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.
15
,
No.
1
,
Febr
uary
20
25
, pp.
425
~
434
IS
S
N:
20
88
-
8708
, DO
I: 10
.11
591/ij
ece.v
15
i
1
.
pp
425
-
434
425
Journ
al h
om
e
page
:
http:
//
ij
ece.i
aesc
or
e.c
om
Q
-
lea
rn
i
ng base
d forec
asting ea
rly landsli
de dete
ction in
internet
of
t
hing
wireless
sensor n
etwork
Devas
ahayam
Joseph
Jey
akumar
1
,
Bo
omi
na
t
han S
han
mathi
2
, P
arap
purathu
Ba
hu
layan
Smith
a
1
,
Sha
li
ni
Chowdary
3
,
Th
amiz
ha
r
as
an
P
ann
eerse
lva
m
1
,
R
aja
go
p
al
an
S
r
inat
h
4
, Mu
th
u
raj M
arisel
va
m
1
,
Mohana
n
Mur
ali
5
1
Dep
artm
en
t of
E
l
ectron
ics an
d
Co
m
m
u
n
icatio
n
E
n
g
in
eering
,
J.N.
N I
n
stit
u
te of E
n
g
in
eering
,
Ch
en
n
ai,
Ind
ia
2
Dep
artm
en
t of
E
l
ectron
ics an
d
Co
m
m
u
n
icatio
n
E
n
g
in
eering
,
Vela
m
m
al
I
n
stitu
te of T
echn
o
lo
g
y
,
Ch
en
n
ai
,
Ind
i
a
3
Dep
artm
en
t of
E
l
ectron
ics an
d
Co
m
m
u
n
icatio
n
E
n
g
in
eering
,
T.
J
.S
Eng
in
eering
C
o
lleg
e,
C
h
en
n
ai,
Ind
ia
4
Dep
artm
en
t of
E
l
ectron
ics an
d
Co
m
m
u
n
icatio
n
E
n
g
in
eering
,
SRM
I
n
stitu
te of Scienc
e and
Techn
o
lo
g
y
,
Ch
en
n
ai,
Ind
ia
5
Dep
artm
en
t of
Bi
o
m
ed
ical E
n
g
in
eer
in
g
,
J.N.N
Ins
titu
te of E
n
g
in
eering
,
C
h
en
n
ai,
Ind
ia
Art
i
cl
e In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
M
a
y 18,
2024
Re
vised
A
ug 19, 2
024
Accepte
d
Se
p 3, 2
024
The
issue
of
cl
i
ma
t
e
modi
fi
ca
t
io
n
and
hum
an
act
ions
termin
at
es
i
n
a
ch
ai
n
of
haz
ard
ous
d
evelopme
nts
,
co
m
pre
hensive
of
la
ndslide
s
.
Th
e
tra
d
it
ion
al
appr
oac
h
es
of
observing
the
envi
ronm
ent
a
l
at
tri
bu
te
s
that
is
a
ct
ua
lly
obta
ini
ng
r
ai
nf
all
d
at
a
fro
m
pl
aces
c
an
be
cru
el
a
nd
suppress
ing
supervising
nec
essit
at
ed
for
ca
r
efu
l
inf
li
c
tion.
Thus
,
l
ands
li
de
fore
ca
sting
and
ea
r
ly
noti
c
e
is
a
sign
i
fic
an
t
appl
i
catio
n
vi
a
wir
eless
sensor
ne
tworks
(W
SN
)
to
red
uce
loss
of
lif
e
and
prop
ert
y.
Bec
ause
of
the
h
ea
vy
p
rep
ar
ation
of
sensors
in
l
andsli
de
pros
tra
t
e
r
egi
ons,
cl
u
steri
ng
is
a
resou
rce
ful
m
et
hod
to
mi
n
im
i
ze
unnec
essary
trans
mi
ss
ion.
In
t
his
art
i
cl
e
we
i
ntroduc
e
Q
-
l
ea
r
ning
base
d
fore
ca
st
ing
ea
r
l
y
la
ndsl
ide
de
tecti
on
(Q
-
LFD)
in
in
te
rn
et
of
t
hing
s
(IoT
)
WSN.
Th
e
Q
-
LFD
mecha
nis
m
u
ti
l
iz
es
a
d
i
ngo
opt
im
i
za
t
io
n
a
lgori
th
m
(DO
A)
to
choose
the
b
est
cl
ust
e
r
h
e
ad
(CH).
Fu
rthe
rmor
e,
th
e
Q
-
le
arn
ing
al
gorit
h
m
fore
c
ast
the
la
ndsl
id
e
by
soil
wa
te
r
c
apa
c
it
y
,
soi
l
la
yer
,
soil
te
mp
era
tur
e,
Sei
smic
vibr
at
i
ons,
and
ra
infa
l
l.
Ex
per
imental
resul
ts
il
lustr
at
e
the
Q
-
LFD
m
echani
sm
rai
s
es
th
e
la
ndslid
e
de
te
c
ti
on
accuracy.
In
add
it
ion
,
i
t
mi
nimize
s
the false
positi
v
e, f
a
lse
neg
at
iv
e
r
at
io
.
Ke
yw
or
d
s
:
Cl
us
te
rin
g
Dingo o
ptimi
z
at
ion
L
an
ds
li
de fo
re
cast
ing
Q
-
le
ar
ning
Wireless
se
nso
r netw
ork
This
is an
open
acc
ess arti
cl
e
un
der
the
CC
BY
-
SA
l
ic
ense
.
Corres
pond
in
g
Aut
h
or
:
Dev
a
saha
yam
Jo
se
ph Jeya
ku
mar
Dep
a
rtme
nt of
Ele
ct
ro
nics
and
C
om
m
unic
at
ion
En
gin
ee
rin
g
,
J
.N.N
Insti
tu
te
o
f
Engine
eri
ng
Chen
nai,
India
Emai
l:
jayaku
marjose
ph33
@
gm
ai
l.co
m
1.
INTROD
U
CTION
The
mo
ti
on of so
il
,
r
oc
k,
and o
the
r
s
ubsta
nc
e
dow
nwar
d
a
slop
i
ng
re
gion o
f
la
nd
is ment
ion
e
d
as
the
la
nd
sli
de
.
Hea
vy
Ra
in,
quak
e
s,
volc
a
no
es
,
a
nd
ad
diti
on
al
na
tural
a
nd
synt
hetic
proces
ses
that
pro
vi
de
a
slop
e
unbalance
d
mi
gh
t
al
l
the
act
ivate
s
for
la
ndsli
des
[1]
.
T
he
i
nt
ern
et
of
thi
ngs
(
Io
T
)
act
s
a
si
gn
i
ficant
funct
ion
in
decidin
g
t
he
la
nd
sli
de
issue
s
[2],
[3
]
.
It
pres
ents
a
se
ver
e
t
hr
eat
to
huma
ns
an
d
the
w
or
l
d's
se
ver
al
cl
as
ses
of
pove
rty
a
nd
su
r
face
s
urr
ound
i
ngs.
La
nds
li
des
pri
ncipal
ly
ha
ppen
be
cause
of
cl
imat
e
change
i
n
the
su
r
rou
nd
i
ngs
[
4]
.
The
re
are
s
ever
al
cau
ses
f
or
la
unc
hing
s
uch
a
s
s
urr
oundin
gs
f
or
disas
te
r
reducti
on,
i
n
t
hat
the
se
nsor
noti
ces
the
el
eme
nt
an
d
i
ns
ta
ntly
f
orwards
the
data,
a
nd
sp
eci
al
machi
ne
le
a
rn
i
ng
al
gorith
ms
a
re
app
li
ed
to
offe
r
data
to
the
pe
op
le
ab
out
the
disas
te
r.
It
is
a
n
i
nexpe
ns
ive
and
eas
y
t
o
i
nst
al
l,
and
it
mi
ght
be
employe
d via a
semi
-
s
kill
ed
i
nd
i
vidual
[
5]
.
Lan
ds
li
des
are
an
im
portance
dete
rmin
e
d
m
ov
e
ment
of
a
bu
l
k
of
ro
c
k,
s
oi
l
an
d
dust
do
wn
a
slo
pe,
and
the
y
ca
n
ori
gin
imp
or
ta
nt
huma
n
death
and
eco
nomica
l
losin
g.
T
he
e
nh
a
ncin
g
num
ber
of
us
ual
t
ra
ged
ie
s
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
omp E
ng,
V
ol.
15
, No
.
1
,
Febr
uary
20
25
:
425
-
434
426
because
of
cl
imat
e
cha
nge
is
of
vital
an
xiety
[6]
.
La
ndsli
de
s
are
on
e
of
the
pr
e
domi
nant
ge
ologic
haza
rd
s
tha
t
resu
lt
in
massi
ve
huma
n
an
d
economic
los
s
es.
The
descr
i
bed
a
nalysis
ha
s
been
e
xec
uted
in
se
ver
al
na
ti
on
s
thr
oughout t
he
world
as ill
us
tr
at
ed
in
Fig
ur
e
1.
Figure
1. Lan
dsl
ide
detect
io
n amo
ng d
i
ff
e
rent
co
untrie
s
Th
us
,
be
f
or
e
oc
currin
g
t
he
la
nd
sli
de,
la
ndsli
de
forecast
in
g
is
a
sig
nifica
n
t
issue
in
wi
re
le
ss
se
ns
or
netw
orks
(
WSN)
.
Tra
diti
on
al
mecha
nisms
-
ba
sed
la
ndsli
des
pr
e
dicti
on
a
na
lysis
not
ade
quat
e
f
or
f
oreca
sti
ng
exact
posit
ion
and
ti
me
f
or
bu
l
k
mo
ti
ons
[7]
.
N
owada
ys
,
the
quic
k
grow
t
h
of
ma
chine
le
ar
ning
(M
L
)
al
gorithm
diss
eminat
e
t
hro
ugho
ut
se
ve
ral
in
vestigat
ors
ha
ve
init
ia
te
d
searc
hing
into
discipli
nary
[
8]
.
Ar
ti
fici
al
intel
li
gen
ce
(AI)
,
a
nd
rem
ote
se
nsi
ng
t
o
detect
i
ng
la
ndsli
de
ha
zard
e
valuati
on.
I
oT
with
M
L
to
ob
s
er
ve
t
he
ge
ologica
l
la
ndsl
ide
act
io
ns
to
protect
the
c
re
at
ur
es
f
rom
t
he
la
ndsli
des
[
9]
.
T
he
i
ncorpo
ra
ti
on
of
a
dee
p
le
ar
ning
m
odel
with
r
ule
-
based
obje
ct
-
base
d
ima
ge
analysis
(
OBI
A)
t
o
noti
ce
la
nd
sli
des.
T
he
va
lue
of
every
pix
el
in
the
heatma
p
i
ndic
at
es
to
the
po
s
sibil
it
y
that
the
pi
xel
bel
ongs
to
more
over
la
ndsli
de
or
non
-
la
nd
sli
de mo
du
le
. H
owe
ver
, it
can
no
t abl
e to
forecast
b
e
fore
occurri
ng
[10
]
. To
s
olv
e t
hes
e issues, Q
-
le
a
rn
i
ng
base
d
f
or
eca
sti
ng
early
la
ndsl
ide
detect
io
n
i
n
I
oT
WSN.
S
ever
al
la
nd
sli
de
te
chn
i
qu
e
s
ca
pab
le
of
detect
ing
t
he
la
nd
sli
de
e
ff
ic
i
ently;
th
ough,
it
do
es
no
t
for
ecast
ing
the
la
nd
sli
de.
T
he
si
gn
i
ficant
ai
m
of
t
his
arti
cl
e
is
to
introd
uce
a
Q
-
le
arn
i
ng
al
go
r
it
hm
to
recog
nize
la
ndsli
des
well
.
Th
e
se
ct
ion
2
ex
plains
Q
-
le
arn
i
ng
base
d
forecast
in
g
ear
ly
la
ndsli
de
detect
ion
in IoT
wireless
se
nsor
netw
ork
.
Sect
ion
3
d
esc
ribes
exp
e
rime
ntal
r
esults
.
F
inall
y,
sect
io
n 4 prese
nts c
oncl
us
i
on and
f
uture
work.
Lan
ds
li
des
pos
e
a
rec
urrin
g
threat
in
the
Himala
yan
re
gion,
le
a
d
in
g
to
de
vastat
ing
c
on
seq
uen
ces
in
te
rms
of
huma
n
casualt
ie
s
a
nd
pro
per
t
y
da
mage.
A
gr
ound
br
ea
king
mecha
nism
is
us
e
d
t
o
obse
rvi
ng
a
nd
forecast
in
g
la
ndsli
des.
T
his
s
ys
te
m
ap
plies
sens
or
no
des
to
i
ncessa
ntly
obser
ve
a
s
urr
oundin
g
sit
uatio
n
a
nd
colle
ct
pe
rtin
e
nt
data.
This
m
echan
is
m
util
iz
es
a
sup
port
ve
ct
or
mac
hin
e
(
SVM),
k
-
nea
re
st
nei
ghbor
(K
-
N
N),
and
decisi
on
tr
ee
(D
T
),
al
gorithms
meas
ured
the
gaine
d
data
[11]
.
T
he
la
ndsli
des
pr
e
dicti
on
s
ys
te
m
util
iz
in
g
the
ra
dio
fr
e
qu
ency
(
RF
)
am
pl
ifie
r
an
d
recti
f
ie
r
desi
gn
e
d
a
nd
in
ve
nted.
Ne
xt,
a
n
bo
unda
r
y
am
ong
t
he
s
ens
or
sign
al
outp
ut
as
well
as
cl
oud
c
omp
utin
g
us
in
g
l
ong
range
(
L
oRa
)
pic
ked
out
a
s
a
Io
T
e
nd
dev
ic
e
corres
po
ndin
gly
[
12]
.
La
ndsli
de
disp
la
ce
ment
predict
arti
fici
al
int
el
li
gen
ce
al
gorithms
rai
nf
al
l
an
d
disp
la
ceme
nt
da
ta
purpos
e
to
al
low
a
reli
abl
e
early
wa
rn
i
ng
wh
ic
h
ca
n
f
oreca
st
la
ndsli
de
disaste
r
a
s
w
el
l
as
pro
vid
e earl
y war
ning
[
13]
. A
G
a
us
sia
n
pr
ocess
to
foreca
st t
he
la
nd
sli
de
in
WSN. T
his
mecha
nism m
i
nimize
s
the
miss
with
wrong
al
ar
m
rati
o
of
la
ndsli
de
[14]
.
T
he
grow
t
h
of
WSN
is
a
cos
t
-
eff
ic
ie
nt
s
olut
ion
f
or
ob
s
er
ving
par
t
ly
sta
ble
slo
pe
s
to
f
or
ecast
la
ndsli
des
i
n
P
un
e
a
nd
M
aha
rash
t
ra.
A
n
e
ne
rgy
e
ff
ic
ie
nt
WSN
app
l
ying
lo
w
powe
r
se
nsor
node
s
to
obser
ve
the
m
oistu
re
of
so
il
,
ru
s
hi
ng
,
a
ngle
of
ti
lt
and
rain
fall
int
ensity
[15]
.
Lan
ds
li
de
s
are
re
peati
ng
i
ncessa
ntly
wh
e
n
cau
sin
g
direct
e
ff
ect
on
human
li
f
e.
The
maj
or
reason
fo
ll
owin
g
rais
e
in
la
ndsli
de
ha
ppeni
ng
is
cras
h
of
c
ha
ng
e
in
cl
imat
e
as
well
as
raisi
ng
human
act
ion.
Lan
ds
li
de
ob
se
rv
i
ng
s
ys
te
m
ut
il
iz
es
a
sever
a
l
typ
es
of
se
nsors
t
hat
can
be
app
li
e
d
to
un
i
nterru
pted
la
ndsli
de
risk
obser
ving
hazar
d
[16
]
.
D
eep
le
ar
ning
al
gori
thm
is
util
iz
ed
to
av
oid
la
nd
sli
des.
T
his
mecha
nism
a
ppli
es
a
long
s
hort
-
te
r
m
me
mor
y
(
L
STM
)
t
o
re
vea
l
IoT
-
base
d
usual
hazar
d
obs
erv
i
ng
a
nd
f
oreca
sti
ng
anal
ytics
in
t
he
hill
s
[17]
.
T
he
enhance
ment
of
forecast
in
g
la
nd
sli
des
an
d
early
al
ert
by
app
l
ying
M
L
wi
th
I
oT
te
c
hniqu
es
us
in
g feat
ures l
ike m
oisture
of so
il
, soil
of
fleece
strength
, ra
in se
ver
it
y,
a
nd
te
rr
ai
n sl
ope
[
18]
.
W
i
r
e
l
e
s
s
s
e
n
s
o
r
n
o
d
e
s
a
r
e
p
r
o
v
i
d
e
d
a
h
i
g
h
s
e
n
s
i
t
i
v
i
t
y
w
i
t
h
hu
g
e
r
e
p
o
r
t
i
n
g
r
e
g
i
o
n
.
T
h
e
s
e
ns
o
r
e
l
e
m
e
n
t
d
i
s
c
o
v
e
r
s
t
h
e
s
h
a
k
i
n
g
o
f
n
a
t
u
r
a
l
d
i
s
a
s
t
r
o
u
s
f
r
o
m
e
a
r
t
h
a
f
t
e
r
t
h
a
t
i
t
w
i
l
l
f
o
r
w
a
r
d
d
a
t
a
t
o
o
b
s
e
r
v
i
n
g
l
o
c
a
t
i
o
n
t
h
r
o
u
g
h
r
a
d
i
o
f
r
e
q
u
e
n
c
y
i
d
e
n
t
i
f
i
c
a
t
i
o
n
(
R
F
I
D
)
.
M
o
i
s
t
u
r
e
,
w
e
i
g
h
t
a
s
w
e
l
l
a
s
t
i
l
t
m
e
t
e
r
i
s
a
p
pl
i
e
d
t
o
o
b
s
e
r
v
e
s
g
l
i
b
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
omp E
ng
IS
S
N:
20
88
-
8708
Q
-
le
arnin
g based f
or
eca
sti
ng ea
rly
l
ands
li
de
d
et
ect
io
n
in
…
(
Dev
asa
ha
y
am Jo
s
ep
h
Jey
akum
ar
)
427
r
e
p
l
y
f
r
o
m
t
h
e
m
o
d
i
f
y
i
n
g
g
e
o
-
t
e
c
h
n
i
c
a
l
s
i
t
u
a
t
i
o
n
.
T
h
u
s
,
d
e
t
e
c
t
s
t
h
e
l
a
n
d
s
l
i
de
w
e
l
l
[19]
.
T
h
e
l
a
n
d
s
l
i
d
e
s
t
r
a
i
n
d
a
t
a
a
n
d
s
u
b
s
e
q
u
e
n
t
d
i
s
t
r
i
b
u
t
e
d
de
c
i
s
i
o
n
a
l
g
o
r
i
t
h
m
i
m
i
t
a
t
e
p
r
e
s
s
u
r
e
v
a
r
i
a
t
i
o
n
o
n
r
o
c
k
t
r
i
e
s
o
u
t
a
n
d
c
a
l
c
u
l
a
t
e
t
he
e
q
u
i
v
a
l
e
n
t
s
t
r
a
i
n
.
T
h
e
v
a
r
i
a
b
l
e
m
e
a
n
G
a
u
s
s
i
a
n
p
r
o
c
e
s
s
i
s
m
o
d
e
l
e
d
t
h
e
s
t
r
a
i
n
d
a
t
a
.
T
h
e
d
i
s
t
r
i
b
u
t
e
d
s
c
a
l
a
r
b
a
s
e
d
r
e
c
o
g
n
i
t
i
o
n
o
f
f
e
r
s
a
s
b
e
t
t
e
r
o
u
t
c
o
m
e
s
r
e
g
a
r
d
i
n
g
f
e
a
s
i
b
i
l
i
t
y
of
m
i
s
s
e
d
d
e
t
e
c
t
i
o
n
,
s
m
a
l
l
e
s
t
e
n
e
r
g
y
u
t
i
l
i
z
a
t
i
o
n
a
n
d
f
a
l
s
e
a
l
a
r
m
[
2
0
]
.
L
a
n
d
s
l
i
d
e
p
r
e
d
i
c
t
i
o
n
a
n
d
e
a
r
l
y
w
a
r
n
i
n
g
s
y
s
t
e
m
i
s
a
n
i
m
p
o
r
t
a
n
t
a
p
p
l
i
c
a
t
i
o
n
w
h
e
r
e
s
e
n
s
o
r
n
e
t
w
o
r
k
s
c
a
n
b
e
d
e
p
l
o
y
e
d
t
o
m
i
n
i
m
i
z
e
l
o
s
s
o
f
l
i
f
e
a
n
d
p
r
o
p
e
r
t
y
.
D
u
e
t
o
t
h
e
d
e
n
s
e
d
e
p
l
oy
m
e
n
t
o
f
s
e
n
s
or
s
i
n
l
a
n
d
s
l
i
d
e
p
r
o
n
e
a
r
e
a
s
,
c
l
u
s
t
e
r
i
ng
i
s
a
n
e
f
f
i
c
i
e
n
t
a
p
p
r
o
a
c
h
t
o
r
e
d
u
c
e
r
e
d
u
n
d
a
n
t
c
o
m
m
u
n
i
c
a
t
i
o
n
f
r
o
m
c
o
-
l
o
c
a
t
e
d
s
e
n
s
o
r
s
.
I
n
t
h
i
s
p
a
p
e
r
w
e
p
r
o
p
o
s
e
t
w
o
d
i
s
t
r
i
bu
t
e
d
c
l
u
s
t
e
r
i
n
g
a
n
d
m
u
l
t
i
-
h
o
p
r
o
u
t
i
n
g
p
r
o
t
o
c
o
l
s
,
c
l
u
s
t
e
r
a
i
de
d
m
u
l
t
i
p
a
t
h
r
o
u
t
i
n
g
(
C
A
M
P
)
a
n
d
h
y
b
r
i
d
b
e
a
c
o
n
v
e
c
t
o
r
r
o
u
t
i
n
g
(
H
B
V
R
)
,
f
o
r
t
h
i
s
p
r
o
b
l
e
m
.
W
h
i
l
e
C
A
M
P
i
s
a
n
e
w
c
l
u
s
t
e
r
i
ng
a
n
d
r
o
u
t
i
n
g
p
r
o
t
o
c
o
l
,
H
B
V
R
i
s
a
n
e
n
h
a
n
c
e
m
e
n
t
o
f
B
V
R
w
i
t
h
h
y
b
r
i
d
e
n
e
r
g
y
-
e
f
f
i
c
i
e
n
t
d
i
s
t
r
i
b
u
t
e
d
(
H
E
E
D
)
.
W
e
f
u
r
t
h
e
r
e
n
h
a
n
c
e
C
A
M
P
a
n
d
H
B
V
R
w
i
t
h
t
h
r
e
s
h
o
l
d
s
e
n
s
i
t
i
v
e
e
n
e
r
g
y
e
f
f
i
c
i
e
n
t
s
e
n
s
o
r
n
e
t
w
o
r
k
(
T
E
E
N
)
,
a
t
h
r
e
s
h
o
l
d
-
b
a
s
e
d
e
v
e
n
t
d
r
i
v
e
n
p
r
o
t
o
c
o
l
.
T
E
E
N
i
s
m
o
s
t
s
u
i
t
a
b
l
e
p
r
o
t
o
c
o
l
f
o
r
t
h
i
s
a
p
p
l
i
c
a
t
i
o
n
s
i
n
c
e
d
i
f
f
e
r
e
n
t
r
o
c
k
t
y
p
e
s
c
a
n
h
a
v
e
d
i
f
f
e
r
e
n
t
t
h
r
e
s
h
o
l
d
s
f
o
r
s
t
r
e
s
s
v
a
l
u
e
s
.
S
i
m
u
l
a
t
i
o
n
r
e
s
u
l
t
s
s
h
o
w
t
h
a
t
C
A
M
P
-
T
E
E
N
g
i
v
e
s
t
h
e
b
e
s
t
p
e
r
f
o
r
m
a
n
c
e
w
i
t
h
r
e
s
p
e
c
t
t
o
n
e
t
w
o
r
k
l
i
f
e
t
i
m
e
a
n
d
e
n
e
r
g
y
c
o
n
s
u
m
p
t
i
o
n
[15]
.
T
h
e
o
b
s
e
r
v
i
n
g
a
n
d
f
o
r
e
c
a
s
t
i
n
g
e
v
e
n
t
s
f
o
r
d
e
t
e
c
t
i
n
g
L
a
n
d
s
l
i
d
e
s
.
A
n
o
p
t
i
m
a
l
l
o
c
a
t
i
o
n
f
o
r
s
i
n
k
a
s
s
i
g
n
m
e
n
t
t
o
e
n
h
a
n
c
e
t
h
e
f
e
a
t
u
r
e
s
f
o
r
e
x
a
m
p
l
e
d
e
l
a
y
,
o
b
t
a
i
n
e
d
r
a
t
i
o
,
a
n
d
j
i
t
t
e
r
[21]
.
A
m
a
g
n
e
t
i
c
i
n
d
u
c
t
i
o
n
e
s
t
a
b
l
i
s
h
e
d
l
o
c
a
l
i
z
a
t
i
o
n
t
h
a
t
p
r
e
c
i
s
e
l
y
a
n
d
p
r
o
f
i
c
i
e
n
t
l
y
p
o
s
i
t
i
o
n
e
d
a
r
b
i
t
r
a
r
i
l
y
d
i
s
t
r
i
b
u
t
e
d
s
e
n
s
o
r
s
b
y
l
e
v
e
r
a
g
e
t
h
e
m
u
l
t
i
p
a
t
h
f
a
d
i
n
g
.
A
t
w
o
-
s
t
e
p
p
o
s
i
t
i
o
n
i
n
g
m
e
t
h
o
d
f
o
r
g
e
t
t
i
n
g
r
a
p
i
d
a
n
d
p
r
e
c
i
s
e
l
o
c
a
l
i
z
a
t
i
o
n
o
u
t
c
o
m
e
s
:
p
r
i
m
a
r
y
,
f
o
r
m
u
l
a
t
i
n
g
t
h
e
f
a
s
t
-
i
ni
t
i
a
l
l
o
c
a
t
i
n
g
d
u
r
i
n
g
a
n
t
a
c
k
i
n
g
d
i
r
e
c
t
i
o
n
i
n
c
r
e
a
s
e
d
L
a
g
r
a
n
g
i
a
n
p
r
o
c
e
s
s
f
o
r
u
n
e
v
e
n
s
e
n
s
o
r
p
o
s
i
t
i
o
n
s
i
n
s
i
d
e
a
s
m
a
l
l
p
r
o
c
e
s
s
i
n
g
t
i
m
e
,
a
n
d
n
e
x
t
s
u
g
g
e
s
t
i
n
g
f
i
n
e
-
g
r
a
i
n
e
d
p
o
s
i
t
i
o
n
i
n
g
f
o
r
e
x
e
c
u
t
i
n
g
g
r
e
a
t
e
x
p
l
o
r
a
t
i
o
n
f
o
r
o
p
t
i
m
a
l
p
o
s
i
t
i
o
n
e
v
a
l
u
a
t
i
on
t
h
r
o
u
g
h
t
h
e
c
o
u
p
l
e
d
g
r
a
d
i
e
n
t
a
l
g
o
r
i
t
h
m
[
2
2
]
.
L
a
n
d
s
l
i
d
e
s
c
a
n
b
e
d
i
s
c
o
v
e
r
e
d
b
y
s
t
u
d
y
i
n
g
t
h
e
e
n
v
i
r
o
n
m
e
n
t
d
a
t
a
t
h
r
o
u
g
h
s
e
n
s
o
r
n
o
d
e
s
.
H
e
n
c
e
,
f
o
r
e
c
a
s
t
e
d
t
h
e
l
a
n
d
s
l
i
d
e
s
.
A
r
t
i
f
i
c
i
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
s
(
A
N
N
)
p
r
o
v
i
d
e
p
r
e
c
i
s
e
f
o
r
e
c
a
s
t
i
n
g
a
n
d
r
e
v
e
a
l
h
i
g
h
l
e
a
r
n
i
n
g
c
a
p
a
b
i
l
i
t
y
.
C
o
n
v
e
r
s
e
l
y
,
t
h
e
A
N
N
a
l
g
o
r
i
t
h
m
t
o
u
n
b
a
l
a
n
c
e
d
d
a
t
a
a
l
l
o
c
a
t
i
o
n
.
A
s
w
i
t
c
h
i
n
g
p
o
l
i
c
y
w
h
i
c
h
c
a
n
d
e
c
i
d
e
b
e
t
w
e
e
n
d
i
s
s
i
m
i
l
a
r
p
r
e
d
i
c
t
o
r
s
a
l
o
n
g
w
i
t
h
e
c
o
l
o
g
i
c
a
l
s
t
a
t
e
s
[
2
3
]
.
T
h
e
f
a
l
l
d
e
t
e
c
t
i
o
n
m
e
t
h
o
d
t
h
a
t
u
t
i
l
i
z
e
s
a
n
I
o
T
h
a
v
e
s
e
v
e
r
a
l
r
e
s
t
r
i
c
t
i
o
n
s
;
f
or
e
x
a
m
p
l
e
,
l
a
t
e
n
c
y
,
e
n
e
r
g
y
u
t
i
l
i
z
a
t
i
o
n
,
a
n
d
l
e
s
s
e
r
f
u
n
c
t
i
o
n
.
T
h
i
s
m
e
t
h
o
d
d
e
t
e
c
t
i
n
g
f
a
l
l
s
a
p
p
l
y
i
n
g
a
w
e
a
r
a
b
l
e
a
c
c
e
l
e
r
o
m
e
t
e
r
[24]
.
A
f
l
e
x
i
b
l
e
a
n
d
p
r
o
f
i
c
i
e
n
t
f
o
r
d
i
s
c
o
v
e
r
i
n
g
r
a
i
n
f
a
l
l
-
c
a
u
s
e
d
l
a
n
d
s
l
i
d
e
s
.
T
he
W
S
N
-
a
l
t
e
r
e
d
a
r
c
h
i
t
e
c
t
u
r
e
f
o
r
r
a
i
n
f
a
l
l
o
b
s
e
r
v
i
n
g
m
e
t
h
o
d
t
o
b
r
o
a
d
c
a
s
t
a
n
d
g
a
t
h
e
r
r
e
a
l
t
i
m
e
d
a
t
a
u
t
i
l
i
z
i
n
g
g
e
n
e
r
a
l
p
o
c
k
e
t
r
a
d
i
o
s
e
r
v
i
c
e
(
G
P
R
S
)
.
T
h
e
S
V
M
a
l
g
o
r
i
t
h
m
t
o
f
o
r
e
c
a
s
t
i
n
g
t
h
e
r
a
i
n
f
a
l
l
;
t
h
u
s
,
i
t
d
e
t
e
c
t
s
t
h
e
l
a
n
d
s
l
i
d
e
[25]
.
2.
METHO
D
In
W
SN,
se
ve
ral
se
ns
ors
tog
et
her
oc
casi
on
al
ly
ob
se
r
ve
the
surr
oundin
g
sit
uati
on
an
d
gath
e
r
correla
te
d
deta
il
s,
ne
xt,
it
f
or
wards
it
t
o
a
ba
se
sta
ti
on
(
BS
)
.
T
he
n
t
he
BS
eval
uates
a
la
nd
sli
de
est
a
blishe
d
on
the
gathe
red
da
ta
an
d
f
orwards
a
n
al
arm
f
or
a
feasi
ble
l
andsl
ide
t
o
a
cl
oud
ser
ver
on
t
he
off
cha
nc
e
the
la
n
ds
li
de
ou
t
pe
rforms
a
pre
set
th
reshold
for
e
arly
la
ndsli
de
a
vo
i
dan
c
e.
F
oreca
sti
ng
la
ndsli
de
is
a
vital
com
pone
nt
of
functi
onal
earl
y
no
ti
ce
s
ys
te
ms
with
Io
T
te
chnolo
gy.
Th
us,
it
is
a
pp
li
ed
to
e
xp
a
nd
a
la
nd
sli
de
forecast
in
g
m
odel
.
WSN
c
onta
ins
numb
e
r
of
sel
f
-
go
vernin
g
m
obil
e
sen
s
or
no
des,
w
hich
util
iz
es
to
obser
ve
env
i
ronme
ntal
sit
uation.
The
BS
is
act
s
the
owne
r
of
th
e
WSN
a
nd
al
l
mobil
e
sens
or
nodes
a
re
reg
i
ste
red
with
the
BS.
The
pro
po
se
d
sy
ste
m
util
iz
es
the
se
ns
or
no
des
li
ke
Diel
ect
ric
mo
ist
ure
sens
or
t
o
mea
s
ur
e
t
he
water
capaci
ty
in
the
s
oil,
te
m
per
at
ur
e
se
ns
or
obse
rv
es
th
e
s
oil
te
mp
e
ratu
re
,
Til
t
se
nsor
as
sessed
la
ye
r
of
so
il
,
Seismi
c
vib
rati
on
s
a
re
e
valua
te
d
by
a
pp
l
ying
acce
le
romet
er
a
nd
Ra
in
gauge
se
nsor
no
ti
ce
the
rai
n
fall
.
T
he
pro
po
se
d
sy
ste
m
obse
rve
s
a
sign
ific
a
nt
in
di
cat
or
via
glob
al
na
vig
at
io
n
s
at
el
li
te
sy
ste
m
(
GNSS)
e
qu
i
pm
e
nt
with
I
oT
se
nsors
.
The
disto
rtion
a
nd
de
va
sta
ti
on
de
velo
pm
e
nt
of
t
he
la
ndsli
de,
to
de
te
rmin
e
a
nd
pr
e
dict
dange
rous
c
on
diti
on
s
in
ti
me
f
or
e
xten
uatin
g
e
valuates
t
o
avo
i
d
the
li
fes
pan
fail
ure
i
nduced
by
unex
pe
ct
ed
disaste
rs.
A
w
ho
le
la
ndsli
de
detect
in
g
s
yst
em
ad
mit
s
da
ta
no
ti
ci
ng,
ac
qu
isi
ti
on,
t
ransmi
ssio
n,
rece
iving,
processi
ng,
de
ci
sion
,
ea
rly
al
ert,
an
d
reacti
on.
Fi
gure
2
e
xpla
ins
c
ompon
ents
of
Q
-
le
ar
ni
ng
base
d
f
ore
cast
ing
early la
nd
sli
de dete
ct
ion (
Q
-
L
FD
)
mec
ha
nism.
In
it
ia
ll
y,
the
s
ens
or
nodes
a
re
posit
io
ned
in
a
la
ndsli
de
reg
i
on,
a
nd
t
hese
nodes
ob
serv
e
s
the
n
forw
a
r
d
to
t
he
BS.
T
he
num
ber
of
se
ns
or
nodes
a
re
gro
uped
into
cl
us
te
rs
[
26]
by
re
c
ei
ved
st
rength
sign
al
ind
ic
at
io
n
(RS
SI
)
the
n
sel
ect
s
the
GH
base
d
on
t
he
di
ngo
opti
miza
ti
on
al
gorithm
(
DOA)
.
T
he
f
orma
ti
on
of
cl
us
te
r
base
d
on
t
hr
ee
le
vels:
le
sser,
mid
wa
y
an
d
hi
gh
e
r.
T
he
l
ow
e
r
le
vel
RSSI
no
des
no
t
able
t
o
c
onne
ct
the
gro
up,
thus;
it
le
aves
the
gr
oup.
T
he
valu
e
of
RSS
I
is
midd
le
,
t
he
se
ns
or
node
co
nsi
der
c
han
ces
a
grou
p
membe
r (
GM
)
.
Th
e
v
al
ue of R
SSI
gr
e
at
that
h
ig
hly cha
nce
s a GM
.
This
a
ppr
oac
h
util
iz
es
a
D
OA
to
pick
ou
t
an
e
ff
ic
ie
nt
GH
base
d
on
D
O
A
pr
ocess
li
ke
ci
rcu
m
fer
e
ntial
,
chasi
ng
a
nd a
tt
acking
t
he
obje
ct
.
Di
ngo
is
capa
ble
to
de
te
rmin
e
the
lo
cat
ion
of
t
he
obje
ct
.
Af
te
r
wa
rd
sear
chin
g
the
loca
ti
on
,
t
he
group
fo
ll
owe
d
by
al
ph
a
ci
rcles
the
ob
je
ct
.
It
is
recog
nized
t
h
at
the
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
omp E
ng,
V
ol.
15
, No
.
1
,
Febr
uary
20
25
:
425
-
434
428
reacha
ble
a
bet
te
r
age
nt
is
t
he
obje
ct
,
wh
ic
h
is
relat
ed
t
o
th
e
opti
mal
since
the
c
hasin
g
a
r
ea
is
not
recog
nized
a priori.
a.
Ci
rcu
m
fer
e
ntial
:
Along
with
t
he
ci
rcumst
anc
es
of
the
obje
c
t
(R*,
S*),
a
din
go
can
resto
r
e
it
s
locat
io
n
a
t
the
lo
cat
ion
of
(R,
S
).
Eac
h
e
xecu
ta
ble
l
ocati
on
s
are
note
d
arou
nd
the
bes
t
age
nt,
co
ns
id
erin
g
t
he
prese
nt
locat
ion
via
m
od
i
fy
i
ng
the
ve
ct
or
s
val
ue.
It
is
e
vid
e
ntly
de
monstrate
d
how
a
r
bitrar
y
vec
tors
1
as
well
as
2
al
locat
e
ding
oe
s
to
par
ti
ci
pat
e
e
very
l
ocati
on
a
m
ong
the
points.
T
he
di
ngoe
s
t
o
a
da
pt
t
heir
locat
io
ns
within t
he hunt
area a
bout t
he ob
je
ct
in
a
ny ra
ndom l
ocati
on is e
xpos
e
d
i
n
(
4
)
a
nd
(
5
)
.
→
=
|
→
.
→
(
)
−
→
(
)
|
(1)
→
(
+
1
)
=
→
(
)
−
→
.
→
(
)
(2)
wh
e
re
,
→
=
2
⋅
1
→
→
=
2
⋅
→
⋅
2
→
−
→
→
=
3
−
(
∗
(
3
)
)
b.
Chasin
g:
I
n
thi
s
par
t,
al
l
G
M
resem
bling
al
pha
as
well
as
be
ta
hav
e
a
n
im
pro
ved
a
wa
reness
re
gardin
g
t
he
ob
je
ct
l
ocati
on.
T
he
al
pha
di
ngo
e
ve
rlast
in
gly
operates
the
e
xp
l
or
i
ng.
On
the
ot
her
hand,
oc
casi
on
al
ly
beta
dingo
e
s
al
so
donate
in
e
xp
l
or
i
ng.
As
pe
r
the
locat
i
on
of
the
be
st
sea
rch
age
nt,
ot
he
r
dingo
e
s
requi
re
to
noti
f
y
thei
r
locat
io
n.
It
a
nnounce
s
that
al
ph
a
a
nd
beta
ding
oes
a
da
pt
their
l
ocati
ons
ra
ndom
l
y
a
nd
cal
culat
e
the
lo
cat
ion
of
t
he
obje
ct
in
the
sea
rch
s
pace.
A
fte
r
t
hat
we
wor
k
out
e
ve
ry
ding
o
i
nt
ensity
(
I)
i
s
sp
eci
fied
in (
6), (7) an
d (
8)
.
→
=
|
1
→
.
→
−
→
|
(3)
→
=
|
1
→
.
→
−
→
|
(4)
→
=
(
1
−
(
1
−
100
)
+
1
)
(5)
→
=
(
1
−
(
1
−
100
)
+
1
)
(6)
c.
Atta
ckin
g
o
bje
ct
:
If
there
are
no
ci
rc
umst
an
ces
modif
y,
it
act
s
din
go
fi
ni
sh
e
d
the
chas
e
via
at
ta
cking
t
he
ob
je
ct
.
Di
ngoe
s
cha
se
for
t
he
ob
je
ct
usual
ly
tog
et
he
r
with
the
gro
up
locat
i
on.
T
he
y
f
or
e
ve
r
tra
vel
proce
e
d
to
path
for
an
d
strike
pr
e
dat
or
s
.
Acc
ordin
gl
y,
it
is
a
ppli
ed
f
or
ra
ndom
va
lues
(RV
),
if
the
R
V
<
–
1
,
it
anno
un
ce
s
obj
ect
is
travel
l
eft
from
t
he
sea
rch
a
ge
nt,
th
ough
if
t
he
RV
>
1
,
it
den
otes
gro
up
nea
rs
the
ob
je
ct
. T
his
int
erv
e
ntio
n helps
the
dingo
e
s to
ob
s
er
ve
the
obj
ect
s g
lo
bally.
Figure
2. Com
pone
nts
of
Q
-
L
FD
mecha
nism
b
s
e
r
i
n
g
n
i
r
o
n
m
e
n
t
t
r
a
ns
m
i
t
t
he
da
t
a
t
o
l
u
s
t
e
r
o
r
m
a
t
i
o
n
e
l
e
c
t
i
o
n
D
a
t
a
a
g
g
r
e
g
a
t
i
o
n
f
o
r
a
r
d
t
h
e
d
a
t
a
t
o
r
e
c
e
i
e
d
t
h
e
d
a
t
a
n
a
l
e
t
h
e
d
a
t
a
o
r
e
c
a
s
t
i
n
g
a
n
d
s
l
i
d
e
o
r
ec
as
t
i
n
g
a
n
d
s
l
i
d
e
l
e
r
t
m
e
s
s
a
g
e
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
omp E
ng
IS
S
N:
20
88
-
8708
Q
-
le
arnin
g based f
or
eca
sti
ng ea
rly
l
ands
li
de
d
et
ect
io
n
in
…
(
Dev
asa
ha
y
am Jo
s
ep
h
Jey
akum
ar
)
429
2.1.
Q
-
le
ar
ning
b
ase
d landsl
ide
f
orecastin
g
Lan
ds
li
de
dete
ct
ing
est
a
blish
ed
on
I
oT
an
d
WSN p
r
ovides
real
ti
me d
et
e
ct
ion
, w
it
h
acc
ur
at
e
ness
a
nd
la
cking
a
ny
pe
rson
mist
ake
.
I
n
a
ddit
ion,
t
he
W
SN
cat
c
hes
an
im
porta
nt
I
oT
data
of
la
ndsli
de
pro
ne
re
gions.
WSN
is
a
pro
misi
ng,
reli
abl
e,
an
d
c
heap
equ
i
pm
e
nt
tha
t
prov
i
des
rea
l
-
ti
me
monit
ori
ng
more
e
xtensiv
e
distances
a
nd
ho
sti
le
te
rr
ai
ns
.
IoT
with
WSN
util
iz
e
adv
a
nce
d
co
m
munica
ti
ng
a
ppr
oac
h
an
d
e
xamine
com
pound
se
nsor
data.
It
is
a
ble
to
no
t
me
r
el
y
no
ti
ce
la
nd
sli
des
bu
t
ca
n
al
so
forecast
th
em.
All
of
the
s
e
data
are
ap
proac
ha
ble
to
t
he
gove
r
nm
e
nt
re
f
err
e
d
by
the
mobil
e
ap
pl
ic
at
ion
.
Re
la
te
d
a
uthoriti
es
li
ke
administrati
on
age
ncies,
a
nd
tra
gedy
ma
na
geme
nt
on
a
r
eal
ti
me
basis.
E
ven
local
pe
op
le
can
al
s
o
ob
ta
i
n
la
nd
sli
de
al
ar
ms
on
t
heir
m
ob
il
e
by
this
s
ys
te
m.
Admini
strat
ion
a
gen
ci
es
ca
n
al
so
dis
tribu
te
Re
scue
pla
ns
with
la
nd
sli
de
imp
resse
d
pe
op
le
s
.
Io
T
bas
ed
forecast
la
nd
sli
de
recog
ni
ti
on
a
nd
obse
rv
i
ng
s
ys
te
ms
giv
e
a
whole tra
ns
mis
sion cha
nnel
.
This
mecha
nis
m
util
iz
es
a
BS
act
s
a
ce
ntral
a
ge
nt
a
nd
eac
h
no
de
plays
a
n
a
ge
nt
,
an
d
the
y
consi
der
at
el
y
distrib
ute
the
da
ta
betwee
n
ne
ighbor
nodes
t
o
create
wh
ic
h
each
se
ns
or
disti
nguish
es
t
he
sta
te
transmissi
on
be
hav
i
or.
Q
-
Le
arn
i
ng
deter
mines
op
ti
mal
f
un
ct
io
n
ai
ms
to
f
or
ecast
t
he
la
nd
sli
d
e
.
T
he
Q
L
ob
je
ct
ives
to
f
or
ecast
the
re
ward
(R
W)
li
ke
la
ndsli
de
of
an
a
ge
nt
by
ac
ti
on
s.
T
he
age
nt
route
c
ollec
ti
on
is
employe
d
t
o
la
nd
sli
de
or
no
r
mal
the
co
ncerned
decisi
on.
Ther
e
f
or
e,
the
bette
r
decisi
on
is
sel
ect
ed
by
rew
a
rd.
The
Q
-
L
value
is
a
pp
li
ed
to
ob
ta
in
a
n
opti
mal
act
ion
pro
cess
pr
e
dict
th
e
f
uture
la
nd
sl
ide.
We
obse
r
ved
the
la
nd
sli
de
re
gion,
i
n
that
f
or
ec
ast
the
la
ndsli
de
via
R
W,
a
nd
le
sser
RW
res
ul
ts
are
el
imi
nat
ed.
He
re
S
a
ct
s
the
Stat
es,
and
A
com
pr
ise
s
th
e
act
ion
.
T
he
co
nclusi
on
to
de
ci
de
the
s
peci
f
ie
d
sta
te
's
act
i
on
s
is
to
im
pr
ov
e
t
he
RW
of
present
and f
uture R
W
.
This
mec
han
is
m
facto
r
a
s
f
ollow
s
{
SA,
AC,
RW,
P
O}.
W
he
re,
S
A
i
nd
ic
at
es
the
sta
te
,
A
C
re
pr
e
sents
the
act
ion
,
R
W
in
dicat
es
the
rew
a
rd,
as
w
el
l
as
PO
corr
esp
onds
the
fe
asi
bili
ty
of
la
nds
li
de.
Allow
curren
t
sta
te
play
s
t
he
,
f
uture
sta
te
in
dicat
es
t
he
,
a
nd
r
efe
rs
t
he
ti
me
of
wa
it
ing
for
gat
he
red
data.
The
QL
-
ta
ble
help
s
to
detect
ing
a
bette
r
act
ion
f
or
eac
h
sta
te
,
The
val
ue
(
,
)
offer
s
t
he
RW
of
c
urren
t
a
nd
fu
t
ur
e
w
h
il
e
a
ct
ion
a
is
pe
rformed
at
.
We
co
ns
i
der
that
the
a
gen
t
sel
ec
ts
an
act
io
n
ac
in
,
dete
rmin
e
s
RW a
nd ex
te
nds i
nto
fu
t
ur
e
s
ta
te
′
. N
e
xt,
t
he QL,
(
,
)
is f
orm
ed
as
(
7)
,
(
,
)
−
(
1
−
)
(
,
)
+
{
+
.
(
′
,
)
}
(7)
w
he
re,
ref
er
s t
he Q
-
le
ar
ning
r
at
e and
dep
ic
t
s the
forthc
om
i
ng RW
d
isc
ount as
pect.
Take
t
he
act
io
n
in
dicat
es
the
ag
gr
e
gated
da
ta
is
transmit
t
he
c
urren
t
sta
t
e
to
nex
t
sta
te
,
the
RW
is
determi
ned
the
curre
nt
sta
te
SA
;
the
act
ion
of
Q
V
-
ta
ble
f
or
sta
te
is
m
odifie
d.
It
e
nha
nces
t
he
qu
al
it
y
of
serv
ic
e
(
Q
oS
)
and
data
a
ggre
gation;
f
ur
t
hermo
re,
it
is
c
ompu
te
d
at
the
f
ut
ur
e
sta
te
.
The
RW
ru
le
is
util
iz
ed
to
choose
a
Q
-
Le
arn
i
ng
best
s
olu
ti
on.
The
n
ca
lc
ulate
the
R
W
by
water
c
apacit
y
(
WC),
te
mp
e
ratu
re
(
T),
so
il
la
yer
(SL),
sei
smic
vib
rati
ons
(
SV)
a
nd
ra
infall
(R
)
an
d
RW
cal
culat
ion
e
qu
at
io
n
is
bel
ow.
W
here,
the
add
it
io
nal
disc
ount
a
sp
ect
is
app
li
ed
to
the
RW,
w
hich
is
need
e
d
to
eva
de
bac
k
wa
r
ding
a
nd
disc
ount
asp
ect
range i
nvolv
i
ng
0
to
1.
=
×
(
+
+
+
)
(8)
3.
EXPERI
MEN
TAL RES
UL
TS A
ND DIS
CUSSIO
N
The
Q
-
le
ar
ning
base
d
f
oreca
sti
ng
earl
y
la
ndsli
de
detect
io
n
(
Q
-
L
FD)
m
echan
is
m
is
e
xecu
te
d
an
d
examine
d
in
la
borato
ry
with
act
ual
se
nsors
f
or
one
-
m
inu
te
.
W
SN
c
omprise
seve
r
al
sens
or
node
s
li
ke
te
mp
erat
ur
e
a
nd
rain
fall
ob
serv
i
ng
se
nsor
,
water
ca
pacit
y
obser
ving
s
ens
or
,
vibrat
io
n
obser
vi
ng
s
ens
or
.
Figure
3
e
xpla
ins
the
ha
r
dware
for
f
or
e
cast
ing
la
ndsli
de
.
T
his
fig
ur
e
con
ta
in
s
te
mp
e
ratur
e
a
nd
rainf
a
ll
ob
s
er
ving
se
nsor
,
water
ca
pac
it
y
obser
ving
s
ens
or
,
vibrat
io
n
obse
rv
i
ng
se
ns
or.
These
se
ns
or
nodes
obs
erv
i
ng
and
ea
rly
wa
rni
ng
platf
orm
offe
rs
a
c
omplet
e
inv
e
sti
gatio
n
of
the
ob
s
er
ving
data
t
hat
al
lots
vis
ualiz
at
ion
of
data f
rom all
obser
ving l
ocati
on
s
f
or the
lan
ds
li
de.
F
i
g
u
r
e
4
d
e
m
o
n
s
t
r
a
t
e
s
t
h
e
O
B
I
A
a
n
d
Q
-
L
F
D
o
f
t
h
e
l
a
n
d
s
l
i
d
e
f
o
r
e
c
a
s
t
i
n
g
a
c
c
u
r
a
c
y
r
a
t
i
o
a
g
a
i
n
s
t
n
o
d
e
d
e
n
s
i
t
y
.
T
h
e
n
o
d
e
d
e
n
s
i
t
y
s
e
n
s
i
t
i
v
e
l
y
d
e
t
e
r
m
i
n
e
s
Q
-
L
F
D
m
e
t
h
o
d
a
c
c
u
r
a
c
y
.
T
h
e
n
o
d
e
s
i
z
e
i
n
c
r
e
a
s
e
s
t
h
e
a
c
c
u
r
a
c
y
r
a
t
i
o
v
a
g
u
e
l
y
i
n
c
r
e
a
s
e
d
i
n
t
h
e
W
S
N
.
T
h
e
Q
-
L
F
D
m
e
c
h
a
n
i
s
m
r
e
s
u
l
t
s
e
x
p
l
a
i
n
t
o
r
e
a
c
h
g
r
e
a
t
a
c
c
u
r
a
c
y
.
F
rom
Figure
4,
w
he
n
raises
t
he
se
nsor
no
des
t
he
a
ccur
ac
y
rati
o
of
OBI
A
a
nd
Q
-
LF
D
me
cha
ni
sm
is
i
ncr
ease
d.
The
Q
-
L
FD
mec
ha
nism
acc
uracy
rati
o
is
high
since
it
forecast
s
the
la
nd
sli
de
eff
ic
ie
nt
ly
.
But
,
the
Q
-
LF
D
c
an
no
t
able
to
f
oreca
s
t
the
la
nd
sli
de
eff
ic
ie
ntly
.
Durin
g
la
ndsli
de
detect
ion,
the
mecha
nism
f
oreca
sts
the
cha
nc
e
of
a
la
nd
sli
de
.
But
there
is
no
c
ha
nce
of
a
la
ndsli
de
occ
urrin
g.
From
t
he
re
s
ults,
the
mech
anism
is
detect
ing
it
wrongl
y,
w
hic
h
is
cal
le
d
a
f
al
se
al
arm.
Fig
ur
e
5
ex
plains
the
False
al
ar
m
cha
nces
of
t
he
OBI
A
a
nd
QLFD
against e
xperi
ments c
ount.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
omp E
ng,
V
ol.
15
, No
.
1
,
Febr
uary
20
25
:
425
-
434
430
Figure
3. Ha
rdwar
e
for
la
ndsli
de
pre
dicti
on
Figure
4. Acc
uracy
of
OBI
A and Q
-
LF
D
a
ga
inst
no
de den
sit
y
Figure
5. False
a
la
rm of OB
I
A
a
nd Q
-
LF
D agai
ns
t
it
erati
ons
0
0.2
0.4
0.6
0.8
1
1.2
20
40
60
80
100
OB
IA
Q
-LFD
Sensor
Nodes
Ac
curacy
Rat
io
0
0.05
0.1
0.15
0.2
0.25
0.3
10
20
30
40
50
OB
IA
Q
-LFD
Number
of Iterat
ion
s
False
Alarm
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
omp E
ng
IS
S
N:
20
88
-
8708
Q
-
le
arnin
g based f
or
eca
sti
ng ea
rly
l
ands
li
de
d
et
ect
io
n
in
…
(
Dev
asa
ha
y
am Jo
s
ep
h
Jey
akum
ar
)
431
Fr
om
Fig
ur
e
5,
e
valuate
t
he
numb
e
r
ex
per
i
ments
t
he
Q
-
L
FD
mec
han
is
m
false
al
ar
m
rate
is
belo
w
0.1
pe
rce
ntage
.
Since,
t
he
Q
-
L
FD
mecha
nism
f
or
e
cast
s
the
la
nd
sli
de
eff
ic
ie
ntly.
It
is
sp
eci
fied
as
the
c
orrelat
ion
am
ong
the
c
ount
of
normal
ti
me
is
i
na
pprop
riat
el
y
pr
e
dicti
ng
a
la
ndsli
de
as
well
as
the
tota
l
co
unt
pr
e
dicti
on.
B
ut
,
the
e
xisti
ng
OBI
A
mecha
ni
sm
raises
t
he
f
al
se
al
arm
bec
ause,
it
ca
n
no
t
forecast
t
he
la
nd
sli
de
eff
ic
ie
ntly
. F
i
gure
6 desc
ribes
the
false p
os
it
ive
rati
o
a
gainst
it
erati
on
s.
Figure
6
il
lust
rates
the
false
posit
ive
rati
o
of
Q
-
LF
D
a
nd
OBI
A
mec
ha
nisms.
W
he
n
increase
s
a
n
it
erati
on
c
ount
,
the
rati
o
of
f
al
se
posit
ive
is
hi
gh.
The
Q
-
LFD
mec
ha
nism
rate
of
the
highest
false
posit
ive
rati
o
is
0.11
a
nd
t
he
lo
west
le
vel
of
false
po
sit
ive
rate
is
0.01.
Be
ca
us
e
,
of
t
he
Q
-
LF
D
mecha
nism
uti
li
zes
a
Q
-
le
ar
ning
t
o
detect
the
ea
rly
la
ndsli
de
ef
f
ic
ie
ntly.
It
is
sp
eci
fied
as
t
he
correla
ti
on
a
mong
the
c
ou
nt
of
normal
ti
me
i
s
inap
pro
pr
ia
t
el
y
predict
in
g
a
la
ndsli
de
a
s
well
as
the
total
count
it
erati
on
s
.
Howe
ver,
the
existi
ng
OB
IA
mec
han
is
m
i
ncr
ease
s
th
e
f
al
se
posit
ive
r
at
io
since
it
c
an
no
t
forecast
the
la
ndsli
de
well
.
Figure
7 desc
ribes
t
he
false
n
e
gative
rati
o
a
ga
inst it
erati
on
s
.
Figure
7
il
lustrate
s
the
false
neg
at
ive
rati
o
of
Q
-
LF
D
an
d
OB
IA
mec
ha
nis
ms.
Wh
e
n
increase
s
a
n
it
erati
on
co
unt,
the
rati
o
of
fa
lse
neg
at
ive
ra
ti
o
is
high.
T
he
rate
of
the
m
aximum
false
neg
at
ive
rati
o
is
0.1
and
the
lo
west
le
vel
of
false
neg
at
ive
rate
i
s
0.0
3.
F
or
the
reas
on
th
at
,
of
the
Q
-
LF
D
m
echan
is
m
a
pplyin
g
a
Q
-
le
ar
ning
t
o
disti
nguish
the
early
la
ndsli
de
powe
rfull
y.
But,
the
existi
ng
OB
IA
mec
han
is
m
lo
west
le
vel
of
false
ne
gative
rate
is
0.0
7
an
d
the
highest
f
al
se
neg
at
ive
r
at
e
is
0.
21.
Sin
ce
it
can
n
ot
ab
le
to
befo
re
f
oreca
st
the lan
ds
li
de.
Figure
6. False
posit
ive r
at
i
o of
OBI
A
a
nd
Q
-
L
FD agai
ns
t
it
erati
on
s
Figure
7. False
ne
gative
rati
o
of
OBI
A
a
nd
Q
-
L
FD agai
ns
t
it
erati
on
s
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
0.2
10
20
30
40
50
OB
IA
Q
-LFD
False
Po
siti
ve
Rat
io
Number
of Iterat
ion
s
0
0.05
0.1
0.15
0.2
0.25
10
20
30
40
50
OB
IA
Q
-LFD
False
Neg
at
ive
Rat
io
Number
of Iterat
ion
s
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
omp E
ng,
V
ol.
15
, No
.
1
,
Febr
uary
20
25
:
425
-
434
432
4.
CONCL
US
I
O
N
This
pa
per
give
s
the
Q
-
le
ar
ni
ng
base
d
forec
ast
ing
early
la
nd
sli
de
detect
ion
mec
ha
nism
desi
gn
a
nd
execu
ti
on
of
Io
T
WSN
est
ablishe
d
on
r
eal
-
ti
me
obser
ving
of
e
nv
i
r
onmental
ar
gu
ments
f
or
la
ndsli
des
pr
e
dicti
on.
I
niti
al
ly,
the
se
nso
r
nodes
a
re
f
orm
the
gro
up
th
en
sel
ect
the
group
hea
d
base
d
on
D
OA
al
gorith
m
fitness
f
unct
io
n.
Re
al
ti
me
obser
ving
of
s
oi
l
par
amet
er
s
a
nd
ea
rly
pr
e
dic
ti
on
syst
em
ap
ply
in
g
par
a
mete
rs
li
ke
so
il
wate
r
ca
pa
ci
ty,
s
oil
la
ye
r,
so
il
te
m
pe
ratu
re,
sei
smic
vibrat
ion
s
,
a
nd
rai
nf
al
l.
The
se
pa
rameters
value
s
are
com
par
e
d
a
nd
early
warnin
gs
are
ge
ner
at
e
d
by
a
pp
l
ying
-
l
earn
i
ng
al
gorit
hm
.
T
hen
al
ar
m
is
a
ct
ivate
d
in
the
reg
i
on
to
noti
fy
a
bout
occurr
ing
of
La
ndsli
de
in
earl
y.
T
hus,
a
voids
m
ore
da
ng
e
r
ou
s
is
su
es
due
to
la
ndsli
de.
The
e
xp
e
rime
ntal
resu
lt
s
re
veals
that
the
Q
-
L
FD
mecha
nism
mi
nimize
s
the
false
posit
ive,
false
neg
at
ive
rati
o.
Furthe
rm
or
e
,
it
incre
a
se
s
the d
et
ect
ion acc
ur
ac
y
rati
o.
In
f
uture,
t
o
prov
i
de
data
sec
ur
it
y
f
or
i
nformat
ion
gathe
rin
g
in
land
sli
de
e
nv
ir
onme
nt.
REFERE
NCE
S
[1]
K.
Das,
.
ajum
d
ar,
.
ou
lik
,
an
d
.
u
jita,
“Re
al
-
tim
e
th
resh
o
ld
-
b
a
sed
lan
d
slid
e
p
redictio
n
sy
stem
for
h
illy
regio
n
u
sin
g
ireless
sen
so
r
n
e
t o
rks
,”
in
2
0
2
0
IE
E
E
Inter
n
a
tio
n
a
l
Co
n
feren
ce
o
n
Co
n
su
mer
Electro
n
ics
-
Ta
iwa
n
(I
C
CE
-
Ta
iwa
n
)
,
Sep
.
2
0
2
0
,
p
p
.
1
–
2
,
d
o
i:
10
.11
0
9
/ICCE
-
Taiwan4
9
8
3
8
.20
2
0
.9
2
5
8
1
8
1
.
[2]
.
m
g
ain
,
.
Ku
m
ar
,
.
ajg
ain
,
a
n
d
.
Rai,
“
and
slid
es
p
redictio
n
an
d
d
etectio
n
u
sin
g
Io
T
s
stem,
”
in
2
0
2
3
2
n
d
Inter
n
a
tio
n
a
l
Co
n
feren
ce
o
n
Vi
sio
n
To
wa
rd
s
Emerg
in
g
Tren
d
s
in
Co
mmu
n
ica
tio
n
a
n
d
Netw
o
rkin
g
Tech
n
o
lo
g
ies
(V
iTE
Co
N)
,
May
2
0
2
3
,
p
p
.
1
–
6
,
d
o
i: 10
.1
1
0
9
/ViTE
Co
N5
8
1
1
1
.20
2
3
.1
0
1
5
7
0
7
7
.
[3]
.
Ku
m
ar
.
an
d
.
V.
Ra
m
esh
,
“
ccurate
IoT
b
ased
slo
p
e
in
stab
ilit
sen
sin
g
s
stem
for
lan
d
slid
e
d
etectio
n
,”
IE
EE
S
en
so
rs
Jo
u
rn
a
l
,
v
o
l.
2
2
,
n
o
.
1
7
,
p
p
.
1
7
1
5
1
–
1
7
1
6
1
,
S
ep
.
2
0
2
2
,
d
o
i: 10
.11
0
9
/JSEN.
2
0
2
2
.3
1
8
9
9
0
3
.
[4]
.
.
,
V
.
.
.
.
,
an
d
.
h
aji,
“
an
d
slid
e
id
en
tification
u
sin
g
m
achi
n
e
learnin
g
tech
n
iq
u
es:
Re
ie ,
m
o
ti
atio
n
,
an
d
futu
re
p
ros
p
ects,”
Ear
th
Scien
ce I
n
fo
rma
tics
,
v
o
l.
1
5
,
n
o
.
4
,
p
p
.
2
0
6
3
–
2
0
9
0
,
Dec. 2
0
2
2
,
d
o
i
: 1
0
.10
0
7
/s1
2
1
4
5
-
022
-
0
0
8
8
9
-
2.
[5]
S.
R.
M
eena
et
a
l.
,
“ and
slid
e
d
etectio
n
in
th
e
im
ala
as
u
sin
g
m
achi
n
e
learnin
g
alg
o
rithms
an
d
U
-
et,”
La
n
d
slid
es
,
v
o
l.
1
9
,
n
o
.
5
,
p
p
.
1
2
0
9
–
1
2
2
9
,
2
0
2
2
,
d
o
i: 1
0
.10
0
7
/s1
0
3
4
6
-
022
-
0
1
8
6
1
-
3.
[6]
.
K
a
u
s
h
a
l
a
n
d
V
.
K
.
e
h
g
a
l
,
“
T
h
r
e
s
h
o
l
d
b
a
s
e
d
r
e
a
l
-
t
i
m
e
l
a
n
d
s
l
i
d
e
p
r
e
d
i
c
t
i
o
n
s
y
s
t
e
m
u
s
i
n
g
l
o
w
-
c
o
s
t
s
e
n
s
o
r
n
e
t
o
r
k
s
,
”
i
n
2
0
2
3
3
r
d
A
s
i
a
n
C
o
n
f
e
r
e
n
c
e
o
n
I
n
n
o
v
a
t
i
o
n
i
n
T
e
c
h
n
o
l
o
g
y
(
A
S
I
A
N
C
O
N
)
,
A
u
g
.
2
0
2
3
,
p
p
.
1
–
7
,
d
o
i
:
1
0
.
1
1
0
9
/
A
S
I
A
N
C
O
N
5
8
7
9
3
.
2
0
2
3
.
1
0
2
6
9
9
3
1
.
[7]
D.
Miya
m
o
to
et
a
l.
,
“ o
n
stru
ctio
n
o
n
irel
ess
lin
k
b
et
een
Io
T
sen
so
r
n
o
d
es
an
d
g
ate a
for
lan
d
slid
es
p
red
ictio
n
s
stem,
”
i
n
2
0
2
1
IE
EE
US
NC
-
URSI
Rad
io
S
cien
ce
Meetin
g
(Joint
with
AP
-
S
S
ymp
o
siu
m)
,
Dec.
2
0
2
1
,
p
p
.
1
2
2
–
1
2
3
,
d
o
i
:
1
0
.23
9
1
9
/USNC
-
URSI5
1
8
1
3
.2021
.
9
7
0
3
6
2
3
.
[8]
F.
S.
T
eh
rani,
G.
San
tin
elli,
an
d
.
er
rer
a
e
rr
e
ra,
“ ulti
-
regio
n
al
lan
d
slid
e
d
etecti
o
n
u
sin
g
co
m
b
in
e
d
u
n
su
p
ervis
ed
an
d
su
p
er is
ed
m
achi
n
e
learnin
g
,”
Ge
o
ma
tics,
Na
tu
ra
l
Ha
za
rd
s
a
n
d
Risk
,
v
o
l.
1
2
,
n
o
.
1
,
p
p
.
1
0
1
5
–
1
0
3
8
,
2
0
2
1
,
d
o
i
:
1
0
.10
8
0
/1
9
4
7
5
7
0
5
.20
2
1
.1912
1
9
6
.
[9]
A.
Jo
sh
i,
D.
P
.
K
a
n
u
n
g
o
,
an
d
R.
K.
Pan
ig
rahi,
“De el
o
p
m
en
t
o
f
lan
d
slide
forecastin
g
s
stem
u
sin
g
d
eep
learnin
g
,”
in
2
0
2
3
IE
EE
App
lied
Sen
sin
g
C
o
n
feren
ce (
AP
S
CO
N)
,
Jan
.
2
0
2
3
,
p
p
.
1
–
3
,
d
o
i: 10
.1
1
0
9
/
APSCON5
6
3
4
3
.2
0
2
3
.1010
1
2
2
3
.
[10
]
O.
Gh
o
rban
zadeh
,
H.
Sh
ah
ab
i,
A.
Criv
ellari
,
.
o
m
a
o
u
n
i,
T.
las
ch
k
e,
an
d
P.
h
a
m
isi,
“ and
slid
e
d
etectio
n
u
sin
g
d
eep
learnin
g
and
ob
ject
-
b
ased
im
ag
e
an
al
sis
,”
La
n
d
slid
es
,
v
o
l.
1
9
,
n
o
.
4
,
p
p
.
9
2
9
–
9
3
9
,
Ap
r.
20
2
2
,
d
o
i: 10
.10
0
7
/s103
4
6
-
0
2
1
-
0
1
8
4
3
-
x.
[11
]
.
R.
u
r a an
sh
i
an
d
U.
.
D
esh
p
an
d
e,
“Re
ie
o
f
ri
sk
m
an
ag
em
en
t
for
lan
d
slid
e
forecas
tin
g
,
m
o
n
ito
ring
an
d
p
redictio
n
u
sing
ireless
sen
so
rs
n
et o
rk,”
in
2
0
1
7
Inter
n
a
tio
n
a
l
C
o
n
f
eren
ce
o
n
Inn
o
va
t
io
n
s
in
Info
rma
tio
n
,
Embed
d
ed
a
n
d
Co
mmu
n
ica
tio
n
S
ystems
(
IC
II
ECS
)
,
Ma
r.
20
1
7
,
p
p
.
1
–
6
,
d
o
i: 10
.11
0
9
/ICII
ECS
.20
1
7
.82
7
6
1
1
3
.
[12
]
.
.
h
m
ed
,
.
Po
th
alaiah,
an
d
D.
.
Rao
,
“Real
-
tim
e
m
o
n
ito
ring
o
f
p
artially
stab
le
slo
p
es
for
lan
d
slid
e
p
r
ed
ictio
n
b
y
u
sing
ireless
sen
so
r
n
et o
rks
,”
in
2
0
1
6
On
lin
e
Inter
n
a
tio
n
a
l
Co
n
feren
ce
o
n
Green
Eng
in
eerin
g
a
n
d
Tech
n
o
lo
g
i
es
(I
C
-
G
E
T)
,
No
v
.
2
0
1
6
,
p
p
.
1
–
5
,
d
o
i:
10
.11
0
9
/GET
.20
1
6
.79
1
6
6
3
8
.
[13
]
T.
Kin
o
sh
ita
et
a
l.
,
“ stim
atio
n
o
f
p
rop
ag
atio
n
p
erf
o
rman
ce
b
et een
IoT
term
in
als
an
d
g
ate
a
u
sin
g
U
-
b
an
d
s
for
lan
d
slid
es
p
redictio
n
s
ste
m
,”
in
2
0
2
1
IE
EE
Asia
-
Pacifi
c
Micr
o
wa
ve
Co
n
feren
ce
(A
P
MC
)
,
No
v
.
2
0
2
1
,
p
p
.
3
2
9
–
3
3
1
,
d
o
i:
1
0
.11
0
9
/APMC5
2
7
2
0
.2021
.9661
8
0
1
.
[14
]
D.
Jo
sep
h
Je
ak
u
m
ar
an
d
.
ing
e
sh
ari,
“
ak
e
sen
so
r
d
etectio
n
an
d
secu
re
d
ata
tr
an
smissio
n
b
ased
o
n
p
redicti
e
p
arser
i
n
s,”
Wir
eless
Perso
n
a
l Co
mmu
n
ica
tio
n
s
,
v
o
l.
1
1
0
,
n
o
.
1
,
p
p
.
5
31
–
5
4
4
,
Jan
.
2
0
2
0
,
d
o
i: 10.
1
0
0
7
/s112
7
7
-
019
-
0
6
7
4
0
-
0.
[15
]
.
n
u
radh
a,
.
b
in
a
a,
.
h
arat
h
i,
.
Jan
an
i,
an
d
.
Kh
an
,
“IoT
b
as
ed
n
atu
ral
d
isas
ter
m
o
n
ito
ring
an
d
p
re
d
ictio
n
an
al
sis
for
h
ills
area
u
sin
g
T
n
et o
rk,”
8
t
h
Inter
n
a
tio
n
a
l
Confer
en
ce
o
n
Adva
n
c
ed
Co
mp
u
tin
g
a
n
d
Co
mmu
n
ica
tio
n
S
ystems
,
IC
ACC
S
2022
,
p
p
.
1
9
0
8
–
1
9
1
3
,
2
0
2
2
,
d
o
i: 1
0
.1
1
0
9
/ICACC
S5
4
1
5
9
.20
2
2
.9
7
8
5
1
2
1
.
[16
]
P.
re
e
id
a,
.
.
b
h
ilash
,
J.
Pau
l,
an
d
.
Rejith
k
u
m
a
r,
“
m
a
ch
in
e
lear
n
in
g
-
b
ased
earl
lan
d
slid
e
arnin
g
s stem
u
sin
g
IoT,”
in
2
0
2
1
4
th
Bienn
ia
l
Inter
n
a
tio
n
a
l
Co
n
feren
ce
o
n
Na
scen
t
Tech
n
o
lo
g
ies
in
Eng
in
eeri
n
g
(I
CN
TE
)
,
Jan
.
2
0
2
1
,
p
p
.
1
–
6
,
d
o
i:
1
0
.11
0
9
/I
CNTE51
1
8
5
.2021
.9487
6
6
9
.
[17
]
.
P.
h
atta
an
d
.
Than
g
ad
u
rai,
“
Detection
an
d
p
redictio
n
o
f
calam
it
o
u
s
lan
d
sli
d
e
in
p
r
ecip
ito
u
s
h
ills,”
in
2
0
1
6
Inter
n
a
tio
n
a
l
Co
n
feren
ce
o
n
Adva
n
ced
Co
mmu
n
ica
tio
n
Co
n
tro
l
a
n
d
Co
mp
u
tin
g
Tech
n
o
lo
g
ies
(I
CACC
CT)
,
May
2
0
1
6
,
p
p
.
2
3
8
–
2
4
0
,
d
o
i:
1
0
.11
0
9
/ICACC
CT.20
1
6
.78
3
1
6
3
8
.
[18
]
P.
eh
ta,
D.
h
an
d
er,
.
h
ah
im
,
K
.
Tej
as i,
.
.
erchan
t,
an
d
U
.
.
Desai,
“Dist
ribu
ted
d
etectio
n
for
lan
d
slid
e
p
redictio
n
u
sin
g
ireless
sen
so
r
n
et o
rk,”
in
2
0
0
7
First
Inter
n
a
ti
o
n
a
l
Glo
b
a
l
Info
r
ma
tio
n
Infr
a
str
u
ct
u
re
S
ymp
o
siu
m
,
2
0
0
7
,
p
p
.
1
9
5
–
1
9
8
,
d
o
i: 10
.1109
/GII
S.
2
0
0
7
.4
4
0
4
1
9
0
.
[19
]
K.
Tej
as i,
P.
e
h
ta,
R.
an
sal,
.
Parekh
,
.
.
er
ch
an
t,
an
d
U.
.
Desai,
“Ro
u
tin
g
p
roto
co
ls
for
lan
d
sli
d
e
p
redictio
n
u
sin
g
ireless
sen
so
r
n
e
t o
rks
,”
in
2
0
0
6
Fou
rth
Inter
n
a
ti
o
n
a
l
Co
n
feren
ce
o
n
Intellig
en
t
S
en
sin
g
a
n
d
Info
rma
tio
n
Pro
cess
in
g
,
Dec.
2
0
0
6
,
p
p
.
4
3
–
4
7
,
d
o
i: 10
.11
0
9
/ICISI
P
.20
0
6
.4286
0
5
7
.
[20
]
.
h
m
ed
,
.
a
h
ajan
,
.
u
p
ta,
a
n
d
.
u
ri,
“ n
o
p
tim
al
selectio
n
o
f
rou
tin
g
p
rot
o
co
l
for
d
iff
er
en
t
sin
k
p
lacem
en
ts
in
a
wirele
ss
sen
so
r
n
et o
rk
for
lan
d
slid
e
d
etectio
n
s
stem,
”
in
2
0
1
4
Inter
n
a
tio
n
a
l
Confer
en
ce
o
n
Co
mp
u
t
a
tio
n
a
l
Intellig
en
ce
a
n
d
Co
mmu
n
ica
tio
n
N
etwo
rks
,
No
v
.
2
0
1
4
,
p
p
.
3
5
8
–
3
6
3
,
d
o
i: 10
.11
0
9
/CICN.2
0
1
4
.87
.
[21
]
S.
-
.
in,
.
.
lsh
eh
ri,
P
.
an
g
,
an
d
I.
.
k
ild
i ,
“
ag
n
etic
in
d
u
ctio
n
-
b
ased
lo
caliz
atio
n
in
rand
o
m
ly
d
ep
lo
y
ed
wireles
s
u
n
d
ergrou
n
d
sen
s
o
r
n
et o
rks
,”
I
E
EE
Inter
n
et
o
f
Th
in
g
s
Jo
u
r
n
a
l
,
v
o
l.
4
,
n
o
.
5
,
p
p
.
1
4
5
4
–
1
4
6
5
,
Oct.
2
0
1
7
,
d
o
i:
1
0
.11
0
9
/JIOT.
2
0
1
7
.27
2
9
8
8
7
.
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
omp E
ng
IS
S
N:
20
88
-
8708
Q
-
le
arnin
g based f
or
eca
sti
ng ea
rly
l
ands
li
de
d
et
ect
io
n
in
…
(
Dev
asa
ha
y
am Jo
s
ep
h
Jey
akum
ar
)
433
[22
]
S.
-
F.
Ch
en
an
d
P.
-
.
siu
n
g
,
“ and
slid
e
p
redicti
o
n
ith
m
o
d
el
s itch
in
g
,”
in
2
0
1
7
IE
EE
C
o
n
feren
ce
o
n
Dep
en
d
a
b
le
a
n
d
S
ecur
e
Co
mp
u
tin
g
,
Au
g
.
2
0
1
7
,
p
p
.
2
3
2
–
2
3
6
,
d
o
i: 10
.1109
/DESE
C.2
0
1
7
.80
7
3
8
4
6
.
[23
]
.
m
ale
an
d
R.
Patil,
“
IoT
b
ased
rainfall
m
o
n
ito
ri
n
g
s
stem
u
sin
g
en
ab
led
arc
h
itectu
re,
”
in
2
0
1
9
3
rd
In
tern
a
tio
n
a
l
Co
n
feren
ce
o
n
Co
mp
u
tin
g
Meth
o
d
o
lo
g
ies
a
n
d
Co
mmu
n
ica
tio
n
(I
CC
MC
)
,
Ma
r.
2
0
1
9
,
p
p
.
7
8
9
–
7
9
1
,
d
o
i:
1
0
.11
0
9
/ICCMC.2
0
1
9
.8819
7
2
1
.
[24
]
.
Zaid
alah
,
.
K.
el
ap
erumal
,
an
d
R.
b
d
u
lla,
“ ccele
romete
r
-
b
a
sed
eld
erly
fall
d
e
tectio
n
sy
stem
u
sin
g
ed
g
e
arti
ficial
in
tellig
en
ce
architectu
re,
”
Inter
n
a
tio
n
a
l
Jo
u
r
n
a
l
o
f
Electrica
l
a
n
d
Co
mp
u
ter
Eng
in
eerin
g
,
v
o
l.
1
2
,
n
o
.
4
,
p
p
.
4
4
3
0
–
4
4
3
8
,
Au
g
.
2
0
2
2
,
d
o
i: 1
0
.11
5
9
1
/ijece.v1
2
i4
.pp
4
4
3
0
-
4
4
3
8
.
[25
]
B.
T
.
Ph
a
m
et
a
l
.
,
“
n
o
el
in
telli
g
en
ce
ap
p
roach
o
f
a
seq
u
en
tial
m
i
n
im
al
o
p
tim
i
zatio
n
-
b
ased
su
p
p
o
rt
v
ecto
r
m
achi
n
e
for
lan
d
slid
e su
scep
tib
ilit
m
ap
p
in
g
,”
S
u
sta
in
a
b
ility
,
v
o
l.
1
1
,
n
o
.
2
2
,
No
v
.
2
0
1
9
,
d
o
i: 10
.3390
/s
u
1
1
2
2
6
3
2
3
.
[26
]
.
K.
Jeml
a
aik
,
.
Par
am
es
ara
p
p
a,
an
d
.
.
Ramachan
d
ra,
“
n
erg
eff
icien
t
d
a
ta
trans
m
iss
io
n
u
sin
g
m
u
ltio
b
jecti
e
im
p
ro
ed
rem
o
r
a
o
p
tim
i atio
n
alg
o
rithm
for
ireless
s
en
so
r
n
et o
rk
ith
m
o
b
ile
sin
k
,”
Inter
n
a
tio
n
a
l
Jo
u
rna
l
o
f
Electrica
l
a
n
d
Co
mp
u
ter Eng
i
n
ee
rin
g
,
v
o
l.
1
3
,
n
o
.
6
,
p
p
.
6
4
7
6
–
6
4
8
8
,
Dec.
2
0
2
3
,
d
o
i: 10
.11
5
9
1
/ijece.v1
3
i6
.
p
p
6
4
7
6
-
6
4
8
8
.
BIOGR
AP
HI
ES OF
A
UTH
ORS
Devasah
ayam
Joseph
Jeyak
um
ar
is
an
associa
t
e
prof
essor
in
the
D
epa
rt
me
nt
of
El
e
ct
roni
cs
and
Comm
uni
catio
n
at
J.N.N
Insti
tut
e
of
Eng
inee
ring
.
He
re
ceiv
ed
his
Ph.D
.
degr
ee
f
rom
An
na
unive
rsi
ty.
H
e
has
a
to
tal
exp
eri
en
ce
of
26
y
e
ars
which
includ
es
te
a
chi
ng
as
well
as
industrial.
His
cur
r
ent
rese
arc
h
int
e
rest
s
are
sign
al
pro
ce
ss
ing,
wire
le
s
s
net
works
,
wire
le
ss
s
ensor
ne
two
rk
and
cogni
ti
v
e
rad
io
ne
tw
ork
.
He
ca
n
be
cont
a
cted
at
em
a
il
:
ja
yaku
ma
rjosep
h33@gma
il.c
o
m
.
Boominathan
Sh
anmat
hi
co
mpl
e
te
d
b
ac
he
lor
of
enginee
ring
degr
ee
i
n
el
e
ct
roni
cs
and
c
omm
unication
fr
om
J.N.N
Insti
tu
te
of
Engi
n
ee
rin
g,
Kann
igaipai
r
aff
iliated
to
Anna
Univer
sity
in
2016.
He
co
mpl
eted
m
aste
r
of
engi
n
ee
ring
i
n
appl
i
ed
e
lectr
o
nic
s
from
Sri
Venka
te
sw
ara
Coll
ege
of
Eng
ine
er
ing,
Sriper
am
ba
thur
aff
i
li
a
te
d
to
Anna
Univer
sity
,
an
d
Chenna
i
in
201
8.
Curre
nt
ly
she
is
pursuing
Ph.D.
at
Anna
Uni
ver
sity,
Ch
enna
i
.
Area
of
h
er
rese
arc
h
is
app
li
c
at
ion
of
i
mage
proc
essing,
com
munica
ti
on
sys
te
m,
signal
and
im
ag
e
proc
essing,
digi
t
al
logi
c
c
irc
u
i
ts
.
She
ca
n
be
con
tacte
d
at email:
shanmathib@j
nn
.
e
du.
in
.
Par
app
urathu
Bah
ulayan
Smi
tha
is
an
as
socia
t
e
profe
ss
or
in
Depa
r
tm
en
t
o
f
El
e
ct
roni
cs
and
Comm
unicati
on
Engi
ne
eri
ng
.
,
J.
N.N
Instit
ut
e
of
Engi
ne
eri
ng
,
K
anni
ga
ipa
ir
,
Thi
ruva
lur
-
6011
02,
Tamil
n
adu,
India
.
She
co
mp
le
t
ed
ba
chelor
of
e
ngineeri
n
g
degr
e
e
in
el
e
ct
roni
cs
and
c
omm
unication
fr
om
Per
ia
y
ar
Ma
nia
mmai
Col
le
g
e
of
Technol
ogy
for
Wom
en
,
Vall
a
m,
Tha
nj
av
ur,
T
am
il
Nadu
aff
iliated
to
Bh
a
rat
hid
asa
n
Univ
ersit
y.
She
com
p
le
t
ed
ma
st
er
of
eng
ine
er
ing
in
e
le
c
tronics
and
con
trol
f
rom
Sathy
abam
a
Inst
it
ut
e
of
Scie
nc
e
and
Te
chno
logy,
Sat
hyaba
m
a
Dee
m
ed
Univer
sity
,
Chenna
i
in
200
5.
Curre
nt
ly
she
is
pursuing
Ph.D.
at Sat
hyab
am
a
Inst
it
ut
e
of S
ci
ence
an
d
Tec
hnology,
Chenn
ai
.
Her
r
ese
arc
h
are
a
includes
cybe
r
physi
cal
s
ystem
s
and
distr
ibut
ed
cont
ro
l
sy
stem
s
.
Her
a
r
ea
of
intere
st
are
cy
ber
physic
al
sys
te
ms,
mi
cro
p
roc
essors
and
m
ic
roc
on
trol
l
ers,
ant
enn
as
and
w
ave
propa
g
at
ion
,
microwave
engi
ne
eri
ng,
co
mm
unica
ti
o
n
sys
te
ms, im
age
pro
ce
ss
ing
,
cont
rol
sys
te
ms.
She
c
an
be
con
tacte
d
at
em
a
il
:
smi
thapb@
jnn.
edu
.
in
.
Sh
ali
ni
Ch
owdar
y
re
cei
ved
a
ba
che
lor'
s
degr
e
e
B.
E
i
n
e
lectr
oni
cs
an
d
com
munica
ti
on
engi
ne
eri
ng
in
2
008
from
Anna
Univer
sity,
a
mas
te
r's
degr
e
e
M.
E
in
applied
el
e
ct
ron
i
cs
in
20
11
from
Anna
Univer
sity
.
She
is
cur
ren
t
ly
workin
g
as
an
assistant
profe
ss
or
in
the
Dep
artme
nt
of
El
e
ct
roni
cs
and
Com
muni
c
a
ti
on,
T
.
J.S
Eng
i
nee
ring
C
ollege
,
Peruvoya
l
,
Ta
milnadu
,
Indi
a.
She
has
mor
e tha
n
13
y
ea
rs of
te
a
chi
ng
expe
ri
e
nce
Curr
ent
ly
she
is pur
suing
Ph.D.
at
Save
e
tha
Univ
ersit
y,
Chenna
i
.
Are
a
of
her
rese
ar
ch
is
appl
i
ca
t
io
n
of
image
proc
essing.
are
a
of
int
er
est:
dig
ital
c
irc
ui
ts
,
i
ma
g
e
and
signa
l
pro
ce
ss
in
g.
She
ca
n
be
contac
te
d
at
em
a
il
:
l
aksha
shali
ni@g
ma
i
l.co
m
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
omp E
ng,
V
ol.
15
, No
.
1
,
Febr
uary
20
25
:
425
-
434
434
Th
amiz
harasan
Pann
ee
rse
lva
m
com
p
le
t
e
d
bac
h
el
or
of
e
ngine
er
ing
degr
ee
i
n
el
e
ct
roni
cs
and
com
munica
ti
on
from
Maha
ra
ja
Engi
ne
eri
ng
Col
le
ge
,
Coim
b
at
or
e
aff
i
li
a
te
d
to
Anna
Univer
si
t
y
in
2006.
H
e
com
pl
et
ed
ma
ster
of
eng
ine
er
ing
in
co
mput
er
and
com
munica
ti
on
from
Sona
Coll
ege
of
T
ec
hno
l
ogy,
Salem
aff
i
li
ated
to
Anna
Univer
sity
Chenna
i
in
200
8.
His
ar
ea
s
of
int
er
est
in
cl
ude
wire
le
ss
co
mm
u
nic
a
ti
on,
signal
proc
essing,
inform
a
ti
on
the
o
ry
and codi
ng
.
H
e
c
an
b
e cont
a
cted
a
t em
a
il
:
pan
nee
rkt@gm
ail.
c
om
.
Rajagop
alan
Srinath
obt
ai
n
e
d
his
B
.
E
.
d
egr
e
e
fro
m
Anna
Un
ive
rsity
,
Ch
enna
i
,
Ta
milnadu
,
Indi
a
in
Apr
il
200
5
and
his
M.E
degr
ee
in
app
li
ed
e
lectr
oni
cs
from
Anna
Univer
sity,
Ch
e
nnai
,
Tamil
n
adu
,
India
in
May
2
007
and
com
p
leted
his
Ph.D.
in
inform
a
ti
on
and
communica
ti
on
eng
ine
e
ring
from
Anna
Un
ive
rsity
,
Chenn
a
i
in
th
e
ye
ar
2
023.
He
is
cur
ren
t
ly
worki
ng
as
an
assistant
prof
essor
in
the
Depa
r
tm
ent
of
Elec
t
ronic
s
and
Comm
unicati
on
Engi
ne
eri
ng,
SR
M
Instit
ute
of
Sc
ie
nc
e
and
T
ec
hn
ology,
Chenn
ai,
Ta
milnadu
,
India
.
His
a
rea
s
of
intere
st
includ
e
dig
it
a
l
sign
al
p
roc
essing,
digita
l
i
ma
ge
proc
essing,
a
rti
f
ic
i
al
int
ellige
n
ce,
neu
ral
net
works
an
d
fuz
zy
logic
.
He
has
publi
she
d
16
articles
in
th
e
r
eputed
Inte
rna
ti
ona
l
Jou
rna
ls,
10
ar
ti
c
le
s
in
the
In
te
rn
at
io
nal
Conf
ere
n
ce
s.
In
his
teac
hing
profe
s
sion
,
he
has
a
v
ast
ex
per
ie
n
ce
o
f
over
17
ye
ars.
He
ha
s
handl
ed
differe
nt
subje
ct
s
for
u
nder
gra
dua
te
and
postgraduat
e
student
s
fro
m
the
EC
E
and
C
SE
strea
ms.
He
has
bee
n
the
Co
ordina
tor
for
Nati
ona
l
Bo
ard
of
a
cc
r
edi
t
at
ion
.
He
has
orga
n
ized
and
attend
ed
work
shops
in
t
he
f
ie
lds
of
signal
and
i
ma
g
e
proc
essing
and
adva
n
ce
d
communicati
on
sys
te
ms.
He
has
a
lso
pla
yed
a
vital
role
in
condu
c
ti
ng
and
coor
d
ina
ti
ng
v
ari
ous
Nat
iona
l
le
v
el
T
ec
hn
ic
a
l
Sympos
ia
and
Confer
ences.
He
is
a
li
fe
m
e
mbe
r
of
ISTE
and
I
ET
E
.
He
ca
n
be
con
tacted
a
t
email:
drsrina
thraja
gop
al
an@gm
ai
l
.
co
m.
Muthu
raj
Mar
i
selvam
is
an
assistant
prof
essor
in
the
D
epa
rt
me
nt
of
E
lectr
on
i
c
s
and
Co
mm
uni
c
at
ion
a
t
J.N.N
Instit
ute
of
En
gine
er
ing,
Kann
iga
ip
ai
r,
Thi
ruv
al
ur
-
601102,
Ta
milnadu
,
Ind
ia
.
He
com
pl
e
te
d
b
ac
he
lor
o
f
engi
n
ee
ring
degr
ee
in
elec
troni
cs
and
com
munica
ti
on
from
Sre
e
Sow
dam
bik
a
Col
le
g
e
of
Engi
n
ee
r
in
g,
Aruppukot
ta
i
aff
i
liate
d
to
Anna
Univer
sit
y
in
2009
.
He
com
pl
et
ed
ma
st
er
of
engi
n
ee
r
in
g
in
VLSI
desi
gn
from
Sri
Venka
te
sw
ara
C
oll
ege
of
En
gine
er
ing
and
Te
chno
logy,
Tirupac
hur
aff
ilia
te
d
to
Anna
Univer
sity,
and
Chenna
i
in
201
3.
His
ar
ea
s
of
i
nte
rest
include
digi
tal
ci
rc
u
it
s,
VLSI
design
,
im
ag
e
and
sig
nal
proc
essing
,
and
low
pow
er
VLSI
.
He
ca
n
be
contac
t
ed
at
email:
ma
rise
lva
m
.
ms
@gma
il.c
o
m
.
Mohanan
Mural
i
obtained
h
is
B.
Te
ch
.
degr
e
e
in
elec
tron
ic
s
and
com
mun
icati
o
n
engi
ne
eri
ng
fro
m
Dr.
M
.
G.R.
U
nive
rsity
Chenn
a
i
and
M.E
.
d
egr
e
e
in
me
d
ical
elec
tro
nic
s
from
Coll
ege
of
Eng
i
nee
ring
,
Guindy
ca
mpus
,
Anna
Univer
sity
,
Ch
enn
ai
-
600
025
.
He
i
s
working
as
assistant
profe
s
sor
in
the
Dep
art
m
ent
of
Bio
me
di
ca
l
Eng
ineeri
ng
of
J.N.N.
Insti
tut
e
of
Engi
ne
eri
ng,
Ka
nniga
ip
ai
r,
Th
iru
val
lur
.
His
f
ie
ld
of
intere
st
inclu
des
signa
l
proc
e
ss
i
ng,
imag
e
proc
essing
,
m
ed
ic
a
l
elec
t
ronic
s
,
wire
le
ss
sensor
net
work,
wir
eless
net
work
an
d
int
egr
at
ed
el
e
ct
roni
cs
.
He
has
at
t
ende
d
nu
mbe
r
of
se
mi
n
a
rs,
short
-
te
r
m
c
ourse,
Summ
er
Schools
and
Confer
ences.
He
publi
shed
13
pa
per
s
in
Nat
iona
l
Confer
ence
proc
ee
ding
s.
He
is
p
ubli
shed
20
pape
rs
in
Inte
rn
at
ion
al
and
UG
C
Journals
proc
ee
dings.
He
is
t
he
Life
ti
m
e
m
e
mbe
r
of
ISTE
(India
).
He
is
working
as
a
ca
de
mi
c
coor
d
i
nat
or
in
J.N.N
.
Instit
u
te
of
Engi
ne
eri
ng,
Kanniga
ip
ai
r,
a
nd
Th
iruva
l
lur
Distric
t
-
601
102
.
He
is
gu
ide
d
3
5
U.G.
proj
ect
s
and
3
P.G
proje
c
ts.
He
c
an be
con
tacte
d
a
t
e
ma
il:
mura
l
im
oh
ana
n@gma
il.c
o
m
.
Evaluation Warning : The document was created with Spire.PDF for Python.