k
=
{1,2,
3,
…
,K
}
.
e.
It
is
no
te
d
fro
m
(2
)
hat
the
i
ns
ta
nt
rea
ding
of
a
no
de
can
be
re
placed
with
the
neig
hbour
nodes
in
ca
se
they
are
valid.
In
orde
r
to
incr
e
ase
the
validit
y
of
est
im
at
ion
an
d
the
a
ccu
ra
cy
,
the
a
ver
a
ge
value
can
be
adopted
am
ongs
t
th
e
valid
c
lose
neig
hbour
s
in
the
est
im
a
ti
on
proces
s.
T
her
e
fore,
the
e
stim
at
ed
value
(
,
)
of r
ea
ding
of a
nod
e
(
,
)
is m
a
them
at
ic
ally wr
it
t
en
as:
(
,
)
=
∑
(
±
,
±
)
=
(3)
Wh
e
re
is t
he
t
otal n
um
ber
of
adopted
n
ei
ghbor
no
des
to
the
und
e
rly
ing n
ode
(
,
)
.
f.
Checki
ng the e
stim
at
ed
value of
a
node
(
,
)
if it
is wit
hin t
he u
pp
e
r
a
nd lo
wer t
h
res
hold
vlu
e
s as:
≥
(
,
)
≥
(4)
g.
Sendin
g
the
v
a
li
d
value
s (sen
so
rs
’ rea
dings) t
o
the
syst
em
f
or futu
re
proce
ssing.
4.
RESU
LT
S
In
order
t
o
te
st
the
pro
po
se
d
m
et
ho
d
that
ha
s
been
e
xpla
in
ed
in
early
sect
ion
s
,
a
si
m
ulatio
n
m
od
el
is
adopted
.
T
he
pro
posed
m
od
el
include
s
a W
S
N
with f
our
cl
us
te
rs
,
each o
f
wh
ic
h
in
vo
l
ves
cl
us
te
r
hea
d
a
nd
si
x
sens
or
s
,
as
show
n
in
Fi
gure
3.
The
cl
us
te
r
hea
ds
are
c
onnecte
d
to
the
dev
ic
e
m
anag
er
as
a
base
sta
ti
on.
Tem
per
at
ur
e
s
ens
or
s
a
re
a
dopted
in
this
si
m
ula
te
d
m
od
el
that
can
be
ch
ang
e
d
to
dif
fere
nt
ty
pes
of
se
ns
ors
.
This
m
od
el
is
bu
il
t
us
in
g
Vis
ual
St
ud
i
o
C
#
la
nguag
e
a
nd
t
he
i
nterf
ace
presents
t
he
resul
ts
in
nu
m
ero
us
col
or
for
m
or
e
ex
planati
on.
T
he
gr
een
bo
xes
re
pr
esent
the
well
work
i
ng
node
s
,
w
h
il
e
the
re
d
boxes
a
re
the
f
aulte
d
nodes
.
I
ns
ide
each
bo
x,
the
instant
sens
or
read
i
ng
is
s
hown.
The
node
s
are
co
nn
ect
e
d
to
the
cl
us
te
r
he
a
d
within eac
h o
ne
ind
i
viduall
y.
Figure
3
.
Sim
ulate
d
m
od
el
interf
ace,
case st
udy
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
Self
-
checki
ng
meth
od for
fa
ul
t t
oleran
ce
s
olu
ti
on in
wi
rel
e
ss sensor
n
et
w
or
k
(
M
ua
y
ad
Sadik Cr
oock
)
4421
Diff
e
re
nt
case
stud
ie
s
hav
e
been
c
onside
red
t
o
te
st
the
syst
e
m
un
de
r
cha
ngeable
conditi
on
s
.
The
case
stu
dy
,
show
n
in
Fi
gure
3
,
e
xp
la
in
s
that
the
cl
us
te
r
one
has
t
hr
ee
fau
lt
ed
se
nsor
nodes,
m
entioned
i
n
red
c
olor.
I
n
add
it
io
n,
cl
ust
er
two
i
nvolve
s
two
fau
lt
ed
sens
or
s
with
red
la
bel.
M
or
e
ov
er,
cl
us
te
r
thre
e
include
s
a
faul
te
d
sensor,
w
hile
the
four
th
cl
us
te
r
po
i
nts
three
sens
or
nodes
to
be
fa
ulted.
De
pe
nd
i
ng
on
the
pro
po
se
d
a
lgorit
hm
and
based
on
eq
uations
(1
-
4),
the
pro
po
se
d
m
et
h
od
c
om
pen
sat
es
the
fau
lt
ed
re
adin
gs
with
est
i
m
at
ed
values.
It
is
i
m
po
rtant
to
no
te
that
the
pro
po
se
d
al
go
rithm
a
m
on
gs
t
al
l
c
lusters
checks
the
read
i
ngs
of
neig
hbor
no
des
if
they
are
within
the
thr
esh
old
s
to
be
consi
der
e
d.
Otherwise,
it
consi
ders
the
nex
t
nei
ghbor
instea
d
to
gu
a
ra
ntee
the
est
i
m
ation
with
m
ini
m
u
m
er
ro
r
oc
cu
r.
T
he
si
m
ulati
on
para
m
et
ers
are
buil
t
based
on
the
upper
a
nd
lo
we
r
thres
ho
l
d
of
(
35
o
c)
and
(
55
o
c),
s
equ
e
ntial
ly
,
for
a
whole
Ju
ly
m
on
th
in Bag
hdad
cit
y at
I
ra
q.
Figure
4
s
how
s
the
sam
e
cas
e
stud
y
of
Fig
ure
3
afte
r
ap
pl
yi
ng
the
pr
opose
d
al
gorithm
.
I
can
be
seen
that
the
fa
ulted
read
i
ng
s of
th
e
sens
or
s
i
n
al
l
cl
us
te
rs
a
re
es
tim
a
te
d
bas
e
d
on
t
he
ne
ig
hbors
an
d
la
be
le
d
i
n
bl
ue
colo
r.
A
s
m
ent
ion
e
d
earli
er,
t
he
res
ults
are
m
easur
ed
for
thirty
days
am
on
g
J
uly
m
on
th.
The
pe
rfor
m
ance
of
each
se
ns
or
in
al
l
cl
us
te
rs
can
be
m
easur
e
d
accuratel
y.
Fi
gure
5
sho
ws
the
pe
rfo
rm
ance
of
se
nsor
no
des
in
cl
us
te
r
one.
Th
e
case
her
e
that
there
is
no
fau
lt
is
detect
ed,
and
al
l
senso
r
s
a
re
work
i
ng
well
in
cl
us
te
r
one
.
It is see
n
that t
he whole
read
i
ng
s
are
w
it
hi
n t
he
validit
y ra
nge
betwee
n upper
and l
ow
e
r
t
hr
es
holds
.
Figure
4
.
Ca
se
stud
y
of
Fig
ure
3
a
fter a
pply
in
g
the
pr
opos
e
d al
gorithm
Figure
5.
Cl
us
t
er
on
e
p
e
rfor
m
ance in
case
of
nor
m
al
w
orki
ng
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
.
4
,
A
ugus
t
2020
:
4416
-
4425
4422
At
the
othe
r
ha
nd,
the
sam
e
clu
ste
r
one
passe
s
by
w
or
st
c
on
diti
on
s
,
in
w
hi
ch
the
re
a
re
t
hree
se
nsors
su
f
fer
f
r
om
fau
lt
s.
This
can
le
ad
to
cha
ng
i
ng
in
the
read
in
gs
of
these
sen
s
or
s
.
These
fa
ul
ts
are
detect
ed
us
in
g
the
pro
pose
d
al
gorithm
that
is
base
d
on
sel
f
-
c
heck
i
ng
process
.
T
he
detect
ed
fau
lt
s
are
c
om
pen
sat
ed
by
app
ly
in
g
th
e
pro
po
se
d
fa
ul
t
tolerance
m
et
hod
base
d
on
the
neig
h
bor
no
des
as
exp
la
ine
d
e
arli
er
.
Figure
6
e
xpla
ins
the
rea
dings
of
sen
sors
inside
cl
ust
er
one,
w
he
re
the
fau
lt
ed
read
i
ngs
excee
d
the
uppe
r
thres
ho
l
d value
s.
It is
sho
wn that the
fa
ults a
re s
ta
rte
d
at
da
y s
eve
n,
t
we
lv
e an
d four
te
e
n, seq
ue
ntial
ly
.
Figure
6
.
Cl
us
t
er
on
e
p
e
rfor
m
ance in
case
of
f
a
ult detect
ion
Now
the
pro
po
sed
m
et
ho
d
shou
l
d
so
l
ve
this
pr
oble
m
by
a
pp
ly
in
g
the
pr
esented
al
gorit
hm
s.
This
is
done
e
ff
ic
ie
ntl
y as show
n
in
Figure
7. T
he
f
aulte
d
rea
dings
are
fixe
d by est
i
m
a
ti
ng
r
e
placed v
al
ues
de
pe
nd
i
ng
on
the v
al
i
d
re
adin
gs
of n
ei
ghbo
r
nodes
. To b
e
m
or
e p
reci
se, sen
s
or
no
de
n
um
ber
f
our
i
n
cl
us
te
r
one ha
s b
een
analy
zed in
Fi
gure
8. It is shown that the es
tim
a
te
d
values of
this se
nsor
a
re w
it
hin
t
h
e va
li
dity
r
ang
e be
tween
thres
ho
l
ds
with
accepta
ble er
r
or.
Figure
7.
Cl
us
t
er
on
e
p
e
rfor
m
ance in
case
of
f
ixi
ng f
a
ults
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
Self
-
checki
ng
meth
od for
fa
ul
t t
oleran
ce
s
olu
ti
on in
wi
rel
e
ss sensor
n
et
w
or
k
(
M
ua
y
ad
Sadik Cr
oock
)
4423
Figure
8
.
Se
nsor
four a
naly
sis re
gardin
g n
orm
al
, f
ault an
d f
ixed
Figure
9
il
lustrate
s
the
perf
or
m
ance
of
th
e
ad
op
te
d
WSN
with
f
our
c
lusters.
For
ea
ch
cl
us
te
r,
the av
e
ra
ge values
of no
rm
al
,
fau
lt
ed
and
fix
ed
cases
are
eva
luate
d
acc
ordi
ng to
t
he follo
wing e
qu
at
io
n:
(
,
)
(
)
=
∑
∑
(
,
)
=
=
(5)
w
he
re
(
,
)
(
)
is
the
a
ver
a
ge
values
of
t
he
cl
ust
er
z
,
an
d
z={
1,
2,
3,
…,
Z}
an
d
Z
is
the
num
ber
of
cl
us
te
rs
within
the
WSN.
is
the
nu
m
ber
of
al
l
rea
d
ing
s
of
sens
ors
inside
the
cl
us
te
r
z
,
re
presente
d
as
a
m
at
rix
of
(
6
X
30)
f
or
six
se
ns
ors
a
nd
thirty
days.
I
is
c
onside
r
ed
as
the
num
ber
of
sens
ors,
wh
ic
h
is
si
x,
wh
il
e
J
represe
nts
the
num
ber
of
da
ys
(thirty
da
ys).
T
hese
values
a
re
com
pa
red
i
n
this
fi
gure
a
nd
sho
wn
that
the
est
i
m
at
ed
values
a
re
ve
r
y
cl
os
e
to
the
norm
al
on
es,
wh
il
e
the
fa
ulted
rea
ding
inc
reased
i
n
nota
ble
way
.
This m
eans th
a
t t
he
pr
opos
e
d al
gorithm
an
d m
et
ho
d
a
re
w
orki
ng in
e
ff
ic
ie
nt w
ay
.
Figure
9
.
The
total
cluster
p
e
r
form
ance f
or
n
or
m
al
, f
aulte
d
and fi
xed cases
Table
1
sho
ws
the
reco
r
de
d
r
eadin
gs
of
sen
so
rs
ov
e
r
thirt
y
days
of
Ju
ly
m
on
th
f
or
cl
ust
er
three
at
the
pr
e
sent
of
fau
lt
occ
ur
,
a
s
an
exam
ple
to
the
pro
pose
d
syst
e
m
fau
lt
m
on
it
or
in
g.
I
t
can
be
seen
that
the
val
ues
of
read
i
ng
s
of
th
e
6
th
sen
sor
a
re
inc
reased
s
harply
after
t
he
10
th
day
as
the
fa
ult
ha
ppened.
In
a
ddit
ion
, t
he
r
eadi
ng
s
of se
ns
or
four
dec
re
ased to
bec
ome
zero
s a
nd thi
s is ind
ic
at
ed
a
s a f
ault as t
he values
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
.
4
,
A
ugus
t
2020
:
4416
-
4425
4424
wen
t
dow
n
the
lower
t
hr
es
hold
sta
rtin
g
f
ro
m
day
11
th
.
At
th
e
oth
er
ha
nd,
s
ens
or
tw
o
pr
ovides
w
ron
g
rea
dings
after
23
rd
day,
wh
e
re
the
values
are
in
creased
no
ta
bly.
The
values
,
highli
gh
te
d
with
ye
ll
ow
col
or
,
are
rec
orde
d
as
fau
lt
s
due
to
t
he
e
xceed
i
ng
of
up
per
an
d
lo
wer
t
hr
es
holds
(25
-
55
o
c)
,
wh
i
ch
are
the
m
axi
m
u
m
and m
ini
m
u
m
t
e
m
per
at
ures a
m
on
g
this m
onth.
Table
1
.
Cl
us
te
r 3 se
ns
ors
afte
r
fa
ult (i
n
se
nsor
2,
4,
6)
Day
Sen
so
r
1
Sen
so
r
2
Sen
so
r
3
Sen
so
r
4
Sen
so
r
5
Sen
so
r
6
1.
36
43
31
38
34
47
2.
40
40
49
40
43
49
3
.
44
37
31
34
31
48
4
.
39
42
41
44
32
32
5
.
47
32
36
49
34
38
6
.
50
36
48
43
41
36
7
.
46
48
41
49
50
34
8
.
46
46
32
30
49
37
9
.
32
33
31
45
40
43
10.
49
44
48
45
37
76
11.
39
37
40
0
39
70
12.
40
39
38
0
43
71
13.
34
38
39
0
31
76
14.
33
30
49
0
45
74
15.
39
43
34
0
35
71
16.
41
49
30
0
45
70
17.
47
42
43
0
38
75
18.
41
40
33
0
38
63
19.
41
50
49
0
40
78
20.
45
34
34
0
34
68
21.
47
45
38
0
40
68
22.
32
42
35
0
32
74
23.
46
65
49
0
49
74
24.
38
65
31
0
32
63
25.
45
59
34
0
31
60
26.
33
75
37
0
31
58
27.
35
66
30
0
47
74
28.
37
68
31
0
36
63
29.
34
64
50
0
40
77
30.
33
75
30
0
39
78
It
is
im
po
rtant
to
no
te
t
hat
th
e
pro
posed
al
gorithm
do
es
not
cha
nge
the
com
plexity
of
the
WSN
as
there
is
no
any
add
it
io
n
to
th
e
hardw
a
re
sid
e.
In
a
dd
it
io
n,
in
fau
lt
ap
peara
nce
cases,
the
so
luti
o
n
is
pro
du
c
e
d
un
ti
l
they
a
re
fixe
d.
T
he
refo
re,
t
he
pr
opose
d
al
go
rithm
consum
ed
a
s
ensible
ti
m
e
t
o
re
co
ver
the
error
s
,
in which
the
re
al
-
tim
e con
cep
ts are sti
ll
sati
sfied.
5.
CONCL
US
I
O
N
This
pa
per
pro
po
s
ed
a
s
of
tw
are
en
gin
ee
rin
g
base
d
fa
ult
tolera
nce
m
et
h
od
for
fa
ult
de
te
ct
ion
an
d
introd
ucin
g
a
s
olu
ti
on
in
WSN.
The
pro
pos
ed
m
et
ho
d
ad
opte
d
t
he
sel
f
-
c
heck
i
ng
pr
oces
s
f
or
fa
ult
dete
ct
ion
.
In
ad
diti
on,
th
e
pr
op
os
e
d
al
gorithm
con
side
red
the
neig
hbor
se
ns
or
no
de
s
to
est
i
m
a
te
t
he
fa
ulted
readi
ng
s
in
eff
ic
ie
nt
way.
The
av
era
ge
va
lue
of
t
he
cl
ose
neig
hbors
was
ad
opte
d
i
n
the
propose
d
al
go
rit
hm
to
increase
the
acc
ur
acy
of
est
i
m
at
ion
as
well
as
the
reli
abili
ty
.
In
ca
se
of
fa
ult
occur
in
the
cl
os
e
nei
ghbor
no
de,
t
he
ne
xt
valid
on
e
wa
s
consi
der
e
d.
It
is
i
m
po
rtant
t
o
not
e
that
the
rob
us
t
res
ponse
of
the
pro
po
sed
m
et
ho
d
al
lowe
d
the
us
e
i
n
rea
l
-
tim
e
syst
e
m
with
hi
gh
act
i
vity
of
fa
ult
tolera
nce.
T
he
pro
po
se
d
m
et
ho
d
was
te
ste
d
ov
e
r
diff
e
re
nt
case
stud
ie
s
at
a
n
a
ssu
m
ed
si
m
ulati
on
m
od
el
with
di
ff
e
ren
t
cl
ust
ers
in
WSN.
The
ob
ta
ine
d
r
esults
sh
owe
d
t
he pr
oved
p
e
rfo
rm
ance of th
e
prop
ose
d
m
et
ho
d i
n
t
erm
s o
f
reli
abil
it
y and
acc
ur
ac
y.
REFERE
NCE
S
[1]
Hos
sam
Mahm
o
ud
Ahm
ad,
“
W
i
rel
ess
Sensor
Networks
Conce
p
ts,
Applicati
ons,
Expe
riment
at
io
n
and
Anal
y
s
is
,”
Springer
,
2016
.
[2]
Ian
Som
m
erv
il
le,
“
S
oftwa
re engineer
ing
,”
10
th
Ed
it
ion, Pears
on
E
ducat
ion
,
In
c
,
20
17.
[3]
Miikka
Kuut
il
a
a
,
Mika
Mant
y
l
a
a,
Um
ar
Farooq
a,
and
Ma
el
i
ck
Cla
es,
“
Ti
m
e
Pr
essure
in
Softw
are
Engi
n
ee
ring
:
A S
y
stem
atic
R
e
vie
w
,”
Esl
iv
er
,
2
020.
[4]
V.
V.
Phoha
and
Shashi
Phoha,
“
Situa
ti
on
-
Aw
are
Software
Eng
ine
er
ing
for
Sensor
Networks
,”
2nd
Inte
rnation
al
Confe
renc
e
on
C
omm
unic
ati
on
S
yste
ms
Soft
ware
and
Middleware
,
2007
.
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
Self
-
checki
ng
meth
od for
fa
ul
t t
oleran
ce
s
olu
ti
on in
wi
rel
e
ss sensor
n
et
w
or
k
(
M
ua
y
ad
Sadik Cr
oock
)
4425
[5]
A
y
uba
J,
S
afwa
na
H,
and
Abdu
la
z
ee
z
Y,
“
Software
Deve
lopment
of
Int
egr
a
ted
W
ire
le
ss
Sensor
Networks
For
Rea
l
Ti
m
e
Monit
oring
of
Oil
and
Gas
Flow
Rat
e
Mete
ring
Infr
astruc
tur
e
,”
Journ
al
of
Informatio
n
Technol
ogy
&
Soft
ware
Engi
ne
ering
,
vo
l. 8, no. 2, p
p.
1
-
10,
201
8.
[6]
Savit
ha
S,
S
C
Li
nga
Redd
y
,
Sanjay
Chit
n
is,
“
Ene
rg
y
Eff
i
ci
en
t
Cluste
ring
and
Routi
ng
Optimi
za
t
ion
Model
fo
r
Maximizi
ng
Li
f
et
ime
of
W
ire
l
ess
Sensor
Network
,”
In
te
rnationa
l
Journal
o
f
E
lect
rical
and
Computer
Eng
ine
erin
g
(
IJE
CE)
,
vol. 10
,
no
.
5
,
2020
.
[7]
Sridhar
R.
,
and
N.
Guruprasad,
“
Ene
rg
y
eff
ici
ent
cha
o
ti
c
wha
le
opti
m
izati
on
te
chn
ique
for
dat
a
ga
the
ring
i
n
wire
le
ss
sensor
net
work
,”
Inte
rn
ati
onal
Journal
of
Elec
tri
cal
an
d
Computer
Eng
ine
ering
(
IJE
C
E
)
,
vol.
10,
no.
4
,
pp.
4176
-
4188
,
2020.
[8]
Abdus
y
S
y
ari
f,
Abdelha
fid
Abouaissa,
Lha
ss
a
ne
Idoum
ghar
,
Achm
ad
Kodar,
Pasca
l
Lore
n
z,
“
Sear
ch
Base
d
Software
Engi
n
e
eri
ng
on
Evol
u
tionar
y
Mul
ti
-
Obj
ec
t
ive
Approac
h
,”
I
E
EE
-
ICC
N
ex
t
-
Gen
erati
on
Net
working
and
Inte
rnet
Symposi
um
,
2016.
[9]
Aadit
i
y
a
V
enkat
Kanna
n,
“
Har
dware
and
Soft
ware
Archi
t
ectu
re
of
W
ire
le
ss
Sensor
Networks
,”
Journal
of
Adv
anc
es
in
Co
mputer
N
et
work
s
,
vol. 2, no. 3, p
p.
207
-
210
,
201
4.
[10]
Emm
anue
l
Eff
a
h,
Ous
m
ane
Thi
are
,
“
Surve
y
:
Fa
ult
s,
Fault
Det
ection
and
Fault
T
ole
ran
ce
T
ec
hni
ques
in
W
ire
le
ss
Sensor Net
works
,”
Int
ernati
onal
Journal
o
f
Com
pute
r Sc
ie
nc
e
an
d
Information
S
e
curit
y
,
2018
.
[11]
Khadidj
a
E
l
Mer
rao
ui,
Abde
ll
a
ziz
Ferdjouni
,
M’h
amed
Bounekhl
a
,
“
Rea
l
ti
m
e
obs
erv
er
-
base
d
st
at
o
r
fau
lt
di
agnosis
for
IM
,”
Int
ernati
onal
Journal
of
El
e
ct
rica
l
and
C
omputer
Engi
n
e
ering
(
IJE
CE)
,
v
ol.
10
,
no
.
1
,
pp
.
210
-
222
,
2020
.
[12]
Sus
hruta
Mishra,
La
m
bodar
Jen
a,
Aart
i
Pradha
n,
“
Fault
Tolera
nce
in
W
ireless
Sensor
Networks
,”
Inte
rnat
ion
al
Journal
of
Ad
va
nce
d
Re
search
i
n
Computer
Sc
ience
and
Sof
tware
Eng
ine
ering
,
v
ol.
2
,
no
.
10
,
p
p
.
146
-
153
,
2014
.
[13]
Mana
s
vi
Manna
n,
and
Shashi
B
.
Ran
a,
“
Fault
T
ole
ran
ce
in
W
ir
el
ess
Sensor
Ne
twork
,”
In
te
rnat
ional
Journal
of
Curr
ent
Engi
n
eer
ing
and
Techno
logy
,
vol
.
5
,
no
.
3,
p
p
.
1785
-
178
8,
2015
.
[14]
A.V.
Sutagunda
r
,
Vid
y
a
S.
Bennur,
Anus
ha
A.
M.
Bhanu
K.
N.,
“
Agent
Based
F
aul
t
To
l
era
n
ce
i
n
W
ire
le
ss
Sens
or
Networks
,”
In
te
r
nati
onal
Con
fe
re
nce
on
Inv
ent
i
ve
Computati
on
Te
chnol
ogi
es
(
ICICT)
,
2016.
[15]
H.S.Annapurna
,
and
M.Sidda
pp
a
,
“
Secur
e
Da
ta
Aggrega
ti
on
wit
h
Fau
lt
Tol
er
ance
for
W
ireless
Sensor
Networks
,”
Inte
rnational
Co
nfe
renc
e
on
Eme
rging R
ese
arch
i
n
Elec
troni
cs,
C
omputer
Scienc
e
and
Techno
logy
,
2015
.
[16]
Ta
oChen
a
,
Miqi
ngLi
,
and
XinYaoc
b
,
“
Standi
ng
on
the
shoulder
s
of
gia
nts:
See
ding
s
ea
rch
-
b
ase
d
m
ult
i
-
objectiv
e
opti
m
iz
ation
wi
th
prior
knowle
dge
for
software
service
compos
it
ion
,”
Journal
of
Informatio
n
and
Soft
wrae
Technol
ogy
,
vo
l. 114, pp
.
155
-
17
5,
2019
.
[17]
Zey
u
Zha
ng
,
Am
ja
d
Mehm
ood,
Le
i
Shu,
Zhi
q
i
ang
Huo,
Yu
Zha
ng
and
Mithu
n
Mukherje
e
,
“
A
Surve
y
on
Fault
Diagnosis
in
W
i
rel
ess Sensor Ne
tworks
,”
I
EE
E
A
cc
ess
,
vol. 8, p
p.
11349
-
11364,
2
018.
[18]
Mehdi
Naz
ari
Chera
ghlou
,
Ah
m
ad
Khade
m
-
Za
deh
and
Ma
ji
d
Haghpa
rast
,
“
A
New
Fault
-
T
ole
ran
t
Clust
ering
Protocol
for
W
i
rel
ess
Sensor
Networks
,”
In
te
rn
ati
onal
Journal
of
Sec
uri
ty
and
Its
Appl
i
cat
ions
,
vol.
12
,
no
.
4,
pp.
29
-
40
,
2018
.
[19]
M
y
eongse
ung
H
an,
Eun
-
Sung
C
ho,
Jun
-
h
y
uk
Lee,
and
Intae
R
yoo,
“
A
Stud
y
o
n
Fault
To
le
r
an
ce
T
ec
hn
ique
fo
r
AGR
-
MA
C
Pro
toc
ol
b
ase
d
Sen
sor
Networks
,”
Inte
rnational
Confe
renc
e
on
In
formation
Net
w
orking
(
ICOIN
)
,
pp.
457
-
460
,
20
19.
[20]
As
m
aa
Shaal
an
Abdul
Munem
,
and
Mua
y
ad
Sa
dik
Croock,
“
Sm
art
Tra
ffi
c
Li
g
ht
Control
S
y
st
e
m
for
Emerge
nc
y
Am
bula
nce
,”
Int
ernati
onal
Jour
nal
of
Adv
an
ce
d
Re
search
in
C
omputer
Engi
ne
ering
&
Technol
ogy
(
IJA
RCE
T)
,
vol.
5
,
no
.
1
,
201
6.
[21]
Feli
pe
T
al
i
ar
Giunti
ni,
De
la
no
Mede
iros
Bede
r
and
Jo´
Ue
y
ama,
“
Expl
oi
ti
ng
Sel
f
-
orga
nizati
on
a
nd
Fault
Tol
er
an
ce
in
W
ire
l
ess
Sensor
Networks:
A
Case
Stud
y
on
W
il
dfire
Dete
c
ti
on
Appli
ca
t
ion
,”
Inte
rna
ti
onal
Journal
of
Distribute
d
S
ensor Ne
tworks
,
vo
l
.
13
,
no
.
4
,
p
p
.
1
-
16,
2017
.
[22]
Ric
ard
o
Vi
llalón
and
Patr
ic
k
G
.
Bridge
s,
“
Fault
-
Tol
er
ant
W
irele
ss
Sensor
Networks
using
Evolutiona
r
y
G
ames
,”
PhD
The
sis,
The Unive
rsit
y
of
N
ew
Mexic
o
,
201
2.
[23]
Jenn
-
W
ei
Li
n
,
P
et
huru
Ra
j
C
.
,
Meng
-
Chie
h
H
.
,
a
nd
Jia
-
Xin
H
.
,
“
Eff
ic
i
ent
Fau
lt
-
Tol
er
ant
Rou
ti
n
g
in
IoT
W
ire
l
ess
Sensor Net
works
Based
on
Bipa
r
ti
te
-
Flow
Graph M
odel
ing
,”
I
EEE
A
ccess
,
vol
.
7
,
pp.
14022
-
1403
4,
2019
.
[24]
L.
Mottol
a
,
G.
P.
Picc
o
,
F.
J.
Opperm
ann
,
et
al
,
“
Sim
pli
f
y
ing
the
Inte
gr
at
ion
of
W
ire
le
ss
Sensor
Networks
in
to
Busin
ess Proce
ss
es
,”
I
EE
E
Tr
ansacti
on
on
Soft
w
are
Engi
n
ee
ring
,
vol. 45, no. 6, p
p.
576
-
596
,
201
9.
[25]
Farna
z
Pakd
el
1
,
Mansour
Esm
aeilpour
,
“
Fuzz
y
L
ogic
Method
for
Enha
n
ce
m
ent
F
aul
t
-
Tol
er
ant
of
Cluste
r
He
ad
in
W
ire
le
ss
Sensor
Networks Cl
uste
ring
,”
TEM
Journal
,
vo
l.
5,
no.
3,
p
p.
268
-
277,
20
16.
[26]
Evi
z
al
Abdul
Kadir
,
Hitoshi
Iri
e
,
Sri
Li
stia
Rosa
,
and
Mahm
od
Othm
an,
“
Mode
ll
ing
of
wire
l
ess
sensor
net
works
for
det
e
ct
ion
land
and
fore
st
f
ire
hotspot
,”
T
ELKOMNIKA
(Tele
communic
a
t
ion,
Computing
,
El
e
ct
ronics
an
d
Control)
,
vol
.
17
,
no
.
6
,
pp
.
2772
-
2781,
2019
.
[27]
Ngan
Ngu
y
en
,
Quoc
Cuong
Ngu
y
en
,
and
Min
h
Thu
y
Le,
“
A
novel
aut
o
nom
ous
wire
le
ss
sensor
node
for
IoT
appl
i
ca
t
ions
,”
T
ELKOMNIKA
(Tele
communic
a
t
ion,
Computing
,
El
e
ct
roni
cs
and
Control)
,
v
ol.
17,
no.
5
,
p
p.
2389
-
2399
,
2019.
Evaluation Warning : The document was created with Spire.PDF for Python.