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.
677
~
688
IS
S
N:
20
88
-
8708
, DO
I: 10
.11
591/ij
ece.v
15
i
1
.
pp
677
-
688
677
Journ
al h
om
e
page
:
http:
//
ij
ece.i
aesc
or
e.c
om
Wireles
s
sensor n
etworks
based eff
icient d
rip i
rr
i
gation
monito
ring s
yste
ms
Ka
r
th
ik
Sag
ar
A
sh
ok
1
, Bas
avar
aj G
angasamu
dra N
ag
endrapp
a
2
, M
ohan
B
anga
l
or
e A
n
j
an
ey
alu
2
,
Pri
ya Nan
dih
al
3
, Veen
a Narayan
a Red
dy
2
, Liy
ak
athu
nisa
Syed
4
1
D
e
p
a
r
t
m
e
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
,
M
a
n
i
p
a
l
I
n
s
t
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
,
M
a
n
i
p
a
l
A
c
a
d
e
m
y
o
f
H
i
g
h
e
r
E
d
u
c
a
t
i
o
n
,
M
a
n
i
p
a
l
,
I
n
d
i
a
2
Dep
artm
en
t of
I
n
f
o
rm
atio
n
Science
,
BMS I
n
stitu
te of
T
echn
o
lo
g
y
and
M
an
ag
em
en
t,
Ben
g
al
u
ru, I
n
d
ia
3
Dep
artm
en
t of
Co
m
p
u
ter
Sci
en
ce, D
ay
an
an
d
a Sagar A
cademy o
f
Techn
o
lo
g
y
and
M
an
ag
em
en
t,
Ben
g
alu
ru,
I
n
d
ia
4
Dep
artm
en
t of
Co
m
p
u
ter
Sci
en
ce,
C
o
lleg
e of Co
m
p
u
ter
Scien
ce
,
T
aib
ah
Univ
ersity
,
Madin
ah
,
Sau
d
i Ar
ab
ia
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
J
ul
3,
2024
Re
vised
A
ug 22, 2
024
Accepte
d
Se
p 3, 2
024
Cott
on
has
profo
und
signifi
c
ance
in
the
te
x
ti
l
e
in
dustry
due
to
i
ts
ver
satilit
y
,
com
fort
and
ea
s
e
of
ca
r
e.
But
t
he
m
ai
n
prob
lem
with
conve
n
tional
cot
ton
far
mi
ng
is
that
i
t
uses
more
w
ater.
The
se
issues
are
m
ade
more
diffi
cu
lt
by
conve
nt
iona
l
irr
iga
ti
on
me
t
hods
,
su
ch
as
drip
irri
ga
ti
on.
To
a
ddre
ss
thi
s
proble
m
res
ea
rc
her
s
ar
e
using
tr
adi
ti
on
al
f
arm
in
g
t
ec
hniqu
es
wi
t
h
adv
anced
wire
le
ss
sensor
net
work
(
WSN)
protoc
ols
to
resolve
cata
str
ophic
issues,
such
as
pip
e
bur
sts
or
bloc
k
ed
e
mi
tters
which
ar
e
de
te
c
te
d
e
ar
ly
to
save
the
wate
r.
Th
is
pap
e
r
int
rodu
ce
s
eff
i
ci
en
t
WSN
arc
h
i
te
c
ture
using
pr
i
orit
y
-
base
d
dire
c
te
d
informa
ti
on
shar
ing
(DIS
)
protoc
ol
for
e
ffic
i
ent
utilizati
o
n
of
w
at
er
.
The
propos
ed
ar
chi
t
ec
tur
e
was
i
mpl
ement
ed
usi
ng
Ti
nyOS
sens
or
net
work
(TOSS
IM)
simu
l
at
ors.
Except
io
nal
qua
li
ty
of
service
(QoS
)
is
ac
hi
eved
using
new
rout
ing
protoc
o
l
e
xcl
usivel
y
for
ca
t
astrophi
c
failure
s.
Th
e
proposed
ar
chi
t
e
ct
ure
is
com
p
ared
with
st
anda
rd
protoc
ols
such
a
s
topol
ogy
geogr
aphic
gre
edy
forwardi
ng
(TPGF
)
,
li
nk
ca
rr
ie
r
sens
e
avoi
d
anc
e
(Li
nkCS
A)
an
d
t
iny
ca
rr
ie
r
sense
avo
idance
(Ti
nyCS
A)
.
Due
to
im
plementat
ion
opt
im
i
ze
d
prio
rit
y,
DIS
la
t
ency
has
be
en
red
uce
d
from
11.
3%
to
11
.
02
%
and
pa
cket
d
el
iv
ery
ra
ti
o
(P
DR)
is
enha
nc
e
d
by
35%
to
78%
conc
e
rnin
g
benc
hm
ark
p
rotoc
o
ls.
The
e
xper
imental
res
ult
s
prove
s
dra
stic
im
prove
me
nt
in
PD
R
a
nd
del
ay
p
erf
or
ma
nc
e
as
com
p
are
d
to
the
exi
sting
WSN pr
otoc
ol
.
Ke
yw
or
d
s
:
Link ca
r
rier se
ns
e a
voida
nce
Pr
eci
sio
n
a
gr
ic
ultur
e
Pr
io
rity
directe
d
in
f
or
mati
on
sh
ari
ng
Tiny ca
r
rier se
ns
e a
voida
nce
To
po
l
ogy pr
e
s
erv
i
ng greed
y
forw
a
r
ding
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
:
Kar
t
hik
Saga
r Asho
k
Dep
a
rtme
nt of
Com
pu
te
r
Scie
nce
a
nd
E
ng
i
ne
erin
g,
M
a
nip
a
l In
sti
tute
of T
echnolo
gy Be
ngal
uru
, Mani
pa
l
Acad
e
my o
f H
igh
e
r
E
du
cat
i
on
M
a
nip
al
, In
dia
Emai
l:
k
art
hik
.
sa@ma
nip
al
.e
du
1.
INTROD
U
CTION
Grow
i
ng
c
otto
n
is
a
major
c
ontrib
utor
t
o
t
he
w
or
l
d
’
s
te
xtil
e
in
du
st
ry.
C
otton
cr
ops
have
hi
gh
wate
r
dema
nd,
m
os
t
areas
are
sca
rc
e
du
e
t
o
de
for
est
at
ion
,
a
nd
poll
ution
le
a
ding
to
s
udde
n
ch
ang
e
i
n
the
cl
imat
e.
Pr
eci
sio
n
a
gri
culture
has
bec
om
e
an
i
nnova
ti
ve
met
hod
in
the
fiel
d
of
a
gri
cultural
pr
act
ic
es
in
the
pa
st
yea
rs
and
in
the
fu
t
ur
e
.
T
he
pr
im
ary
obje
ct
ive
i
s
to
monit
or
t
he
la
nd
’
s
physi
cal
and
histo
r
ic
al
char
act
eri
sti
cs
to
maximize
cr
op
o
ut
pu
t
wh
il
e preser
ving the e
nv
i
ronme
nt,
a
nd en
e
r
gy
[1]
. A
s a so
luti
on to
t
his,
wi
reless se
ns
or
netw
orks
hav
e
been
us
e
d
in
the
agr
ic
ulture
industr
y
[
2]
.
In
the
c
onve
ntion
al
a
ppr
oa
ch,
se
nsor
node
s
are
dep
l
oy
e
d
un
de
rgrou
nd
to
m
onit
or
pa
ramete
rs
li
ke
so
il
m
oistur
e
,
water,
an
d
miner
al
c
on
te
nt,
a
nd
th
ey
a
re
li
nk
ed
to
a
w
i
r
el
ess trans
cei
ve
r
th
rou
gh ph
ysi
cal
cables
[
3]
, [
4]
.
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
:
677
-
688
678
Nev
e
rtheless
,
e
ns
uri
ng
dep
e
ndable
wireless
commu
nicat
io
n
in
un
derg
rou
nd
e
nviro
nme
nt
s
remains
a
recent
subje
ct
of
ex
pl
or
at
io
n
for
resea
rch
e
rs
in
the
fiel
d
of
preci
sio
n
a
gri
culture
.
T
o
m
onit
or
fa
rmla
nd,
the
mo
st
ef
fecti
ve
irrigati
on
stra
te
gy
f
or
water
con
se
r
vation
is
the
dr
i
p
irri
gation
s
ys
te
m
,
par
ti
cularl
y
wh
e
n
com
bin
e
d
with
wireless
sens
or
net
wor
ks
(WSNs
)
.
In
ad
diti
on
to
t
his,
a
pote
ntial
malf
unct
ion
i
ng
m
onit
or
in
g
sy
ste
m
has
be
en
inc
orp
or
at
e
d
f
or
ea
rly
det
ect
ion
of
cat
as
tro
ph
ic
fail
ur
e
s.
The
main
c
ha
ll
eng
e
s
are
getti
ng
accurate
data
on
ti
me,
re
duci
ng
har
m
to
the
en
vir
onment
,
set
ti
ng
up
r
el
ia
ble
way
s
to
s
ha
re
in
f
or
m
at
io
n,
an
d
deali
ng
with
s
ys
te
m
br
ea
kdowns
.
T
he
mai
n
goal
s
are
t
o
se
t
up
fast
r
eal
-
ti
me
co
mm
un
ic
at
ion
a
nd
to
te
st
how
well
it
w
or
ks
t
hro
ugh
sim
ulati
on
s
[
5]
.
This
researc
h
work
ai
ms
to
imp
rove
a
pr
i
ori
ty
-
base
d
D
IS
protoc
ol
withi
n
a
wi
reless
se
ns
or
net
work
to
en
ha
nce
wa
te
r
us
e
ef
fici
ency
in
c
otto
n
far
mi
ng.
B
y
opti
mizi
ng
the
protoc
ol,
it
will
detect
an
d
a
ddres
s
major
ir
rigati
on
sy
ste
m
br
e
ak
dow
ns
,
e
ns
uri
ng
faster
respo
nse
s
a
nd
more
r
el
ia
ble
pac
ket
deliver
y
co
mpa
red
to
tradit
ion
al
WSN
meth
ods.
T
he
e
nh
a
nce
d
pr
oto
c
ol
fo
c
us
es
on
mi
nimizi
ng
wate
r
sup
ply
disruptio
ns,
re
du
ci
ng
crop
stress
a
nd
bo
os
ti
ng
pro
du
ct
ivit
y.
P
rio
riti
zi
ng
crit
ic
al
data
impro
ves
commu
nicat
ion
sp
ee
d
a
nd
s
ucces
s
rates,
al
l
ow
i
ng
f
or
bette
r
i
rr
i
ga
ti
on
ma
nag
e
ment,
ef
fici
ent
water
us
a
ge,
a
nd
inc
rease
d
operati
onal
resil
ie
nce
i
n
cotton fa
rmi
ng.
This
a
rtic
le
is
orga
nized
a
s
f
ollows:
sect
io
n
2
disc
us
ses
rel
at
ed
work
design
e
d
f
or
ef
fici
ent
ir
rigati
on
sy
ste
ms
.
Sect
ion
3
disc
us
ses
dri
p
ir
rigati
on
s
ys
te
m
desi
gn
an
d
pr
io
rity
-
ba
sed
r
ou
ti
ng
prot
oco
l.
In
sect
ion
4,
the
re
su
lt
s
of
our
pro
posed
scheme
th
rou
gh
sim
ulati
ons
and
e
xp
e
rime
nt
s
are
prese
nted.
I
n
sect
i
on
5,
we
con
cl
ud
e
the
work a
nd
prov
i
de
futur
e
p
e
rs
pe
ct
ives.
2.
RELATE
D
W
ORK
This
sect
ion
e
xp
l
or
es
th
e
la
t
est
resea
rc
h
on
issue
s
a
nd
c
ha
ll
eng
es
in
aut
om
at
ed
ir
rigati
on
s
ys
te
ms
,
fo
c
us
in
g
on
s
ys
te
m
i
neffici
encies
a
nd
te
ch
no
l
og
ic
al
li
mit
at
ion
s.
Ke
y
c
ha
ll
eng
es
i
nclu
de
unreli
able
sens
or
data, comm
un
i
cat
ion
d
el
a
ys
, a
nd
s
ys
te
m
br
e
akdo
wn
s t
hat di
srupt w
at
er
distribu
ti
on. W
hile po
te
ntial
so
l
ution
s
li
ke
imp
r
ov
e
d
sens
or
accu
rac
y
a
nd
fau
lt
det
ect
ion
al
gorith
ms
e
xist,
th
ey
hav
e
dr
a
wb
ac
ks
su
c
h
as
high
costs
and
c
omplex
impleme
ntati
on.
A
ddressi
ng
t
hese
c
halle
ng
e
s
is
cru
ci
al
f
or
en
ha
ncin
g
the
ef
fecti
ve
ne
ss
of
automate
d
ir
rig
at
ion
in
preci
si
on agric
ultu
re.
The
spe
ci
al
iz
e
d
li
te
ratu
re
ha
s
not
pro
pose
d
util
iz
ing
WSN
with
a
pr
i
or
it
y
-
base
d
r
ou
ti
ng
prot
oco
l
to
monit
or
t
he
m
al
functi
on
of
dr
i
p
irri
gatio
n
sy
ste
ms
.
T
he
fo
ll
owin
g
sec
ti
on
discuss
e
s
the
work
rela
te
d
to
irrigati
on
s
ys
t
em
c
on
tr
ol
[
6],
[
7]
.
The
stu
dies
[8],
[
9]
pro
posed
a
n
e
ne
rgy
-
savi
ng
te
chn
i
qu
e
f
or
w
irel
ess
sens
or
tra
nsmi
ssion
i
n
sma
rt
agr
ic
ultur
al
ir
r
igati
on
s
ys
te
m
s.
This
te
c
hn
i
que
le
ssen
s
ra
dio
inter
fer
e
nce
wh
il
e
conser
ving
e
ne
rgy
for
the
no
de
.
A
nothe
r
stu
dy
pro
pose
d
a
com
pu
te
rized
i
rr
igati
on
sche
me
cente
red
on
WSN
te
chnolo
gy
t
o
imp
rove
water
us
a
ge
in a
gr
ic
ul
ture
[
10]
. T
he met
hod use
d
two
se
nsor
s to gathe
r
inf
orma
ti
on
o
n
water
co
nte
nt
and
s
oil
te
mp
e
ratur
e
i
n
the
ar
ea
wh
e
re
the
pl
ant
roots
are
l
ocated.
To
gat
her
se
nsor
data
,
tur
n
on
act
uato
rs,
a
nd
delive
r
the
data
to
an
on
li
ne
ap
p,
a
gate
way
was
us
ed
.
To
re
gula
te
the
flow
of
wat
e
r,
the
sci
entist
s
then
pro
gr
a
mme
d
a
sy
ste
m
int
o
an
em
be
dded
sy
ste
m
a
nd
es
ta
blished
preci
se
te
mp
e
ratu
re
s
an
d
mo
ist
ure in
the
so
il
th
res
ho
l
d values
.
The
c
ontrib
ution
in
st
udy
[
11]
is
to
t
he
de
sign
an
d
im
ple
mentat
io
n
of
a
low
-
c
os
t
ene
r
gy
-
ef
fici
en
t
irrigati
on
man
ageme
nt
s
ys
te
m
c
ombinin
g
meas
ur
i
ng
de
vices
a
nd
ac
tuators
within
a
WSNs.
I
n
their
con
cl
ud
i
ng
obs
erv
at
io
ns
,
t
he
auth
or
s
st
ress
the
imp
or
ta
nce
of
placi
ng
sens
or
nodes
in
farms
an
d
sug
ges
t
that
reducin
g
t
he
se
par
at
io
n
a
mon
g
th
e
se
nsor
no
des
is
esse
ntial
f
or
en
ha
ncin
g
sy
ste
m
ef
ficacy.
H
ow
e
ve
r,
t
he
fact
that
the
res
ear
ch
us
e
d
just
fi
ve
sens
ors
is
a
majo
r
draw
ba
ck
of
t
he
w
ork
.
Kumar
et
al.
[12]
ai
m
to
pr
opos
e
a
com
pr
e
he
ns
ive
irri
gatio
n
so
l
ut
ion
cat
ered
t
o
pr
ese
nt
re
qu
i
re
ments
wi
thi
n
the
bac
kdr
op
of
ass
ociat
ed
res
earch
init
ia
ti
ves.
The
key
iss
ue
rais
ed
to
day
is
how
to
create
a
f
ully
aut
onomo
us
irri
gatio
n
s
ys
te
m
that
op
ti
mize
s
water
wa
sta
ge
wh
il
e
bei
ng
finan
ci
al
ly
f
easi
ble.
Sa
kth
ivel
et
al.
[13]
doe
s
no
t
del
ve
int
o
t
he
spe
ci
fics
of
how
mu
c
h
e
ne
rgy
the
a
utomat
ed
i
rr
igati
on
s
ys
te
m
mi
gh
t
co
nsu
me
a
fact
or
,
for
s
us
ta
ina
bili
ty
an
d
co
st
ef
fici
ency
.
M
ore
ov
e
r,
t
here
is
no
me
ntio
n
of
the
iss
ues
or
facto
rs
[
14]
to
co
ns
i
der
wh
e
n
it
come
s
to
mainta
ini
ng
an
d
cal
ibrati
ng
t
he
se
ns
ors
in
th
e
se
ns
or
ne
tw
orks
cr
ucial
f
or
pr
eci
se
data
gather
i
ng
a
nd
decisi
on
-
ma
king.
It
bar
el
y
e
xp
la
in
s
how to e
nhanc
e the s
ys
te
m t
o co
ver bi
gg
e
r
a
gr
ic
ultur
al
a
rea
s or
diff
e
re
nt c
rop
t
yp
es
[
15]
.
M
iya
et
al.
[
16]
is
s
hallow
in
it
s
re
view
of
the
possibl
e
dow
ns
i
des
a
nd
di
ff
ic
ulti
es
that
w
ould
inh
e
ren
tl
y
ste
m
f
rom
us
in
g
wireless
ga
dg
e
ts
for
water
s
uper
visio
n
pur
poses
a
mon
g
a
gri
cultural
ecos
ys
te
m
s.
The
pap
e
r
do
e
s
not
pro
vid
e
i
ns
ig
ht
into
t
he
scal
abili
ty
of
t
he
wate
r
m
onit
or
i
ng
s
ys
te
m
be
ing
pro
po
s
ed
,
wh
ic
h
ho
l
ds
ce
rtai
n
r
eper
c
us
si
on
s
f
or
ho
w
it
can
be
pr
act
ic
al
ly
mana
ged
on
a
wide
r
scal
e.
K
um
a
r
et
al.
[17
]
deals
on
l
y
with
l
ong
range
a
nd
i
ntern
et
of
thi
ngs
(IoT)
i
n
aut
o
irrigati
on
s
ys
te
ms
sp
eci
fical
ly
wh
il
e
the
y
m
ention
no
t
hing
ab
out
oth
er
s
mart
t
echnolo
gies
th
at
can
be
co
mb
ine
d
to
e
nhance
t
he
sma
rt
agr
ic
ultur
e
sy
ste
ms
eff
ect
ivel
y.
A
lt
ho
ug
h
the
pa
per
sta
te
s
th
e
ne
cessi
ty
of
monit
ori
ng
the
e
nviro
nm
e
ntal
par
am
et
er
s
i
n
agr
ic
ultur
e
,
the
y
fail
to
dee
ply
discuss
the
po
te
ntial
cyb
e
rse
cur
it
y
t
hr
eat
s
a
nd
data
pr
i
vac
y
co
nce
rns
rela
te
d
to
smart
a
gr
ic
ultu
re
s
ys
te
ms
im
pl
emented
us
i
ng
IoT
dev
ic
es
a
nd
wi
reless
se
ns
or
netw
orks
wh
ic
h
a
re
sig
ni
ficant
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
Wi
rel
ess sen
s
or netw
ork
s
bas
ed
ef
fi
ci
ent d
ri
p
irri
gati
on
monitorin
g
sy
ste
m
s
(
Ka
rthik
Sagar As
hok
)
679
factors
to
c
on
s
ider
i
n
t
he
im
pl
ementat
ion.
T
he
ai
m
of
t
he
pa
per
[
18]
is
to
sh
ow
ho
w
WSN
a
nd
I
oT
are
bein
g
app
li
ed
in
ma
na
ging
water
ne
eds
on
farms,
bu
t
it
does
not
go
i
nto
a
ny
de
ta
il
s
about
dif
ficult
ie
s
enc
ou
ntere
d
wh
il
e
implem
e
nting
t
hem.
Th
e
researc
h
pa
pe
r
unde
rsc
or
e
s
sev
eral
a
dv
a
nt
ages
of
us
in
g
a
n
I
oT
-
base
d
s
yst
em
in
fa
rming
al
t
houg
h
it
fail
s
to
exten
sively
ev
al
uate
possible
pr
i
vac
y
iss
ues
an
d
c
yb
e
r
me
naces
t
hat
ma
y
ste
m
from
a
doptin
g Io
T
d
e
vices i
n a fiel
d
set
ti
ng.
M
ost
of
the
a
uthor
’
s
main
c
on
t
rib
ution
is
dev
el
op
i
ng
a
gro
undbrea
king
wireless
se
nsor
netw
ork
desig
n
that
inc
lud
es
a
pr
i
or
it
y
-
base
d
disast
er
inf
ormat
io
n
sh
ari
ng
(DIS
)
protoc
ol
to
boos
t
water
pro
duct
ivit
y
in
cotto
n
farmi
ng.
T
he
a
uth
or
ma
kes
a
big
im
pact
on
the
area
of
smart
a
gr
ic
ultur
e
by
ma
king
wate
r
mana
geme
nt
methods
bette
r
for
gr
ow
i
ng
c
otton.
H
oweve
r,
the
mentio
ne
d
w
orks
do
not
ad
dr
es
s
the
reli
able
so
luti
on
to
t
he
current
ir
rigat
ion
s
ys
te
m
fail
ur
es
a
nd
le
ss
s
cop
e
has
bee
n
giv
e
n
to
qual
it
y
of
se
rv
ic
e
(
QoS)
.
The
e
xtensi
ve
li
te
ratur
e
s
urvey
sho
ws
a
hi
gh
dema
nd
f
or
pri
ori
ty
-
base
d
protoc
ols
f
or
the
a
f
or
eme
ntion
e
d
pro
blems.
3.
PROP
OSE
D MET
HO
D
This
sect
io
n
de
scribes
a
ne
w
wireless
se
nsor
a
nd
act
uat
or
net
work
(
WSAN
)
m
odel
f
or
op
ti
mizi
ng
dr
i
p
ir
rigati
on sy
ste
ms
.
By
a
pplyin
g
i
mpro
ve
d
ca
rr
ie
r
se
nse
awa
re
-
mu
lt
ip
at
h
ge
ogra
phic
r
ou
ti
ng (
CS
A
-
M
GR
)
protoc
ol
c
har
a
ct
erist
ic
s
and
f
un
ct
io
nalit
y
to
the
pro
pose
d
model
an
d
le
ve
rag
i
ng
t
he
du
al
traff
ic
le
vels
c
on
ce
pt
to
boos
t
th
rou
ghput
a
nd
del
ay,
the
perfor
mance
of
the
pro
po
se
d
m
odel
is
im
pro
ved.
Ou
r
pro
pose
d
model
employs dual
traf
fic
le
vels
to en
s
ur
e reli
able d
at
a
delive
ry
in
dri
p
i
rr
igati
on
s
ys
te
ms.
Firs
t,
traff
ic
is generat
ed
by
se
nsors
li
ke
te
mp
e
ratu
re
a
nd
s
oil
se
ns
ors
.
T
he
seco
nd
is
bas
ed
on
pri
or
it
y
w
hich
is
ge
ner
at
e
d
by
pr
e
ssu
re
sens
or
s
,
w
hic
h
is
more
cr
ucia
l
an
d
nee
ds
im
mediat
e
at
te
nti
on
to
pr
e
ve
nt
wastage
of
water
a
nd
da
mage
to
the
crop.
To
im
pr
ov
e
the
perfor
mance
of
the
pro
po
se
d
s
ys
te
m,
feat
ur
es
an
d
f
unct
ion
al
it
y
of
the
CS
A
-
MGR
protoc
ol
ha
ve
been
co
ns
ide
re
d
by
util
iz
ing
t
he
c
on
ce
pt
of
du
al
tra
ff
ic
le
ve
ls
to
en
ha
nce
thr
oughput
an
d
delay
.
Figure
1
s
hows
the
flo
w diagr
am of t
he pr
opos
e
d
s
ys
te
m.
Figure
1.
Flo
w
d
ia
gram
of t
he
prop
os
ed
wo
r
k
3.1.
W
orkin
g pri
ncipl
e of th
e propos
ed sy
stem
The
first
ap
pr
oach
is
dri
p
ir
rigati
on,
w
hic
h
i
nvolv
e
s
t
he
util
iz
at
ion
of
pip
es
c
on
ta
ini
ng
emit
te
rs
to
deliver
water
directl
y
to
the
r
oo
ts
of
the
pl
ants.
T
his
sys
te
m
co
mprise
s
esse
ntial
el
em
ents
s
uc
h
a
s
a
wate
r
so
urce,
a
pri
m
ary
pi
peline,
a
nd
a
series
of
se
co
ndar
y
pi
pes
li
nke
d
t
o
the
mai
n
one
thr
ough
ma
nu
al
or
automate
d
val
ves.
T
he
sec
on
d
te
c
hn
i
qu
e
is
known
as
s
pr
in
kler
i
rr
i
gation,
wh
e
re
press
ur
i
zed
water
is
pu
mp
e
d
and
subse
quen
tl
y
directe
d
to
nozzl
es
that
disp
e
rse
water
into
t
he
ai
r
.
But
this
met
hod
is
le
ss
c
ompete
nt
because
of
the
wastage
of
w
at
er
owed
t
o
e
vapo
rati
on
a
nd
run
off
w
hich
is.
Conseq
ue
ntly,
the
dr
i
p
ir
rigati
on
method
e
mer
ge
s
as
the
s
up
e
ri
or
c
hoic
e
f
or
our
desig
n.
Fig
ure
2
show
s
the
dep
l
oyment
a
nd
data
flo
w
between
sens
or
node
s a
nd actuat
or
s
[
19]
.
The
m
odel
w
e
pro
po
s
e
em
p
lo
ys
a
cl
os
e
d
-
l
oop
met
hodo
l
ogy,
w
her
e
in
the
s
ys
te
m
co
ns
ist
entl
y
ob
s
er
ves
it
s
re
act
ion
s
a
nd
uti
li
zes
the
in
pu
t
to
enact
necess
ary
modific
at
ion
s
in
it
s
c
on
t
r
ol
mec
han
is
ms
.
This
model
is
cra
fte
d
with
a
fo
c
us
on
ta
il
or
e
d
ir
r
igati
on
f
or
s
pe
ci
fic
sit
es,
ena
bling
adj
us
tme
nts
in
c
r
op
wa
te
ring
that
account
f
or
both
te
m
pora
l
and
s
patia
l
con
si
der
at
io
ns.
The
pr
opos
e
d
desig
n
’
s
main
go
al
is
to
a
ddr
ess
the
sh
ort
co
min
gs
of
the
c
urre
nt
dr
i
p
irri
gation
method.
Va
riat
ion
s
in
so
il
ty
pe
,
crop
var
ie
ty
,
and
weathe
r
pa
tt
ern
s
are
t
he
ca
us
es
of
these
disc
rep
a
ncies,
al
ong
with
a
f
oc
us
on
ad
dressi
ng
the
iss
ues
associat
ed
with
dr
ip
irrigati
on
s
et
up
malfu
nctio
ns.
T
o
a
ddress
th
ese
chall
e
ng
es
,
it
bec
om
es
es
sentia
l
to
c
onti
nuously
m
on
it
or
the
water
fl
ow
rate
within
the
dri
p
irri
gatio
n
s
ys
te
m
an
d
at
the
same
ti
me
water
the
cr
ops
with
a
bala
nced
qu
a
ntit
y of wat
er
[
20]
, [2
1]
.
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
:
677
-
688
680
Figure
2. De
pl
oyment
of se
nsor
no
des
a
nd a
ct
uator n
et
wor
ks
3.2
.
P
ri
ority
-
ba
sed
D
IS
In
the
pr
opos
e
d
protoc
ol,
tw
o
pri
mar
y
sorts
of
traf
fic
a
re
pr
oduce
d
by
t
he
sens
or
s
.
Te
mpe
ratur
e
s
a
nd
so
il
mo
ist
ure
are
the
fi
rst
c
at
egory,
w
hich
we
de
scribe
as
typ
ic
al
traf
f
ic
becau
se
no
immediat
e
ac
ti
on
is
need
e
d.
T
he
ne
xt
kind,
wh
ic
h
is
obta
ine
d
f
rom
se
n
s
or
s
t
hat
measu
re
press
ur
e
,
is
cat
eg
ori
zed
as
pri
ori
ty
traff
i
c
since
an
y
dif
ficult
ie
s
that
are
disco
ve
red
m
us
t
be
res
olv
e
d
rig
ht
awa
y,
ei
t
her
by
cl
os
i
ng
the
pri
mar
y
va
lve
or
need
i
ng
huma
n
assist
ance
.
Wh
e
ne
ver
mu
l
ti
ple
traff
ic
ty
pe
s
are
ope
rati
ng
at
the
sam
e
t
ime,
it
is
obvio
us
t
hat
pr
i
or
it
y
traf
fic
fo
c
us
es
on
de
pe
nd
a
bili
ty
a
nd
punctuali
ty
ov
er
regular
tra
ffi
c
[22
]
.
As
a
re
su
lt
,
t
o
c
omple
te
the
qu
al
it
y
-
of
-
se
r
vi
ce
crit
eria
f
or
e
ach
tra
ff
ic
ty
pe
an
d
preve
nt
c
olli
sion
s
am
ong
man
y
dif
fer
e
nt
s
ources
of
tr
aff
ic
,
a
s
uitable
for
w
ard
i
ng
proce
dure
must
be
us
e
d.
I
n
this
w
ork
,
a
n
ex
planati
on
f
or
how
pat
hways
with
va
ri
ou
s
traff
ic
pr
i
or
it
ie
s mig
ht
be bu
il
t t
o
meet
t
hese
issues is
pro
vi
ded in
t
he
sect
i
on
s
that
fo
ll
ow
.
3.3. M
athem
ati
cal m
od
el
for priori
ty
ba
s
ed D
I
S p
r
otoc
ol
Pr
io
rity
-
base
d
DI
S
im
pro
ves
co
mm
un
ic
at
io
n
by
set
ti
ng
pri
or
it
y
le
vels
f
or
di
ff
e
ren
t
co
nd
it
io
ns
o
r
QoS
de
pendin
g
upon
it
s
re
quirements.
T
his
protoc
ol
pro
vid
es
i
nterf
e
re
nc
e
sup
pr
es
sio
n
f
rom
diff
e
re
nt
nodes
,
acce
ssing
of
s
pectr
um
,
dy
na
mica
ll
y
set
ti
ng
up
of
pr
i
or
it
y
an
d
gu
a
ran
te
e
d
QoS.
A
sim
plifie
d
mathe
mati
cal
model
f
or
m
ul
ti
path
pri
or
it
y
-
base
d
D
IS
ca
n
be
f
ormulat
ed
as
an
opti
miza
ti
on
pr
ob
le
m
.
Let
be
the
total
numb
e
r
of
node
s
in
the
sy
ste
m,
pi
is
the
pr
i
or
it
y
le
vel
f
or
the
no
de
w
he
r
e
=
1,
2,
3
,
…
,
and
the
range
of
pr
i
or
it
y
val
ue
is
[
1,
M]
w
her
e
is
the
maxim
um
pri
or
it
y.
,
,
an
d
represe
nt
inter
fer
e
nce
exp
e
rience
d,
t
he
re
quireme
nt
of
data
rate,
QoS
re
quireme
nt
an
d
a
vaila
ble
sp
ect
r
um
res
ources
of
t
he
it
h
no
de
resp
ect
ivel
y.
The
pro
pose
d
model
is
re
pr
e
sented
in
si
x
s
ta
ges
as
sig
nme
nt
of
pr
i
or
it
y,
inter
fer
e
nce
s
uppressi
on
,
al
locat
ion
of
r
eso
ur
ces
,
guar
anteed
Q
oS,
a
ll
ocati
on
of
da
ta
rate
a
nd
dyna
mic
a
dju
st
ment.
As
sig
nme
nt
of
pr
i
or
it
y
(
):
Pr
i
or
it
y
is
denote
d
by
pi
w
hich
is
deter
mine
d
base
d
on
re
quireme
nt
of
se
r
vi
ce
requeste
d
f
rom
the it
h node
. Set
ti
ng
up o
f pr
i
or
it
y i
s formul
at
ed
as a
fun
ct
i
on of
node
’
s
re
qu
i
reme
nt whi
ch
is
giv
e
n b
y
(1),
=
(
ℎ
)
(1)
The fu
nctio
n
ta
kes dif
fer
e
nt val
ues based
on fi
xe
d,
dyna
mic an
d user s
pecific p
rio
rity
=
{
(
)
ℎ
(
)
ℎ
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
Wi
rel
ess sen
s
or netw
ork
s
bas
ed
ef
fi
ci
ent d
ri
p
irri
gati
on
monitorin
g
sy
ste
m
s
(
Ka
rthik
Sagar As
hok
)
681
−
In
te
r
fer
e
nce
s
uppressi
on
(
):
This
ca
n
be
modele
d
us
i
ng
powe
r
-
c
ontr
ol
le
d
strat
eg
y.
Let
de
note
s
transmitt
ed
power
of
the
th
de
vice
a
nd
represents
c
hanne
l
gai
n
betwee
n
th
de
vice
a
nd
th
de
vice
s
o
that
(
2)
,
=
∑
whe
re
i
,
j
=
1
,
2
,
…
.
.
N
≠
(2)
Eq
uation
(2)
r
epr
ese
nts
inter
fer
e
nce
ex
per
i
enced
from
al
l
ot
her
nodes
to
th
node
as
a
s
um
of
pro
du
ct
s
of
their
tran
smit
te
d
po
wer
a
nd
channel
gai
n.
To
s
uppress
on
e
m
us
t
co
ntr
ol
the
po
wer
a
nd
a
djust
valu
es
dynamical
ly
w
it
h
ob
je
ct
ive
i
s
to
minimi
ze
the
powe
r.
Th
e
ob
je
ct
ive
functi
on
t
o
s
uppress
can
be
.
Algorith
m
1
,
i
mp
le
me
ntati
on of
pr
io
rity
DIS pro
t
oc
ol.
=
∑
=
1
(3)
su
bject
e
d
to
co
nd
it
io
n
≤
≤
w
her
e
are
maxim
um
and
minim
um
powe
r
of
t
he
dev
ic
e.
T
he
e
quat
ion
(
3)
min
imi
zes
the
tota
l
interfe
ren
ce
consu
med
by
al
l
nodes
with
res
pect
to
po
wer
le
vels.
−
Re
so
urce
al
l
oc
at
ion
(
λ
)
:
Allocat
ion
of
re
qu
ire
d
res
ources
in
pro
posed
m
odel
can
be
done
by
c
onside
rin
g
sp
ec
tr
um
of
re
so
urces
λ
to
ea
ch
no
de
by
f
ulfill
ing
Ω
at
dat
a
rate
.
T
he
w
ho
le
re
source
al
locat
io
n
process
ca
n
be
f
ormulat
ed
as
a
pro
blem
of
op
ti
miza
ti
on
s
ub
je
ct
ed
to
goal
of
maximizi
ng
s
ys
te
m
util
it
y
base
d
on
a
vaila
bili
ty
of
s
pec
trum
a
nd
node
’
s
re
qu
ireme
nt
,
util
it
y
of
eac
h
node
ca
n
be
ex
pr
esse
d
a
s
a
functi
on
of
it
s
al
locat
ed
s
pe
ct
ru
m
res
ourc
e,
sa
y
(
λ
).
Ne
xt
,
ai
m
is
t
o
maximize
t
he
al
locat
ion
of
resou
rces
us
in
g
(4),
λ
=
∑
(
λ
)
=
1
(4)
with a c
onstrai
nt
{
∑
λ
=
1
≤
λ
,
Ω
(
λ
)
≥
Ω
,
(
λ
)
≥
,
λ
≤
λ
i
≤
λ
max
},
∀
i=
1,
2, ……
…
., N.
Using
(
4)
prot
oco
l,
we
are
a
ble
t
o
al
locat
e
the
s
pectr
um
of
res
ources
to
al
l
nodes
s
o
th
at
overall
util
it
y
is
maximize
d.
−
Qu
al
it
y
of
ser
vice
(
Ω
):
Q
ualit
y
of
ser
vice
in
propose
d
m
od
el
is
ac
hie
ve
d
by
ma
ximizi
ng
t
he
util
it
i
es
der
i
ved
f
rom
a
ll
ocated
resou
r
ces.
T
he
pri
ma
ry
goal
of
the
model
is
t
o
ac
hi
eve
higher
pac
ket
deliver
y
rat
io
and
minim
um
delay
in
delive
ry
of
a
pack
et
.
To
achie
ve
th
e
go
al
le
t
us
c
on
si
der
Ω
is
the
qu
al
it
y
of
the
serv
ic
e
re
qu
ir
e
ment
of
ℎ
node wh
e
re
=
1, 2
,
…
,
. Gua
ra
nteed Qo
S ca
n be
for
mu
la
te
d as
(
5),
Ω
=
∑
(
)
=
1
(
5)
with a c
onstrai
nt
{
∑
≤
,
Ω
(
)
≥
Ω
,
(
)
≥
=
1
,
≤
≤
}
,
∀
=
1
,
2
,
3
…
.
Using
(5) pr
ot
oco
l ca
n
al
l
oca
te
the r
es
ource
s so that s
ys
te
m u
ti
li
ty is ma
ximize
d w
hile
mainta
in
Q
oS
.
−
Data
rate
al
loc
at
ion
(
):
Data
rate
al
locat
ion
in
t
he
pro
po
s
ed
a
rch
it
ect
ure
is
ac
hieve
d
by
maximi
zi
ng
sy
ste
m
t
hroug
hput
without
disturbin
g
us
e
r
re
quireme
nts
.
Let
denote
data
re
qu
i
rem
ent
of
ℎ
node
wh
e
re
=
1, 2, …
,
. Object
ive
f
unct
ion t
o ma
ximize
d
at
a
rate i
s g
i
ven by
(6),
=
∑
=
1
(6)
with a c
onstrai
nt
{
∑
≤
,
(
)
≥
,
=
1
≤
≤
}
,
∀
=
1
,
2
,
3
…
.
−
Dynamic
a
djust
ment:
dyna
mic
adjustme
nt
is
achieved
by
c
ha
ng
i
ng
data
rate
an
d
interfe
ren
ce
adap
ti
vel
y.
De
pendin
g
upon
t
he
co
ndit
ion
s
of
netw
ork
pr
i
or
it
y
sc
he
du
li
ng
w
il
l
be
done
.
This
is
achie
ve
d
by ma
ximizi
ng
the s
ys
te
m
pe
r
forma
nce
,
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
:
677
-
688
682
∑
(
)
=
1
(
7)
with a c
onstrai
nt
{
≤
,
(
)
≥
,
Ω
(
)
≥
Ω
,
≤
≤
}
,
∀
=
1
,
2
,
3
…
.
Fr
om
(7)
ov
e
r
al
l
sy
ste
m
perf
ormance
will
be
e
nh
a
nce
d
by
e
ns
uri
ng
t
he
interfe
re
nce
r
ang
e
is
withi
n
the
desire
d ran
ge.
Algorith
m l.
I
mp
le
me
ntati
on of
p
rio
rity
DIS pro
t
oc
ol
1
.
Definition:
HPQ: High Priority Queue
LPQ: Low Priority Queue
Node: A device in a network
MAT: Time of
message arrival to a node.
NSI: Information about length of queue, network states.
MsgInt: Message interval at nodes exchange the information.
2
.
Initialization:
Initialization of HPQ and LPQ at each node
Node initialization to keep track of queue length a
nd status of queue.
3
.
Message Arrival:
While network is in operation:
If message Priority
=
-
High
HPQ enqueue (message)
Else:
LPO enqueue (message)
Update NSI
4
.
Exchange of information:
At each interval:
Exchange NSI information with
neighboring
node.
Update NSI
5
.
Propagation of message
If HPQ is not empty:
Prioritize the sharing of high priority message from HPO
Else:
sharing of high priority message from LPQ
6
.
Dynamic Adjustment: Dynamically adjust the frequency of propagation messages based on
current load and queue length
If length HP)>>length (LPO):
Increase the frequency of high priority messages
If length (LPQ)>length (HPQ):
Increase the frequency of l
ow priority messages.
7
.
Congestion Control:
If congestion Detected
Signal the
neighbor
Adjust the propagation rate or reroute message
Update NSI
8
.
Stop
The
ab
ove
-
des
cribe
d
al
gorith
m
is
i
mp
le
me
nt
ed
us
in
g
the
T
OS
S
I
M
sim
ula
tor
to
m
od
el
th
e
be
ha
vior
of
t
he
p
rio
rity
DIS
pr
oto
c
ol
in
a
co
ntr
olled
en
vir
onme
nt.
TO
SSIM
al
l
ow
s
f
or
preci
se
te
sti
ng
of
m
essag
e
pr
i
or
it
iz
at
ion
,
qu
e
ue
ma
na
ge
ment,
a
nd
dyna
mic
a
dju
stme
nts
i
n
netw
ork
loa
d.
T
he
pro
tocol
’
s
perf
ormanc
e
unde
r
dif
fer
e
nt
congesti
on
s
cenari
os
is
ev
al
uated
by
s
i
mu
la
ti
ng
real
-
world
c
onditi
ons.
T
he
ex
pe
rimental
resu
lt
s,
wh
ic
h
will
be
detai
le
d
i
n
t
he
ne
xt
se
ct
ion
,
hi
ghli
gh
t
the
protoc
ol
’
s
eff
ic
ie
nc
y
in
ha
nd
li
ng
high
-
pri
or
it
y
message
s a
nd
con
t
ro
ll
in
g network
c
ongestio
n.
4.
EXPERI
MEN
TATION
AN
D RESULT
A
NA
L
YS
I
S
4.1.
E
xper
im
enta
l
s
etup
The
e
xperime
nt
wa
s
c
onduct
ed
us
i
ng
the
T
OS
S
I
M
simulat
or,
util
iz
ing
the
Tel
os
B
m
ote
config
ur
at
io
n
ou
tl
ine
d
i
n
Ta
ble
1.
This
set
up
e
ns
ures
a
c
cur
at
e
simulat
ion
of
netw
ork
be
ha
vio
r
unde
r
the
pro
po
se
d
prot
oco
l.
Fig
ure
3
visu
al
ly
il
l
us
trat
es
t
he
ne
twork
la
yout
,
hi
gh
li
ghti
ng
the
placeme
nt
an
d
config
ur
at
io
n o
f nodes
f
or
c
omp
reh
e
ns
ive
te
sti
ng
a
nd
performa
nce e
valuat
ion
.
The
netw
ork
to
po
l
ogy
i
nclu
de
s
tw
o
s
ources
(
mo
te
1
a
nd
2),
12
inter
mediat
e
m
otes,
a
nd
a
sing
le
si
nk
(m
ote
15).
Th
e
sink
is power
e
d
via
a
un
i
ver
s
al
serial
bu
s (
U
SB
)
cable
c
onne
ct
ed
to
a
mai
n
per
s
onal
com
pute
r
(
PC
)
,
an
d
al
l
mo
te
s
a
re
pr
ogra
mme
d
th
rough
th
e
U
S
B
port.
A
U
SB
hub
with
13
ports
is
us
e
d
to
com
pile
mu
lt
iple
mo
te
s
simult
ane
ou
sl
y
us
i
ng
a
s
hell
script.
A
n
a
ddit
ion
al
mo
te
se
rv
es
as
a
rem
ot
e
con
tr
ol
to
e
ns
ure
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
Wi
rel
ess sen
s
or netw
ork
s
bas
ed
ef
fi
ci
ent d
ri
p
irri
gati
on
monitorin
g
sy
ste
m
s
(
Ka
rthik
Sagar As
hok
)
683
sy
nc
hro
nized
s
ta
rtup
of
al
l
ne
twork
m
otes.
This
m
ote
is
pr
ogram
med
with
NE
SC
c
ode
t
o
broa
dcast
a
s
pecial
pack
et
at
high
transmissi
on
powe
r,
prom
pting
a
ny
receivi
ng
m
ote
to
perform
a
syst
em
reset.
De
buggi
ng
was
cond
ucted
on
t
he
ce
ntral
PC,
with
mo
te
s
se
nd
i
ng
de
bug
message
s
via
US
B
inte
rf
ace
s.
Alt
hough
th
is
set
up
represe
nts
a
s
mall
-
scal
e
sen
s
or
net
work,
t
he
ex
per
ime
nt
i
s
val
uab
le
as
it
de
monstrate
s
protoc
ol
beh
a
vi
or
i
n
a
real
ha
r
dw
a
re
env
i
ronme
nt.
The
TO
SSIM
simulat
or
retai
ns
t
he
phys
ic
al
an
d
li
nk
-
la
yer
pro
pe
rtie
s
of
WSNs
,
pro
vid
in
g
a
re
al
ist
ic
env
ir
on
ment
f
or
e
valuati
ng
var
i
ous
pr
oto
c
ols
a
nd
al
gorithm
s
[
18]
.
We
util
iz
e
the
Cutec
om
t
oo
l t
o
rec
ord
the te
stbed
’
s p
e
rfo
r
mance in
a trac
e file
, s
pecifica
ll
y
trackin
g
the
p
acket
deliver
y
rati
o
and
the
a
verag
e
pac
ket
dela
y
betwee
n
t
he
s
ource
a
nd
de
sti
nation
.
A
ddit
ion
al
ly
,
the
pro
tocol
al
lo
ws
f
or
t
he
accum
ulati
on
of
noise
,
w
hic
h
can
ei
ther
i
nc
rease
or
decr
e
ase
the
li
nk
gai
n.
T
o
a
ddress
t
his
issue
,
our
prot
oco
l
is
adj
us
te
d
as
fo
ll
ows
:
d
ur
i
ng
t
he
disc
ov
e
r
y
pe
rio
d,
eve
r
y
node
kee
ps
track
of
the
c
ount
of
hello
pack
et
s
receive
d
from
each
neig
hbor
within
it
s
inter
nal
me
mor
y.
T
his
is
acco
mp
li
sh
e
d
by
c
omp
uting
the
per
ce
ntage
diff
e
re
nce
between
the
total
num
ber
of
he
ll
o
messa
ges
that
a
re
r
ece
ived
by
the
t
ot
al
amo
un
t
t
ha
t
wa
s
antic
ipate
d.
T
he
s
ub
se
quent
hop
is
sel
ect
ed
thr
ough
ou
t
the
routin
g
proce
dure
dep
e
ndin
g
on
the
c
onnect
ion
’
s
trustwo
rthin
es
s
,
wit
h
a
predet
ermine
d
t
hr
es
hold.
I
n
t
he
sit
ua
ti
on
at
ha
nd,
70%
is
t
he
nec
essar
y
a
nd
s
uff
ic
ie
nt
per
c
e
ntage
.
Ti
ny
ca
rr
ie
r
sen
s
e
avo
i
dan
c
e
(T
iny
CS
A)
is
a
n
al
te
red
va
riant o
f
t
he
CSA
-
MGR
prot
oco
l
in
wh
ic
h
every sta
rting
node
j
us
t creat
es a sin
gle
path
way to t
he
e
ndpo
i
nt.
Table1
. Con
fig
ur
at
io
n para
me
te
rs
Sl.
No
.
Para
m
eter
Sp
ecifica
tio
n
1
MAC
lay
e
r
f
ra
m
e
wo
rk
IE
E
E
8
0
2
.15
.4
2
Path
lo
ss
exp
o
n
en
t
4
.7
3
Distan
ce
3
m
eters
4
Path
lo
ss
at
re
fer
en
ce dis
tan
ce
5
6
.4 d
B
5
Pack
et
s
ize
5
4
by
tes
6
Frequ
en
cy
2
.4 GHz
Figure
3. Ex
pe
rimental
to
polo
gy
4.1
.
R
esult
analy
si
s
In
simulat
in
g
wireless
c
omm
un
ic
at
io
n
s
ys
te
ms,
ke
y
par
am
et
ers
li
ke
dela
y
a
nd
la
te
ncy
are
crit
ic
al
,
especial
ly
f
or
app
li
cat
io
ns
su
c
h
as
irri
ga
ti
on
wi
reless
sens
or
netw
orks.
T
hese
fac
tors
determi
ne
how
eff
ic
ie
ntly
data
is transm
it
te
d. High d
el
a
ys
c
an
le
ad
to
i
neff
ic
ie
nt w
at
er ma
nag
e
ment a
nd sy
ste
m f
ai
lu
re
s
[23],
[24]
.
4.1.1.
Key
met
ri
cs
Av
e
ra
ge
delay
:
cal
culat
ed
as
the
a
ver
a
ge
ti
me
ta
ken
for
pack
et
s
to
tra
ve
l
f
rom
t
he
source
to
t
he
destinat
io
n.
Pa
cket
delive
ry
r
at
io
(
PD
R
):
T
he
rati
o
of
s
u
c
cessf
ully
deliv
ered
pack
et
s
t
o
t
he
total
nu
mb
e
r
of
pack
et
s
sent.
T
he
propose
d
protoc
ol
ex
hib
it
e
d
lo
wer
ave
rage
delays
co
mpa
red
t
o
oth
e
r
prot
oco
ls,
in
dicat
in
g
eff
ic
ie
nt
data
trans
missi
on
an
d
reduce
d
la
te
ncy
.
T
his
fin
ding
s
up
ports
the
obj
ect
ive
of
e
nhancin
g
commu
nicat
io
n
e
ff
ic
ie
nc
y
i
n WS
Ns.
The
ave
rag
e
d
el
a
y
is
wr
it
te
n
as
(
9),
_
=
∑
=
1
(9)
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
:
677
-
688
684
w
he
re
is
the
total
numb
e
r
of
pack
et
s
receiv
ed
an
d
do
e
s
ℎ
pack
et
ex
per
ie
nc
e
the
dela
y.
It
is
no
t
eas
y
to
measu
re
di
rectl
y
but
can
be
est
imat
ed
us
i
ng
t
he
traf
fic
load
of
a
network
.
Packet
-
deliver
y
-
rati
o
can
be
cal
culat
ed
as
(
10)
.
=
(
)
×
100%
(10)
Table
2
co
mpa
re
s
a
ver
a
ge
de
la
y
pe
rce
ntag
es
betwee
n
C
SA
-
MGR
a
nd
the
p
rio
rity
DI
S
prot
oc
ol
acro
s
s
diff
e
re
nt
protoc
ols
li
ke
topol
ogy
ge
ographic
gr
ee
dy
f
orwardin
g
(T
P
GF
)
,
L
I
NK
A
war
e
,
Ti
ny
CS
A,
a
nd
Tiny
Hop,
m
easur
e
d
at
va
ry
i
ng
packet
s
per
sec
ond
(
PPS
)
rates.
T
he
res
ults
sho
w
that
p
rio
rity
D
IS
consi
ste
ntly
re
du
ce
s
delay
co
mp
a
red
to
C
S
A
-
MGR,
par
ti
cularly
in
TP
GF
,
w
he
re
delays
dro
p
si
gn
i
ficantl
y
from 20
% t
o
13% at 1
0
PPS and
from 3
2% t
o
26% at 2
1
P
PS.
Acr
os
s all
p
r
oto
c
ols,
p
rio
rity DIS d
e
m
onstrat
es
bette
r
dela
y
pe
rformance
,
im
pr
ovin
g
net
work
e
ff
ic
ie
nc
y.
The
P
DR
wa
s
measu
red
acr
oss
dif
fer
e
nt
pr
oto
c
ols
su
c
h
as
Tin
yCSA
,
LI
N
K
-
Aware
, a
nd TP
GF.
−
Tiny
CS
A
:
De
monstrate
d
s
uperi
or
PD
R
pe
rformance
,
e
ffec
ti
vely
a
vo
i
din
g
the
c
ar
rier
sense
e
ff
ect
.
T
his
su
ggest
s t
hat T
iny
CS
A
is
mor
e reli
able
for
a
pp
li
cat
io
ns
requirin
g hig
h dat
a integ
rity.
−
LINK
-
A
war
e
protoc
ol
:
s
how
ed
relat
ively
good
PD
R
pe
rfo
rma
nce,
c
onsid
erin
g
li
n
k
reli
abili
ty
duri
ng
pa
th
const
ru
ct
io
n. T
his in
dicat
es a
balance
d
a
ppr
oa
ch betwee
n re
li
abili
ty an
d pe
rformance
.
−
TPG
F
p
ro
t
oc
ol
:
disp
la
ye
d
lo
wer
PD
R
res
ul
ts
du
e
t
o
it
s
gr
ee
dy
f
orwa
r
ding
a
ppro
ac
h,
wh
ic
h
pr
i
or
it
iz
es
paths
closest t
o t
he
si
nk. T
his
highli
gh
ts
a
tra
de
-
off bet
wee
n path
opti
mali
ty a
nd r
el
ia
bili
ty.
−
Tiny
-
h
op
p
ro
t
oco
l:
Ha
d
the
lo
west
PD
R,
at
trib
uted
to
ack
nowle
dge
ment
an
d
retr
ansmissi
on
iss
ues,
emp
hasizi
ng
th
e
need
f
or
rob
us
t
ack
nowle
dgme
nt
mec
hani
sms
in
W
SNs.
Diff
e
re
nt
valu
es
obta
ined
fro
m
exp
e
rime
nt
is
dep
ic
te
d
in
Ta
ble
3
an
d
it
c
ompare
d
with
sta
nd
a
rd
sen
sor
net
wor
k
protoco
ls
li
ke
TP
GF
,
LINK
A
war
e
,
t
iny
cl
us
te
r
-
bas
ed
sel
f
-
organ
iz
at
ion
alg
ori
thm
, tiny h
op
-
base
d routi
ng
proto
col.
Table
2.
C
omp
ariso
n of ave
ra
ge dela
y vs
. PP
S w
it
h di
ff
e
rent
protoc
ols
Delay (%
)
in
CSA
-
MGR vs
Priority
DIS
PPS
TPGF
LI
NK
Awa
re
Tiny
CSA
Tiny
Hop
CSA
-
M
GR
Priority
DI
S
CSA
-
M
GR
Priority
DI
S
CSA
-
M
GR
Priority
DI
S
CSA
-
M
GR
Priority
DI
S
10
20
13
22
21
23
25
110
112
21
32
26
23
23
24
21
105
108
22
33
23
24
24
25
27
110
110
24
29
28
25
26
26
25
115
113
18
25
19
27
23
28
24
125
111
19
26
21
28
21
29
29
130
106
20
27
25
29
21
34
32
135
105
*
PPS
-
Pack
ets Pe
r
Seco
n
d
Table
3.
Packe
t deli
very
rati
o vs
P
PS
PDR
(%
)
in
CSA
-
MGR vs
P
riority
DIS
TPGF
LI
NK
Awa
re
Tiny
CSA
Tiny
h
o
p
PPS
CSA
-
M
GR
Priority
DI
S
CSA
-
M
GR
Priority
DI
S
CSA
-
M
GR
Priority
DI
S
CSA
-
M
GR
Priority
DI
S
10
25
35
45
52
64
78
78
88
15
17
27
44
51
69
85
76
87
20
15
19
32
5
0
.41
66
8
4
.12
74
86
25
8
16
30
50
60
78
66
78
30
9
12
27
48
55
73
60
72
35
10
11
23
45
52
68
56
68
40
5
10
21
40
45
64
52
65
45
4
9
19
38
36
59
45
59
50
5
8
12
35
27
54
37
56
Re
gardless
of
the
PP
S
num
be
r,
Ti
nyH
op
c
on
sist
e
ntly
e
xhibit
s
higher
la
te
ncy
co
mp
a
r
ed
t
o
oth
e
r
protoc
ols,
w
hich
is
no
te
w
or
t
hy.
T
his
diff
e
r
ence
is
at
tri
bu
t
ed
to
the
desi
gn
of
t
he
Ti
nyH
op
prot
oc
ol.
W
hile
it
op
e
rates
simi
la
rly
to
the
ad
-
hoc
on
-
de
ma
nd
distance
ve
ct
or
(
A
ODV)
protoc
ol,
it
incorp
or
at
e
s
a
s
epar
at
e
ackno
wled
gem
ent
pr
ocess
for
eve
ry
co
ntr
ol
and
in
formati
on
pack
et
.
H
ow
ever,
e
nab
li
ng
ackno
wled
gem
ents
for
i
nfo
rmati
on
pac
kets
co
ul
d
e
xte
nd
the
w
ai
ti
ng
ti
me
du
e
to
po
te
ntial
co
ll
isi
on
s
or
c
ha
nnel
util
iz
at
ion
i
ssu
es,
especial
ly
at
hi
gh
e
r
PPS
rate
s.
Fig
ur
e
4
il
lustrate
s
the
a
ve
rag
e
delay
vs.
PPS
,
sho
wing
that
the
pro
po
s
ed
protoc
ol
sig
ni
ficantl
y
reduc
es
la
te
ncy.
O
n
the
ot
her
hand,
existi
ng
proto
cols
e
xhibit
simi
la
r
delay
performa
nce.
Table
3
c
omp
ares
the
PD
R
vs
.
PPS
ac
r
oss
dif
fer
e
nt
pr
oto
c
ols,
dem
on
strat
in
g
the
s
up
e
rio
r
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
Wi
rel
ess sen
s
or netw
ork
s
bas
ed
ef
fi
ci
ent d
ri
p
irri
gati
on
monitorin
g
sy
ste
m
s
(
Ka
rthik
Sagar As
hok
)
685
performa
nce
of
Ti
ny
C
SA
[
25
],
[
26]
.
T
he
av
erag
e
P
DR
cha
ng
e
s
for
an
inc
reasin
g
num
be
r
of
P
PS
(sta
rting
a
t
10
a
nd
goin
g
up
to
50).
PDR
decli
ne
s
at
TPG
F
a
nd
val
ues
dec
rease
f
r
om
35
per
ce
nt
to
8
per
ce
nt
as
PP
S
increases
.
I
n
c
on
t
rast,
LI
N
K
Aw
a
re
sho
ws
a
relat
ivel
y
sta
bl
e
PD
R
of
a
rou
nd
50%,
w
hile
CSA
de
monstr
at
es
a
gr
a
dual
decr
ea
se
from
78
%
t
o
54%.
Ti
ny
Hop
e
xh
i
bits
the
hi
gh
est
P
D
R,
sta
rting
at
88%
an
d
dec
re
asi
ng
to
56%
as
PPS
i
ncr
ease
s.
T
his
analysis
sug
ge
sts
that
TP
GF
may
strug
gle
t
o
mai
ntain
pac
ket
deli
very
ra
te
s
as
PPS
inc
reases,
w
hile
L
INK
Aw
a
re
an
d
CS
A
offe
r
more
c
on
sist
e
nt
performa
nce,
an
d
T
iny
H
op
c
onsi
ste
ntly
achieves
h
i
gh
e
r
P
DRs ac
ro
s
s
the r
a
nge
of
PP
S v
al
ues.
Figure
5
sho
w
s
the
PD
R
for
var
i
ou
s
prot
oc
ols
agai
ns
t
the
PPS
pa
ramete
r,
hi
gh
li
gh
ti
ng
the
superi
or
performa
nce
of
Tin
yCS.
The
obser
ved
t
rend
s
hows
that
t
he
P
DR
ge
neral
ly
decr
ea
ses
as
the
PP
S
pa
r
amet
er
increases
.
Re
m
ark
a
bly,
Tin
yC
SA
ac
hieve
s
be
tt
er
PD
R
c
ompare
d
t
o
th
e
ot
her
prot
oco
ls
,
wh
ic
h
is
e
xp
e
c
te
d
as
the
paths
c
onstructe
d
f
or
bo
t
h
s
ources
in
Ti
nyCSA
a
void
th
e
car
rier
se
ns
e
eff
ect
.
T
he
LI
NK
-
A
wa
re
pro
tocol
al
so
de
monstr
at
es
relat
ively
good
P
DR
performa
nce,
as
the
pat
h
c
on
st
ru
ct
io
n
process
c
onside
r
s
li
nk
reli
abili
ty
.
Howev
e
r,
the
T
P
GF
pr
oto
c
ol
e
xh
i
bits
lo
we
r
PD
R
beca
us
e
of
it
s
gr
ee
dy
f
orwardin
g
a
ppro
ac
h.
Anothe
r
c
hara
ct
erist
ic
ob
se
r
ve
d
is
t
hat
du
rin
g
the
path
c
ons
tructi
on
proces
s,
it
al
wa
ys
c
hoose
s
the
on
e
c
losest
to
the
si
nk.
L
ast
ly,
t
he
Ti
ny
-
Hop
pr
oto
c
ol
show
s
t
he
l
east
PD
R
re
sul
t,
w
hich
can
be
at
tri
bu
te
d
t
o
t
he
chall
enges r
el
at
ed
to ac
knowl
edg
e
ments
and
potenti
al
r
et
ra
ns
missi
on
s
.
Figure
4. A
verage
delay vs
P
PS
Figure
5. N
umber
of PPS
vs
PD
R
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
:
677
-
688
686
4.2.
Discussio
n
This
resea
rc
h
discuss
e
s
pe
rformance
opti
miza
ti
on
f
or
pr
eci
sio
n
a
gri
culture
us
i
ng
f
our
WSN
protoc
ols,
na
m
el
y
Tin
yCS
A,
Link
-
awa
re,
T
PG
F
,
a
nd
Ti
ny
-
Hop.
T
he
pr
iority
-
base
d
pr
oto
c
ol
of
Tin
yC
SA
pro
vid
es
a
bette
r
pack
et
deliv
ery
rati
o
(
PD
R
)
without
ca
rr
i
er
se
ns
e
issues
,
a
nd
he
nce,
th
e
net
work
is
hi
gh
l
y
reli
able.
Th
e
pe
rformance wa
s
al
so
well
-
bal
anced
w
it
h
reli
abili
ty
du
e
t
o
the
li
nk
a
wa
reness
of
t
he
Lin
k
-
awa
re
protoc
ol.
H
owever,
t
he
perf
orma
nce
of
TP
GF
was
no
t
th
at
good
in
de
nse
net
wor
ks
,
wh
il
e
Tin
y
-
Hop
face
d
pro
blems
wit
h
ackno
wled
gme
nt.
T
he
res
ults
show
prot
oc
ols
su
c
h
as
Tiny
CS
A,
base
d
on
pri
ori
ty
pr
oto
c
ols,
gr
eat
ly
imp
r
ove commu
nicat
ion ef
fici
e
nc
y.
−
Tiny
CS
A:
Ac
hieve
d
bette
r
PD
R
by
co
ns
t
ru
ct
in
g
pat
hs
that
a
void
the
carrier
se
ns
e
e
ff
ect
,
e
nh
a
nci
ng
ov
e
rall
netw
ork
pe
rforma
nce
.
This
fi
nd
i
ng
al
ign
s
wit
h
th
e
hypothesis
t
hat
pr
i
or
it
y
-
ba
sed
pr
oto
c
ols
can
impro
ve netw
ork
r
el
ia
bili
ty.
−
Link
-
awa
re
pr
oto
c
ol:
Ba
la
nc
ed
betwee
n
pat
h
reli
abili
ty
a
nd
perf
ormance
,
w
hich
s
upport
s
the
obje
ct
ive
of
achievin
g ef
fici
ent and
reli
able com
munica
ti
on.
−
TPG
F
pr
oto
c
ol:
Strugg
le
d
with
lo
wer
P
DR
due
to
it
s
path
s
el
ect
ion
strat
e
gy,
i
nd
ic
at
in
g
th
e
li
mit
at
ion
s
of
gr
ee
dy fo
rw
a
r
di
ng
i
n dense
ne
twork
scena
rio
s.
−
Tiny
-
H
op
prot
oco
l:
Faced
c
ha
ll
eng
es
with
ackno
wled
gm
e
nts,
le
a
ding
to
re
duced
P
DR,
hi
gh
li
ghti
ng
t
he
importa
nce
of
eff
ic
ie
nt ac
knowle
dgme
nt me
chan
is
ms.
The
resu
lt
s
fro
m
Tin
yCS
A
in
dicat
e
that
pri
or
it
y
-
ba
sed
pr
oto
c
ols
are
hi
g
hly
e
ff
ect
ive
in
op
ti
mizi
ng
WSN
perform
ance,
pa
rtic
ularly
i
n
sce
nar
i
os
re
qu
i
rin
g
hig
h
data
integ
r
it
y
an
d
l
ow
la
te
ncy
.
T
he
s
uperi
or
performa
nce
of
Tin
yCS
A
in
te
rms
of
PD
R
su
ggest
s
t
hat
pri
or
it
y
-
based
prot
oco
ls
ca
n
si
gn
i
ficantl
y
im
pro
ve
netw
ork
reli
ab
il
it
y,
al
ign
ing
with
the
stu
dy
's
ob
je
ct
ive
of
enh
a
ncin
g
c
ommu
nicat
io
n
eff
ic
ie
nc
y
in
WSNs
.
Com
par
e
d
to
pr
e
vious
stu
dies,
Ti
ny
CS
A's
a
bili
ty
to
av
oid
t
he
c
arr
ie
r
sens
e
e
ff
ect
ma
r
ks
a
nota
ble
impro
veme
nt.
Howe
ver,
the
study's
reli
anc
e
on
simulat
io
ns
li
mit
s
t
he
gen
e
ra
li
zabil
it
y
of
the
se
fin
dings.
Un
e
xpect
ed
re
su
lt
s
in
the
T
P
GF
prot
oco
l
in
dicat
e
pote
ntial
issues
with
greedy
forw
a
r
ding
i
n
de
ns
e
net
works,
warrantin
g
f
ur
t
her
in
vestigat
ion.
T
his
resea
r
ch
w
ork
is
ai
m
ed
at
de
velo
ping
a
n
e
ff
ic
ie
nt
WSN
a
rch
it
ect
ur
e
f
or
bette
r
w
at
er
util
iz
at
ion
in
preci
sion
ag
ricul
ture.
T
he
fi
nd
i
ng
s
unde
rsc
ore
the
im
portan
ce
of
pr
io
rity
-
base
d
protoc
ols,
bu
t
fu
t
ur
e
re
searc
h
sho
uld
te
st
these
prot
oco
l
s
in
real
-
worl
d
ag
ricult
ural
set
ti
ng
s
an
d
c
on
si
der
add
it
io
nal
Q
oS
pa
rameters
.
Un
a
nswere
d
qu
e
sti
on
s
i
nc
lud
e
the
pro
tocols'
perfor
mance
unde
r
varyin
g
env
i
ronme
ntal
conditi
ons a
nd
their a
pp
li
cabil
it
y
to
oth
e
r WS
N
a
pp
li
cat
io
ns
.
5.
CONCL
US
I
O
NS
A
ND FUT
UR
E
WO
RK
In
this
wor
k,
a
WSAN
-
ba
sed
arch
it
ect
ur
al
model
for
dr
ip
irrigati
on
syst
em
is
propose
d.
T
his
m
od
e
l
include
s
se
nso
rs
to
meas
ure
pr
ess
ure,
te
mpe
ratur
e
,
a
nd
s
oi
l
mo
ist
ure
to
track
irri
gation
op
e
rati
ons,
inc
lud
in
g
instances
of
s
ys
te
m
fail
ure.
We
achie
ve
d
hi
gh
Q
oS
pe
rformance
w
he
n
c
ompar
ed
with
t
he
CS
A
-
MGR
protoc
ol
a
nd
c
on
si
der
i
ng
tw
o
traf
fic
le
vels.
O
ur
m
et
hod
performs
bette
r
t
han
the
e
xisti
ng
prot
oco
ls
in
the
te
rminolo
gy
of
la
te
ncy
a
nd
P
DR
f
or
pri
ori
ty
tra
ff
ic
,
as
s
how
n
by
e
xtens
ive
sim
ulati
ons
us
i
ng
the
T
OS
S
I
M
simulat
or.
We
al
so
pe
rforme
d
te
sts
in
a
re
al
te
st
env
ir
onment,
c
onfir
ming
a
pprecia
ble
adv
a
nce
ments
ov
e
r
current
te
c
hn
i
ques.
We
inte
nd
to
c
oncentr
a
te
our
f
uture
work
on
act
ua
l
agr
ic
ultu
ral
la
nds
with
a
dd
ed
Q
oS
par
a
mete
rs.
REFERE
NCE
S
[1]
M.
T.
Ng
u
y
en
,
C.
V
Ng
u
y
en
,
an
d
H.
N.
Ng
u
y
en
,
“
Visu
alizatio
n
-
b
ased
m
o
n
ito
ring
in
early
w
arnin
g
sy
stems
wit
h
wireless
s
en
so
r
n
etwo
rks
,
”
Ind
o
n
esia
n
Jo
u
rnal
o
f
Electrica
l
Eng
in
e
erin
g
a
n
d
Co
mp
u
ter
S
cien
ce
,
v
o
l.
3
1
,
n
o
.
1
,
p
p
.
2
8
1
–
2
8
9
,
Ju
l.
2
0
2
3
,
d
o
i: 10
.1159
1
/ijee
cs.v
3
1
.i1.p
p
2
8
1
-
2
8
9
.
[2]
A.
C
am
illi
,
C.
E.
Cu
g
n
asca,
A.
M
.
Saraiva,
A
.
R
.
Hi
rakawa,
an
d
P
.
L.
P.
Co
rr
êa,
“
F
rom
wi
reless
s
en
so
rs
to
field
m
ap
p
in
g
:
An
ato
m
y
o
f
an
ap
p
licatio
n
for
p
r
eci
sio
n
ag
ricultu
re,
”
Co
m
p
u
ters
a
n
d
Electro
n
ics
in
Agr
ic
u
ltu
re
,
v
o
l.
5
8
,
n
o
.
1
,
p
p
.
2
5
–
3
6
,
Au
g
.
2
0
0
7
,
d
o
i: 1
0
.10
1
6
/j.compag
.20
0
7
.01.
0
1
9
.
[3]
R.
Bas
k
ar,
G.
A
.
Ku
m
ar,
an
d
D.
Kara
n
,
“
S
m
a
rt
ag
ricultu
ral
re
m
o
te
m
o
n
ito
ring
sy
ste
m
for
b
etter
so
il
h
ealth
u
sing
IoT,
”
Inter
n
a
tio
n
a
l Jo
u
r
n
a
l
o
f Health
Scien
ces
,
v
o
l.
6
,
n
o
.
8
,
p
p
.
1
2
3
9
–
1
2
5
1
,
Jun. 20
2
2
,
d
o
i: 10
.5
3
7
3
0
/ijh
s.v
6
n
S4
.9
8
8
5
.
[4]
D.
Mahes
war
i,
R
.
Harish
,
S
.
So
w
m
iy
a,
an
d
K
.
A
iswary
a,
“
S
m
a
rt
c
rop
p
rotectio
n
sy
stem
for
ag
ricultu
re
u
sin
g
I
o
T,
”
Inter
n
a
tio
n
a
l
Jo
u
rn
a
l of Cr
ea
tiv
e
Resear
ch
Tho
u
g
h
ts (I
JCRT
)
,
v
o
l.
22
,
n
o
.
1
,
p
p
.
4
4
3
–
4
4
7
,
2
0
2
3
.
[5]
L
.
G
a
r
c
í
a
,
L
.
P
a
r
r
a
,
J
.
M
.
J
i
m
e
n
e
z
,
J
.
L
l
o
r
e
t
,
a
n
d
P
.
L
o
r
e
n
z
,
“
I
o
T
-
b
a
s
e
d
s
m
a
r
t
i
r
r
i
g
a
t
i
o
n
s
y
s
t
e
m
s
:
a
n
o
v
e
r
v
i
e
w
o
n
t
h
e
r
e
c
e
n
t
t
r
e
n
d
s
o
n
s
e
n
s
o
r
s
a
n
d
I
o
T
s
y
s
t
e
m
s
f
o
r
i
r
r
i
g
a
t
i
o
n
i
n
p
r
e
c
i
s
i
o
n
a
g
r
i
c
u
l
t
u
r
e
,
”
S
e
n
s
o
r
s
,
v
o
l
.
2
0
,
n
o
.
4
,
F
e
b
.
2
0
2
0
,
d
o
i
:
1
0
.
3
3
9
0
/
s
2
0
0
4
1
0
4
2
.
[6]
D.
W
o
h
we
Sa
m
b
o
,
A.
Fo
rster
,
B.
O
.
Yen
k
e,
I.
Sa
rr
,
B.
Gu
ey
e,
an
d
P.
Day
an
g
,
“
W
ireless
u
n
d
ergrou
n
d
sen
so
r
n
etwo
rks
p
ath
lo
ss
m
o
d
el
for
p
recisi
o
n
ag
ricultu
re
(W
USN
-
P
LM
),
”
I
EE
E
S
en
so
r
s
Jo
u
rn
a
l
,
v
o
l.
2
0
,
n
o
.
1
0
,
p
p
.
5
2
9
8
–
5
3
1
3
,
May
2
0
2
0
,
d
o
i:
1
0
.11
0
9
/JSEN.20
2
0
.29
6
8
3
5
1
.
[7]
K.
Riah
i,
G.
K
ah
n
,
B.
D
aff
lo
n
,
an
d
J.
Laval,
“
A
faulty
IoT
n
etwo
rk:
si
m
u
latin
g
sen
so
rs
an
d
p
erturbatio
n
s,
”
in
Lectu
re
No
tes
in
Netw
o
rks
an
d
Sys
t
ems
,
v
o
l.
4
7
0
,
Sp
ri
n
g
er
Internatio
n
al
Pu
b
lish
in
g
,
2
0
2
2
,
p
p
.
8
7
–
9
7
.
Evaluation Warning : The document was created with Spire.PDF for Python.