Indonesi
an
Journa
l
of El
ect
ri
cal Engineer
ing
an
d
Comp
ut
er
Scie
nce
Vo
l.
23
,
No.
1
,
Ju
ly
2021
, p
p.
273
~
284
IS
S
N: 25
02
-
4752, DO
I: 10
.11
591/ijeecs
.v
23
.i
1
.
pp
273
-
2
84
273
Journ
al h
om
e
page
:
http:
//
ij
eecs.i
aesc
or
e.c
om
Energy e
ffici
ent WS
N usi
ng
hybri
d modifi
ca
tion P
EGASIS
with
ant lion opti
miz
ation
Ah
med
Abdul
A
z
eez
A
smae
l
, Ba
sm
an Al
-
Ned
awe
Middle
Te
chn
ica
l
Univer
si
t
y
,
Tec
hnic
a
l
Insti
tut
e
o
f
Baquba,
B
aqub
a,
D
a
y
ala,
Ira
q
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
Feb
20
, 202
1
Re
vised Ju
n 7
,
2021
Accepte
d
J
un
18
, 202
1
W
ire
le
ss
sensor
nodes
consist
of
ti
n
y
e
lectr
on
ic
device
s
tha
t
ca
n
sense,
tra
nsm
it
,
and
m
ea
sure
data
from
phy
sic
al
environm
ent
s
such
a
s
the
fie
ld
of
m
int
er
surveil
l
a
nce
.
The
se
sens
or
nodes
signifi
ca
nt
l
y
d
epe
nd
o
n
bat
te
r
ie
s
to
gai
n
energ
y
whi
ch
is
used
to
oper
ations
associa
t
ed
with
comm
unic
a
ti
on
an
d
computat
ion
.
Gene
ra
lly
,
design
ing
comm
unic
ation
protoc
o
ls
is
fea
sibl
e
t
o
ac
hi
eve
eff
ec
t
iv
e
usage
of
the
se
ene
rg
y
r
esourc
es
of
the
sensor
node.
Both
rep
o
rte
d
m
edi
u
m
ac
ce
ss
cont
ro
l
and
routi
ng
c
a
n
ac
h
ie
ve
en
erg
y
-
saving
th
at
supporting
re
al
t
ime
func
t
iona
l
ity
.
Th
is
pap
er
e
m
phasiz
es
the
u
se
of
h
y
brid
m
odifi
ed
PEGA
SIS
-
Ant
li
on
opti
m
iz
at
ion
.
Sever
al
steps
ar
e
enta
il
ed
in
thi
s
rese
arc
h
.
First
is
ran
dom
distri
bu
t
ion
of
node
fo
llow
ed
b
y
c
luste
ri
ng
the
m
ap
as
a
ci
r
cul
ar
reg
i
on.
Th
en,
the
no
des
are
conn
ec
t
e
d
to
the
cl
osest
node
in
that
reg
ion.
In
conse
quence,
PEGA
SIS
-
Ant
li
on
o
pti
m
iz
ation
is
appl
i
ed
to
enha
nc
e
the
con
nec
t
ion
of
the
no
des
and
accom
pli
sh
the
m
axi
m
um
li
fe
batte
r
of
the
sensor.
A
t
la
st
,
the
expe
r
i
m
ent
s
per
form
ed
in
th
is
work
d
emons
tra
te
tha
t
th
e
propose
d
opti
m
iz
a
ti
on
t
ec
hniqu
e
oper
ates
well
in
te
rm
s
of
net
work
la
t
ency
,
power
dura
ti
on
and
en
erg
y
’s
consum
p
ti
on.
Furthermo
re,
the
li
f
e
span
of
the
nodes
has
im
prove
d
gre
at
l
y
b
y
87%
over
the
origi
n
al
al
gor
it
hm
tha
t
accom
pli
she
d
a
r
ate
of
li
f
e
n
odes
of
60%
.
Ke
yw
or
ds:
An
t l
io
n o
ptim
iz
at
ion
Energy c
onsum
pt
ion
Nodes
PEGAS
IS
W
i
reless se
nso
r
This
is an
open
acc
ess arti
cl
e
un
der
the
CC
B
Y
-
SA
l
ic
ense
.
Corres
pond
in
g
Aut
h
or
:
Ba
s
m
an
Al
-
Ne
daw
e
Dep
a
rtm
ent o
f c
om
pu
te
r
syst
e
m
Mi
dd
le
Tec
hnic
al
U
ni
ver
sit
y
Em
a
il
: b.
al
-
ne
daw
e
@m
tu.edu.iq
1.
INTROD
U
CTION
Adva
ncem
ents
an
d
in
novatio
ns
occurri
ng
in
the
fiel
d
of
wir
el
ess
com
m
unic
at
ion
play
a
key
r
ole
in
dev
el
op
i
ng
wir
el
ess
sensor
ne
tworks
that
c
on
sist
of
m
ino
r
de
vices
that
can
gat
her
in
f
or
m
at
ion
by
w
orkin
g
coope
rati
vely
[
1]
-
[
3]
.
T
hese
ti
ny
sen
sin
g
devi
ces
are
known
as
no
des
a
nd
con
ta
in
the
f
ollow
i
ng
com
pone
nts:
(1)
central
pr
oc
essing
un
it
,
util
iz
ed
to
pr
oce
ss
data,
(
2)
m
e
m
or
y,
util
iz
ed
to
store
da
ta
,
(
3)
batte
ry,
uti
li
zed
to
ob
ta
in
e
ne
rg
y,
(4)
tra
ns
cei
ve
r,
util
iz
ed
to
rec
ei
ve
an
d
se
nd
sign
al
s
am
on
gst
var
io
us
node
s.
It
is
im
po
rtant
t
o
reali
ze
that
diff
ere
nt
nodes
ha
ve
dif
fer
e
nt
siz
es
dep
e
nd
i
ng
on
the
f
un
ct
i
on
it
carries
out.
For
instanc
e,
i
n
app
li
cat
io
ns
re
la
te
d
with
m
ili
ta
ry
or
s
urveil
la
nce
as
pects,
t
he
siz
e
of
the
sens
or
node
m
ay
be
inv
isi
ble
to
the
nak
e
d
ey
e.
I
n
te
rm
s
of
co
st,
facto
rs
li
ke
m
e
m
or
y
sp
ace
avail
able
f
or
data
sto
ra
ge,
ba
tt
ery
and
the
sp
ee
d
require
d
to
pr
ocess
data
(i
nfo
rm
at
ion
)
al
l
influ
e
nce
m
assivel
y
[4
]
,
[
5].
N
owadays
,
there
are
se
ver
al
app
li
cat
io
ns
of
W
S
Ns
i
n
var
i
ou
s
fiel
ds,
wh
i
ch
ca
n
al
l
be
e
m
plo
ye
d,
how
ever
are
not
li
m
it
ed.
For
e
xa
m
ples,
areas
of
t
rad
e
and
in
du
st
ry
s
uch
as
healt
hc
are
ben
e
fit
by
m
on
it
or
ing
en
vir
on
m
ent
an
d
hab
it
at
,
or
sur
ve
il
la
nce
(like
m
ilit
ary
fiel
ds
).
The
use
of
WSNs
is
w
it
nessing
a
noti
ceable
prolifer
at
ion
a
nd
at
th
e
sam
e
tim
e
their
use
is ham
per
ed by
an
iss
ue o
f
e
ne
rg
y
-
relat
ed
c
onstrai
nts
r
e
gardin
g
th
e
b
at
te
r
y’s lim
it
ed
dur
at
ion
[6
]
, [
7].
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
23
, N
o.
1
,
Ju
ly
2021
:
273
-
2
8
4
274
Since
each
no
de
is
ene
rg
y
-
de
pende
nt
in
it
s
act
ivit
ie
s,
this
op
e
rati
on
pos
es
a
chall
en
ge
in
W
S
Ns;
wh
e
re
t
he
dys
functi
on
of
an
y
node
inter
rupts
the
w
hole
syst
e
m
ov
erall
.
Ty
pical
ly
,
ea
ch
node
can
t
ake
a
nu
m
ber
of
patte
rn
s.
At
so
m
e
po
i
nts,
it
m
igh
t
be
a
n
ac
tive
p
attern
(
m
od
e)
(in
te
r
m
s
of
receivi
ng
a
nd
transm
itti
ng
da
ta
/
inform
at
ion
),
or
a
n
idl
e
p
attern
or
a
sle
ep
p
attern
.
N
od
e
s
are
see
n
t
o
be
i
n
act
ive
m
od
e
wh
il
st
recei
vin
g
or
se
nd
i
ng
data.
T
he
fo
l
lowing
points
are
s
uggeste
d
to
sa
ve
e
ne
rg
y
or
reduce
energy
consum
ption
pro
duced
b
y c
om
m
un
ic
at
ion
proces
se
s in
WSNs
[8
]
, [
9]
,
a)
S
c
he
du
l
i
ng
t
h
e
m
od
e
of
t
he
no
de
,
i
.
e
.
r
e
c
e
i
vi
n
g
m
od
e
,
t
r
a
nsm
i
t
ti
ng
m
od
e
,
s
l
e
e
p,
or
i
d
l
e
m
o
d
e
(
p
a
t
t
e
r
n
)
.
b)
A
l
t
e
r
i
ng
the
r
a
n
g
e
o
f
transm
issio
n
a
m
on
g
se
ns
i
ng
n
od
e
s
.
c)
Em
plo
yi
ng
w
e
l
l
-
or
ga
ni
z
e
d
w
h
i
l
s
t
e
n
s
u
r
i
n
g
h
i
g
h
l
y
e
f
f
i
c
i
e
n
t
ro
uting
a
nd
da
t
a
–
ga
t
he
r
i
ng
m
e
t
h
o
d
s
.
d)
Pr
e
ven
t a
nd a
void
deali
ng
with a
ny
u
nd
e
s
i
r
a
bl
e
da
t
a
.
P
r
a
c
t
i
c
a
l
l
y
,
b
a
t
t
e
r
y
i
s
r
e
g
a
r
d
e
d
a
s
t
h
e
s
o
l
e
s
o
u
r
c
e
f
o
r
l
i
f
e
s
u
s
t
a
i
nm
e
n
t
o
f
n
o
d
e
s
i
n
WS
N
s
.
T
hi
s
i
s
be
c
a
us
e
a
ny
com
m
un
ic
at
ion
ta
kin
g
place
w
i
t
h
ot
he
r
no
de
s
or
a
ny
s
ensin
g
a
c
t
i
vi
t
ie
s
w
o
u
l
d
ne
e
d
great
a
m
ou
nts
of
e
ne
rg
y
f
or
bo
t
h
proces
sin
g
an
d
t
r
a
nsm
it
t
i
ng
the
ga
t
he
r
e
d
data
to
the
ba
s
e
s
t
a
t
i
on
(
B
N
)
.
I
n
m
ost
c
a
s
e
s
s
uc
h
a
s
s
ur
ve
i
l
l
a
nc
e
a
pp
l
i
c
a
ti
on
s
,
r
e
pl
a
c
i
ng
ba
t
t
e
r
y
th
a
t
de
pl
e
t
e
s
or
dr
a
i
ns
e
ne
r
gy
i
s
hi
gh
l
y
s
ug
g
e
s
t
e
d.
T
he
r
e
f
o
r
e
,
a
l
o
t
of
r
e
s
e
a
r
c
he
r
s
w
e
r
e
ha
vi
ng
m
ul
ti
pl
e
a
tt
em
pt
s
a
t
di
s
c
ov
e
r
i
ng
s
ui
t
a
bl
e
e
ne
r
gy
-
e
f
f
i
c
i
e
nt
pr
ot
oc
ol
s
f
or
W
S
N
s
.
T
he
a
im
be
hi
nd
s
u
c
h
a
c
t
i
on
i
s
t
o
e
l
im
i
na
te
i
s
s
ue
s
s
uc
h
a
s
t
h
os
e
m
e
nt
i
on
e
d
a
bo
ve
[
1
0
]
-
[
1
2]
.
I
t
i
s
s
a
i
d
t
ha
t
a
pr
ot
oc
ol
ha
s
r
e
a
l
-
t
im
e
s
up
po
r
t
a
s
l
on
g
a
s
i
t
s
s
pe
e
d
a
nd
r
e
l
i
a
bi
l
it
y
t
ow
a
r
ds
r
e
a
c
t
i
on
s
in
t
he
ne
t
w
or
k
a
r
e
f
ou
nd
i
n
t
he
i
r
up
m
os
t
e
xt
e
nt
. M
or
e
o
ve
r
,
a
p
r
ot
oc
ol
i
s
e
xp
e
c
t
e
d
t
o
pr
o
vi
de
r
e
du
nd
a
nt
da
t
a
t
o
B
S
by
us
i
ng
t
he
ga
t
he
r
e
d
data
f
r
o
m
t
he
w
ho
l
e
s
e
ns
i
ng
no
de
.
A
p
pr
o
pr
i
a
t
e
l
y
,
a
ny
de
l
a
y
f
ou
n
d
i
n
t
he
ne
t
w
or
k
w
hi
l
s
t
t
r
a
ns
f
e
r
r
i
n
g
da
t
a
f
r
om
se
ns
or
no
de
s
t
o
ba
s
e
s
t
a
ti
on
s
ho
ul
d
be
s
ho
r
t
.
T
hu
s
,
t
he
r
e
s
ul
t
i
s
a
r
a
pi
d
r
e
a
c
t
i
on
[
1
3]
,
[
14
]
.
T
he
w
or
k
a
c
c
om
pl
i
s
he
d
i
n
t
hi
s
pa
pe
r
m
a
ke
s
us
e
of
hy
br
i
d
a
nt
l
i
on
op
t
i
m
i
z
at
i
on
a
s
w
e
l
l
a
s
m
od
i
f
i
e
d
P
E
G
A
S
I
S
.
T
he
ob
je
c
t
i
ve
s
of
t
hi
s
w
o
r
k
a
r
e
t
o
im
pr
ov
e
t
he
n
od
e
s
s
e
ns
or
s
n
e
t
w
or
k
i
n
t
e
r
m
s
of
ne
t
w
or
k
l
a
t
e
nc
y
,
po
w
e
r
d
ur
a
t
i
on
a
nd
e
ne
r
gy
’
s
c
on
s
um
pt
i
on
.
T
hi
s
w
i
l
l
be
a
c
hi
e
ve
d
us
i
n
g
hy
br
i
d
m
od
i
f
i
e
d
P
E
G
A
S
I
S
-
A
nt
l
i
on
op
t
im
i
z
at
i
on
.
S
e
ve
r
a
l
s
t
e
ps
w
il
l
be
e
nt
ai
le
d
in
t
hi
s
r
e
s
e
a
r
c
h.
F
i
r
s
t
,
a
r
a
nd
om
di
st
r
i
bu
t
i
on
of
n
od
e
f
ol
l
ow
e
d
by
c
l
us
t
e
r
i
ng
the
m
a
p
a
s
a
ci
r
c
ul
a
r
r
e
gi
on
.
T
he
n,
t
he
n
od
e
s
a
r
e
c
on
ne
c
t
e
d
t
o
t
he
c
l
os
e
s
t
no
de
i
n
t
ha
t
r
e
gi
on
.
C
on
s
e
q
ue
nt
l
y
,
P
E
G
A
S
I
S
-
A
nt
l
i
on
op
t
i
m
i
z
at
i
on
w
i
l
l
be
a
p
pl
i
e
d
t
o
e
nh
a
n
c
e
t
he
c
on
ne
c
t
i
on
o
f
t
he
n
od
e
s
a
nd
a
c
c
om
pl
i
s
h
t
he
m
a
xi
m
um
li
f
e
ba
t
t
e
r
of
t
he
s
e
ns
or
.
2.
RELATE
D
W
ORK A
N
D
B
ACKG
ROUN
D
TH
ORY
2.1
.
Li
ter
atu
re sur
vey
In
the
la
st
te
n
ye
ars,
resea
rch
e
rs
ha
ve
c
onduct
ed
a
lo
t
of
stu
dies
c
on
ce
r
ning
WSN
ene
rg
y
consum
ption
i
ssu
es
w
her
e
a
su
r
vey
of
le
adin
g
a
ppro
ac
hes
a
nd
prot
oc
ols
w
it
h
the
m
a
in
aspects
of
it
s
com
par
ison are
provide
d,
a)
Ri
na
Ma
haku
d
et
al
.
in
2016
[15],
pr
opos
e
d
PEGAS
IS
a
ne
ar
opti
m
a
l
chain
-
based
prot
oco
l,
w
hich
w
as
us
e
d
for
exte
ndin
g
net
work
li
fetim
e.
In
PE
G
AS
I
S,
eac
h
no
de
can
c
omm
u
nicat
e
with
only
a
neighborin
g
node,
pe
rfor
m
s
a
chain
,
an
d
el
ect
s
a
le
ader
f
ro
m
the
chain.
This
w
ould
c
ollec
t
an
y
data
from
t
he
su
r
rou
nd
i
ng
node
s
w
hich
ar
e
sent
to
t
he
BS.
As
a
res
ul
t,
it
is
po
ssibl
e
to
achie
ve
a
reduce
d
pow
er
consum
ption
t
hat can
b
e
u
ti
li
zed to i
nc
rease
both netw
ork'
s co
m
pe
te
nce a
nd li
fetim
e.
b)
Dok
o
Ba
ndur
Đoko
Banđur
,
Brani
m
ir
Jakšić
,
Miloš
Banđur
,
and
Srđan
Jov
ić
.
in
20
19
[
16]
,
fo
c
us
e
d
on
analy
sis
li
nk
e
d
to
energy
eff
ic
ie
ncy
in
(
WSNs)
that
is
e
m
plo
ye
d
in
the
fiel
d
of
s
m
art
agr
ic
ultu
re
,
env
i
ronm
ent
i
nputs
an
d
re
quirem
ents,
analy
sis
and
plan
ning,
an
d
desi
gn
ph
ase
s.
Sim
ula
ti
on
was
by
dep
l
oying
500
sensor
no
des
on
in
area
of
800
width
by
800
le
ngth
m
2
.
The
a
ver
age
of
ene
r
gy
consum
ption
w
as foun
d
to
b
e
49.29 %
.
c)
Ra
j
Pr
iy
adarsh
ini,
in
2019
[17]
pr
ese
nted
w
ork
that
app
li
e
d
a
hybr
i
d
m
e
chan
ism
fo
r
im
pr
ov
in
g
ene
r
gy
eff
ic
ie
ncy
to
ob
ta
in
faster
transm
issi
on
of
data
in
an
unde
r
water
WSN.
The
re
fore,
antc
olon
y
op
ti
m
iz
ati
on
(
ACO
)
r
ou
ti
ng
with
m
ark
ov
-
c
hain
m
on
te
-
ca
r
lo
(MCM
C)
al
gorithm
wer
e
app
li
ed
.
T
his
is
a
m
et
ho
d
us
e
d
to
ha
nd
le
t
his
t
rou
ble
an
d
to
captu
rin
g
a
ny
t
ran
sm
issi
on
l
oss
in
the
Ma
rkov
Chai
n
M
onte
Ca
rlo
Me
th
od.
More
s
pecific
al
ly
,
channel
sta
tus
in
form
ation
f
or
ec
ast
pred
ic
ti
on
al
gorith
m
was
us
e
d.
I
n
par
ti
cula
r,
t
he
process
of
e
val
uation
of
e
xpe
rim
ental
si
m
ul
at
ion
s
was
car
r
ie
d
out
by
us
in
g
perform
ance
evaluati
on.
d)
So
m
aur
oo
an
d
V.
Ba
ssooin
2019
[
18
]
,
il
lu
strat
ed
r
ou
ti
ng
al
go
rithm
s
that
increase
the
sensor
no
des’
li
fetim
e
in
3D
area
wireless
s
ens
or
netw
ork
s.
This
was
ac
com
plished
by
us
in
g
PE
GAS
IS
pr
oto
c
ol
an
d
gen
et
ic
al
gorit
hm
to
est
ablis
h
the
f
ull
cha
in.
T
he
prot
oc
ol
was
e
xec
uted
f
or
netw
or
ks
co
ns
i
der
i
ng
separ
at
e
cases
of
a
fixe
d
B
S
bo
t
h
outsi
de
and
insi
de
th
e
netw
ork.
T
he
resu
lt
s
sho
w
ed
a
sign
i
ficant
enh
a
ncem
ent in th
e
li
fetim
e o
f
PE
G
AS
I
S
by
81.7%.
e)
K.
K
ar
un
a
nith
y,
B.
Velusam
y
in
2020
[19
]
,
dem
on
strat
ed
a
cl
us
te
rtree
based
e
nergy
eff
ic
ie
nt
dat
a
gathe
rin
g
(CT
EED
G)
prot
oc
ol
to
le
ng
t
hen
bo
t
h
th
r
oughput
an
d
li
feti
m
e
of
the
W
S
N.
In
t
he
phase
of
inter
-
cl
ust
er
co
m
m
un
ic
at
ion
,
tree
topolo
gy
is
fo
rm
ed
a
m
ongs
t
cl
us
te
rs
to
wards
the
(BS
)
that
gu
ara
ntee
s
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c
En
g
&
Co
m
p
Sci
IS
S
N:
25
02
-
4752
Ener
gy
ef
fi
ci
en
t WSN
us
in
g hy
br
id
mod
if
ic
ati
on PE
GA
SIS
w
it
h
ant l
io
n…
(
Ah
med Ab
dul
Azeez As
m
ael)
275
the
obta
ina
bili
ty
of
th
e
c
onge
sti
on
-
fr
ee
s
hortest
path
to
t
he
base
sta
ti
on
.
From
the
si
m
ulati
on
res
ults,
CTEED
G
ha
s
perform
ed
bette
r
tha
n
the
D
L
-
LE
ACH
by
38.28%
w
her
e
FA
MACR
O
W
wa
s
bette
r
by
28.81%
in
t
he
t
hro
ughput.
2.2.
PEG
ASI
S (
p
ower e
ff
ic
ie
nt
gath
eri
n
g i
n sens
or
in
f
or
mat
i
on
sy
s
te
ms
)
T
he
m
a
i
n
c
on
c
e
pt
be
hi
nd
P
E
G
A
S
I
S
i
s
t
ha
t
it
us
e
s
al
l
node
s
t
o
t
r
a
ns
m
i
t
or
r
e
c
e
i
ve
da
t
a
w
i
t
h
t
he
ne
a
r
e
s
t
neig
hbor
i
ng
no
de
s
.
A
c
ha
i
n
i
s
f
or
m
ed
i
n
or
de
r
t
o
a
c
hi
e
ve
t
hi
s
,
a
s
di
s
pl
a
y
e
d
i
n
F
ig
ur
e
1.
A
l
l
t
he
no
de
s
t
ha
t
a
r
e
r
e
s
po
ns
i
bl
e
f
or
ga
t
he
r
i
ng
da
t
a
,
c
om
bi
ne
w
i
t
h
da
t
a
f
rom
t
he
ne
i
gh
bo
r
i
ng
n
od
e
s
w
h
e
r
e
ov
e
r
a
l
l
a
r
e
s
e
nt
t
o
t
he
ne
a
r
e
s
t
ne
ig
h
bo
r
i
n
g
no
de
.
Sp
eci
fi
cal
ly
,
t
his
m
e
tho
d
e
nsures
t
hat
al
l
nodes
recei
ve
and
fu
se
t
heir
data,
wh
e
re
t
he
data
is
la
te
r
passe
d
to
the
ne
xt
ne
ighbor
i
n
the
form
of
a
c
hai
n
un
ti
l
al
l
data
arr
i
ves
t
o
BS
.
It
is
ackno
wled
ged
that
each
node
in
the
netw
ork
r
otate
s
as
the
chain
le
ader
and
is
res
ponsi
ble
fo
r
tra
nsfe
rr
i
ng
fu
ll
y i
nteg
rated
d
at
a that
is obt
ai
ned
by the
node
c
hain
t
o
t
he B
S [20
]
.
Fig
ure
1. F
or
m
at
ion
of chai
n usin
g n
od
es
in PEG
ASIS
Using
this
a
pp
ro
ac
h,
e
ne
rg
y
l
oad
is
eq
ually
disp
e
nded
bet
ween
sen
sin
g
nodes
i
n
the
ne
twork
w
her
e
the
w
hole
ne
twork
nodes
is
us
e
d
to
f
or
m
the
c
hain
an
d
pe
rfor
m
un
c
ompli
cat
ed
data
-
f
orwardin
g
ope
rati
on
s
.
If
a
no
de dies i
n
the
pr
oduce
d chain
, a ne
w
c
hain
is
m
ade wher
e
the
dea
d n
od
e
s ar
e
r
em
oved [2
1].
2.3
.
Radio
and ener
gy m
odel
Dem
on
strat
ed
in
Fig
ur
e
2
is
the
first
-
orde
r
ra
dio
m
od
el
wh
ic
h
is
us
e
d
f
or
c
om
pu
ti
ng
e
nergy
consum
ption
a
sso
ci
at
ed
with
com
m
un
ic
at
ing
no
des.
T
his
sta
ge
is
sim
ple
to
be
a
ppli
ed
and
e
xploit
ed
m
os
tly
in
li
te
ratur
e
of
wireless
sen
so
r
netw
orks
[22].
He
nce,
it
beco
m
es
ea
sie
r
and
m
or
e
reli
able
to
m
ake
com
par
isons
w
it
h
pr
e
vious
pr
oto
c
ols.
It
is
a
fact
that
energ
y
is
an
essenti
al
factor
f
or
run
ning
tran
sm
i
tt
e
r
an
d
receiver
circ
uits. T
he
la
tt
er is
a f
un
ct
io
n of t
he
num
ber
of
bits (
k)
only
[23
]
as prese
nted
i
n
(
1
)
.
E
Tx
Rx
−
e
le
c
=
E
e
lec
∗
k
(1)
Wh
e
re
is
the
energy
us
e
d
in
each
bit
t
o
operate
a
tra
ns
m
i
tt
er
or
a
receiv
er.
Tra
ns
m
issio
n
po
wer
that
is
us
e
d by the am
plifie
rs rep
rese
nts a
functi
on
of both
d
ist
anc
e of tra
ns
m
issio
n an
d
t
he num
ber
o
f bit
s.
E
Tx
−
a
mp
{
∈
fs
∗
k
∗
d
2
,
if
d
<
d
0
∈
mp
∗
k
∗
d
4
,
if
d
<
d
0
(2)
0
is a
th
res
ho
l
d
t
hat is
util
iz
ed
f
or d
et
erm
ining m
ult
ipath, fr
ee
sp
ace
and is a
ccom
plished
by
,
d
0
=
√
∈
fs
∈
mp
(3)
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
23
, N
o.
1
,
Ju
ly
2021
:
273
-
2
8
4
276
w
he
re
the
co
nst
ants
∈
fs
a
nd
∈
m
p
are
e
m
plo
ye
d
i
n
the
m
ulti
pa
th
m
od
el
and
free
spa
ce
ind
e
pe
nd
e
ntly
.
Th
us
,
the cons
um
ption
e
nergy re
qui
r
ed fo
r
tra
ns
m
itti
ng
pack
et
s
is,
E
Tx
=
{
E
e
lec
×
k
+
∈
fs
×
k
+
d
2
,
if
d
<
d
0
E
e
lec
×
k
+
∈
mp
×
k
+
d
4
,
if
d
<
d
0
.
.
(4)
t
he
energy
that
is
essenti
al
for
the
proce
ss
of
rece
ption
c
onsist
s
of
t
he
energy
nee
ded
t
o
operate
the
c
ircuit
on
ly
[24].
E
Rx
=
E
e
le
c
×
L
…
…
(5)
The
c
os
t
of
ene
rg
y
for dat
a a
ggre
gatin
g
is,
E
DA
−
tot
=
S
×
k
×
E
da
…
…
(6)
w
he
re
s
repres
ents
the
sign
al
nu
m
ber
that
is
gathe
red
a
nd
E
(d
a
)
represe
nts
th
e
ener
gy
f
or
each
bit
sp
ent
in
the
process
of a
ggr
egati
ng [2
2].
Fig
ure
2. First
-
order ra
dio m
od
el
[2
0]
2.4
.
An
t
li
on
op
timi
z
er (
ALO)
The
AL
O
is
a
n
al
go
rithm
that
is
able
t
o
de
te
rm
ine
the
m
os
t
s
uitable
de
sign
s
f
or
a
m
ajo
r
portio
n
of
the
e
xisti
ng
tr
aditi
on
al
e
ngineeri
ng
issue
s.
This
idea
po
ints
out
that
this
ty
pe
of
al
gorithm
has
va
rio
us
adv
a
ntage
s
in
fin
ding
so
l
utio
ns
to
c
onstrai
ne
d
pr
ob
le
m
s
in
m
any
researc
h
fiel
ds.
T
he
ALO
al
gorith
m
’s
cor
e
insp
irat
io
n
is
init
ia
ll
y
fr
om
the
f
or
a
ging
beh
a
vior
of
a
nt
-
li
on’s
la
r
va
e.
Consi
de
rab
l
y,
the
co
ne
e
dg
e
is
sat
isfact
or
il
y
s
harp
for
i
ns
ect
s
to
be
t
rappe
d
without
a
ny
dif
ficult
y.
It
i
s
kn
own
t
hat
An
t
Lio
n
Op
ti
m
iz
e
r
al
gorithm
enco
ur
a
ges
inte
racti
on
s
to
ta
ke
pl
ace
betwee
n
a
nts
an
d
ant
li
ons
in
a
s
nar
e
.
Fo
r
t
his
intera
c
ti
on
to
happe
n,
a
nts
a
r
e
re
qu
ire
d
t
o
w
al
k
in
t
he
sea
rc
h
s
pace
w
her
e
antli
on
s
are
al
lowe
d
t
o
c
hase
them
and
m
ake
us
e
of
t
he
m
os
t
su
it
able
sn
a
res.
F
or
c
om
pu
ta
ti
on
pu
rposes
,
th
e
rand
om
m
ov
e
m
ent
of
a
nt
is
caref
ully
chos
en
f
or
m
od
el
ing
a
nts’ m
ov
e
m
ents. T
his is
fur
the
r u
nd
e
rstoo
d
[25]
,
X
(
t
)
=
[
0
,
cums
um
(
2r
(
t
1
−
1
)
,
…
,
cumsum
(
2r
(
t
n
−
1
)
)
(7)
r
(
t
)
=
{
1
if
rand
>
0
.
5
0
if
ra
nd
<
0
(8)
w
he
re
t
ind
ic
at
es
the
ste
ps
of
rand
om
walk
and
ra
nd
re
pr
es
ents
a
rando
m
nu
m
ber
create
d
in
interval
of
[0
,
1].
The
be
hav
i
our
noti
ced
by
A
ntli
on
s
’
hunt
s
hows
that
t
he
rand
om
walk
us
e
d
m
ay
po
ss
ibly
var
y
ar
ound
the
so
urce
(r
e
d
c
ur
ve).
It
m
ay
ei
t
her
ha
ve
a
de
s
cend
i
ng
be
ha
vi
or
(a
blu
e
cu
r
ve),
or
a
n
i
ncrea
sing
tren
d
(a
black
curve)
. In
cons
equ
e
nce,
the
a
nts’ p
os
it
ion i
s
no
te
d
a
nd appl
ie
d
th
rou
g
h o
pt
i
m
iz
ation
in
th
e n
e
xt m
at
rix
[
23
]
,
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c
En
g
&
Co
m
p
Sci
IS
S
N:
25
02
-
4752
Ener
gy
ef
fi
ci
en
t WSN
us
in
g hy
br
id
mod
if
ic
ati
on PE
GA
SIS
w
it
h
ant l
io
n…
(
Ah
med Ab
dul
Azeez As
m
ael)
277
M
An
t
=
[
A
1
,
1
A
1
,
2
…
A
1
,
d
A
2
,
1
⋮
A
2
,
2
⋮
…
A
2
,
d
⋱
⋮
A
n
,
1
A
n
,
2
…
A
n
,
d
]
…
.
.
(9)
w
he
re
M(A
nt)
is
the
m
at
rix
to
reco
r
d
the
fitn
ess
of
eac
h
ant
,
A(
i,
j
)
represe
nts
the
value
of
the
j
-
t
h
dim
e
ns
io
n
of
t
he
i
-
th
a
nt
;
f
dem
on
strat
es
the
ob
j
ect
ive
f
unct
ion
a
nd
finall
y
n
r
e
pr
ese
nts
the
nu
m
ber
of
a
nt
s.
It
is
su
ggest
e
d
to
ke
ep
in
m
ind
that
possi
bly
a
nts
m
ay
hid
e
so
m
ewh
ere
i
n
the
sear
ch
sp
a
ce.
T
o
m
ai
ntain
thei
r
locat
ion
s
and
f
it
ness
-
values, t
hese m
at
rices are
us
e
d
,
M
An
t
lion
=
[
AL
1
,
1
AL
1
,
2
…
AL
1
,
d
AL
2
,
1
⋮
AL
2
,
2
⋮
…
AL
2
,
d
⋱
⋮
AL
n
,
1
AL
n
,
2
…
AL
n
,
d
]
…
.
.
(10)
wh
e
re
M
(Ant
-
l
ion)
is
t
he
m
at
rix
to
kee
p
each
ant
-
li
on
’s
post,
A
L(i,
j
)
in
dica
te
s
the j
-
t
h
dim
ensio
n’s v
al
ue
o
f
i
-
th
antli
on
,
d
is
the
num
ber
of
var
ia
bles
(
dim
ensio
n)
a
nd
la
s
tl
y
bu
t
no
t
le
as
t
is
n
wh
ic
h
re
pr
es
e
nts
the
nu
m
ber
of antli
ons
.
M
OA
=
[
f
(
[
AL
1
,
1
AL
1
,
2
AL
1
,
d
]
)
f
(
[
AL
2
,
1
AL
2
,
2
AL
2
,
d
]
)
⋮
f
(
[
AL
n
,
1
AL
n
,
2
AL
n
,
d
]
)
]
…
.
.
(11)
M(O
A)
is
t
he
2D
ar
ray
tha
t
m
ai
ntains
the
fitness
of
e
ver
y
a
ntli
on
,
AL(
i,
j)
hi
gh
li
ghts
the
j
-
t
h
dim
ension
’s
va
lue
of
i
-
th
a
nt
li
on
,
f
dep
ic
ts
the
obj
ect
ive
functi
on
an
d
f
inall
y
n
repres
ents
the
num
ber
of
antli
on
s
.
D
ur
i
ng
the
process
of
op
ti
m
iz
ation
,
al
l
Ra
ndom
walks
a
re
de
pe
nd
e
nt
on
(
7
)
.
The
idea
be
hind
su
c
h
process
is
that
ants
w
ork
by
updat
in
g
t
heir
locat
ion
s
with
rand
om
walk
a
t
each
ste
p
of
op
ti
m
iz
ation
.
Sinc
e
each
searc
h
spa
ce
has
a
boundar
y,
ne
ve
rthel
ess,
(
7
)
can
not
be
us
ed
im
m
e
diate
ly
to
up
da
te
the
ants’
posit
ion.
More
ov
e
r,
to
m
ai
ntain
rand
om
walks
inside
the
s
pace
of
s
earch
,
norm
al
i
zat
i
on
ta
kes
place
w
her
e
(
12
)
is
us
e
d
(m
in
–
m
ax
nor
m
al
iz
ation
) [
23]
,
X
i
t
=
(
X
i
t
−
a
i
)
×
(
d
i
−
c
i
t
)
(
d
i
t
−
a
i
)
…
…
(12)
w
he
re
a(i
)
port
rays
the
m
ini
mu
m
ran
dom
walk,
rep
res
ents
the
m
axi
m
u
m
of
(i)
-
t
h
va
riab
le
at
t
-
th
it
eration
and
la
stl
y
re
pr
ese
nts
the
m
ini
m
u
m
of
w
hole
va
riables
a
t
(t)
-
th
it
erati
on.
Eac
h
it
erati
on
is
e
xpect
ed
to
e
m
plo
y
(
13
)
t
o
guara
ntee
that
rand
om
w
al
ks
are
pe
rfo
rm
ed
inside
the
sp
ace
of
s
earch
.
From
wh
at
is
com
pr
ehe
nd
e
d,
antli
on
s
’
tra
ps
hav
e
an
i
nf
l
ue
nce
on
rand
om
walks
of
ants.
Th
us
,
t
o
m
od
el
this
assum
pti
on,
a
m
at
he
m
at
ic
a
l m
et
ho
d
is
u
se
d as
bo
t
h
(
13
)
a
nd
(
14
)
pro
pos
e
,
c
i
t
=
Antlion
i
t
+
c
t
…
.
.
(13)
d
i
t
=
Antlion
j
t
+
d
t
…
.
.
(14)
b
ot
h
(
14
)
a
nd
(
15
)
dem
on
str
at
e
that
ants
ar
bitraril
y
wal
k
i
n
a
hype
r
sphe
re
re
so
l
ved
by
the
vecto
rs
c
and
d
arou
nd
a
sel
ect
ed
a
nt
li
on.
Fi
g
ure
3
il
lustrat
es
a
the
or
et
ic
al
m
od
el
li
nk
ed
to
this
be
hav
i
or
w
he
re
it
dis
pl
ay
s
a
search
s
pace
c
on
sist
in
g
of
tw
o
dim
ension
s.
It
is
note
d
that
ants
a
re
requir
ed
to
tra
ns
fe
r
i
ns
i
de
a
hyper
s
ph
e
re
arou
nd
the
sel
e
ct
ed
ant
li
on.
Using
the
s
ugge
ste
d
m
echan
ism
s,
the
pan
ts
c
an
f
or
m
traps
pro
per
t
o
their
fi
tness
le
vel.
More
ov
e
r,
ants
are
re
qu
ired
to
m
ov
e
arou
nd
ra
ndom
l
y.
On
the
ot
her
hand,
ants
rele
ase
sands
ou
t
of
the
ho
le
ce
nter
as
so
o
n
as
they
re
al
iz
e
that
an
ant
is
trapp
e
d
in
the
sn
are
.
Thi
s
form
of
act
ion
dro
ps
the
stu
ck
ant
that
is
atte
m
pti
ng
to
r
un
away.
To
m
od
el
su
ch
be
hav
i
or
re
ga
rd
i
ng
m
at
he
m
at
ic
s,
the
ran
do
m
walking
ra
diu
s
of
the an
ts
d
ec
rea
ses, in an
ad
a
pt
ive m
ann
er.
T
he
se
(
15
)
a
nd
(
16)
w
e
re
pro
vide
d
in
this as
pec
t
,
c
t
=
c
t
I
…
…
(15)
c
t
=
d
t
I
…
…
(16)
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
23
, N
o.
1
,
Ju
ly
2021
:
273
-
2
8
4
278
Fig
ure
3. A
n
a
nt r
a
ndom
ly
w
al
kin
g i
ns
i
de
a
n
a
ntli
on’s
tra
p
The
la
st
ph
ase
of
huntin
g
ta
ke
s
place
wh
e
n
a
n
a
nt
ar
rives
at
the
bott
om
of
the
hole
a
nd
is
stuck
i
n
a
n
antli
on’s
j
a
w.
Fo
ll
owin
g
this,
the
ant
can
drag
the
a
nt
into
the
san
d
an
d
in
gest
it
s
body
f
ully
.
Fo
r
e
nc
ourag
i
ng
su
c
h
proce
dur
e,
pr
ey
-
cat
chin
g
is
assum
ed
to
occ
ur
w
hen
ants
get
m
or
e
com
petent
to
go
into
t
he
san
d
th
a
n
their
op
pone
nt
.
A
fter
t
hat,
t
he
a
ntli
o
n
is
exp
ect
e
d
to
al
te
r
it
s
posit
ion
to
the
la
te
st
ant
-
huntin
g
place
to
i
m
pr
ove it
s op
portu
nity
to hunt a
new prey
. In
term
s o
f wha
t was
discu
sse
d,
(
17
)
is a
ppli
ed
,
Antlion
j
t
=
Ant
i
t
if
f
(
Ant
i
t
)
>
f
(
Ant
li
on
j
t
)
…
.
.
(17)
w
he
re
t
re
prese
nts
the
e
xisti
ng
it
erati
on
,
a
nt
i hig
hligh
ts
t
he
s
po
t of
i
-
th
a
nt at
t
-
th
it
erati
on
and
antli
on j
r
efers
to the si
te
of th
e sele
ct
ed j
-
t
h antl
ion
at
t
-
th i
te
rati
on
.
The
te
rm
Eli
t
ism
is
reg
ar
ded
as
a
m
ajo
r
feat
ur
e
f
ro
m
ev
olut
ion
ary
al
gorithm
s
that
has
the
a
bili
ty
t
o
pro
vid
e
best
s
olu
ti
ons
f
or
an
y
sta
ge
at
the
process
of
opt
i
m
iz
ation
.
I
n
this
stud
y,
the
finest
antli
ons
wer
e
store
d
in
each
it
erati
on
an
d
nam
ed
as
el
it
es.
Con
side
ra
bly,
ensurin
g
that
the
el
it
e
is
the
fitt
est
antli
on
,
it
can
po
s
e
an
im
pact
on
ants
’
m
ov
e
m
e
nt
ov
e
rall
by
it
erati
on
.
He
nce,
a
n
ass
umpti
on
is
in
dicat
ing
w
hich
sta
t
es
that
each
ant
m
ov
e
s
rand
om
l
y
around
a
s
pecifie
d
antli
on
by
the
roulet
te
wh
eel
and
t
he
el
it
e
si
m
ultaneou
sly
.
This
is sh
own
i
n
(
18
)
,
Ant
i
t
=
R
A
t
+
R
E
t
2
…
…
.
(18)
wh
e
re
R
(
t,
A)
is
the
rand
om
m
ov
e
m
ent
around
the
a
ntli
on
sel
ect
ed
by
th
e
roulet
te
wh
e
el
at
t
-
th
it
eratio
n,
R
(t,E)
points
ou
t
the
ra
ndom
m
ov
em
ent
arou
nd
the
el
it
e
at
t
-
th
it
erati
on,
a
nd
la
stl
y
An
t
(t,i)
ref
e
rs
to
t
he
s
po
t o
f
i
-
th a
nt
at
t
-
th
it
erati
on
.
3.
METHO
DOL
OGY
On
e
of
t
he
m
os
t
essenti
al
as
pects
po
si
ng
a
n
im
pact
on
hi
gh
-
qu
al
it
y
cl
ust
ering
is
the
sim
ul
ta
neo
us
consi
der
at
io
n
of
the
tw
o
crit
eria
con
sist
in
g
of
the
distanc
es
within
a
cl
us
te
r
(inter
-
cl
ust
er)
an
d
the
di
sta
nces
within
tw
o
or
m
or
e
cl
us
te
rs
(intra
-
cl
us
te
r
)
.
Most
rec
omm
end
ed
m
et
hods
in
dicat
e
that
on
ly
one
of
these
crit
eria
is
co
nsi
der
e
d
or
e
ve
n
none
ha
ve.
Wh
en
e
xam
ining
the
m
et
ho
ds
at
wh
ic
h
thes
e
cri
te
ria
are
per
cei
ved,
bo
t
h
qual
it
y
and
pr
eci
si
on
of
the
cl
us
te
r
are
no
t
com
pu
te
d
at
the
end
sta
ge
of
cl
us
te
rin
g.
Ce
rtai
nly,
err
or
rate
in
cl
us
te
r
pro
duct
ion
is re
garded
a
s
one o
f
t
he
m
os
t
sign
ifi
cant
pro
blem
s
i
m
pacti
ng
hi
gh
-
qual
it
y
cl
us
te
ring
. In
this
w
ork,
so
m
e
reco
m
m
end
a
ti
on
s
a
re g
ive
n
to
be
able
t
o
m
easur
e
the
qu
al
it
y
of
the
cl
ust
er.
T
hese
c
rite
ria
are
li
ste
d
,
a)
T
he
first
cri
te
ri
on
de
pends
on
the
densi
ti
es
within
a
so
le
cl
us
te
r
a
m
ongst
two
or
m
or
e
cl
us
te
rs.
I
n
the
case w
her
e
the
crit
erion is sm
al
l, the cluste
r’s qual
it
y i
s co
nsi
der
e
d bett
er.
b)
The
second
cr
iteri
on
points
ou
t
a
n
erro
r
ta
king
place
du
ring
cl
us
te
rin
g.
T
he
act
io
n
of
arr
a
ng
i
ng
node
s
in
a
cl
us
te
r
le
ads
to
the
high
-
qu
al
it
y
cl
us
te
ring
c
rite
ria.
To
further
unde
rs
ta
nd
the
i
dea,
if
m
or
e
regul
a
r
nodes
c
ov
e
r
th
e
su
r
face
of
th
e
cl
us
te
r,
the
c
luster
will
be
balance
d
great
ly
,
and
no
des’
consum
ption
of
energy
will
d
e
c
rease.
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c
En
g
&
Co
m
p
Sci
IS
S
N:
25
02
-
4752
Ener
gy
ef
fi
ci
en
t WSN
us
in
g hy
br
id
mod
if
ic
ati
on PE
GA
SIS
w
it
h
ant l
io
n…
(
Ah
med Ab
dul
Azeez As
m
ael)
279
c)
The
th
ir
d
cri
terion
em
ph
asi
zes
the
nodes’
balance
withi
n
the
cl
us
te
rs
.
The
preci
sio
n
range
li
nk
e
d
t
o
this crit
erio
n
is
establi
sh
e
d
ac
cordin
g
t
o bo
t
h m
easur
e
m
ent error an
d
sta
ti
s
ti
cal
assu
m
ption
s.
As
th
e crit
eria
is o
verviewe
d and com
pr
e
hended, the
cl
us
te
r
ing
proces
s is e
xp
la
ine
d st
ep
-
by
-
ste
p
,
a)
The
first
step
:
N
sens
or
no
de
s
sp
rea
d
ra
ndom
ly
(n
on
-
deter
m
inist
ic
dep
loy
m
ent)
in
a
sq
uar
e
fiel
d
with
a
le
ng
th
of M
wi
th a stat
ic
BS.
b)
The seco
nd
s
t
ep
: Eve
ry no
de
h
as
a c
on
sta
nt
a
m
ou
nt
of e
nergy E
o.
c)
The
th
ir
d
st
ep
:
O
nce
the
no
des
a
re
s
pr
ea
d,
the
m
ap
is
cl
ust
ered
i
nto
ci
r
c
ular
reg
i
on
s
en
su
ri
ng
they
are
arou
nd
the
bas
e
sta
ti
on
.
Ty
pical
ly
,
the
first
reg
i
on
w
ould
c
on
ta
in
al
l
node
s
with
le
ss
than
20m
fr
om
the
base
sta
ti
on.
T
hen,
the
sec
on
d
re
gion
woul
d
co
ntain
a
ny
nod
e
that
is
in
the
ra
ng
e
of
21
-
40m
fr
om
th
e
base
sta
ti
on
.
L
ast
ly
,
the
third
reg
i
on
w
ould
c
on
ta
in
al
l
node
s
hav
in
g
41
-
60
m
fr
om
the
base
sta
ti
on
.
This
patte
rn is co
nti
nu
e
d.
d)
The
fo
ur
th
st
ep
:
On
ce
the
c
luster
m
ap
is
c
om
plete
d,
al
l
n
od
e
s
are
li
nk
e
d
in
betwee
n
wh
e
re
each
no
de
connects
to
it
s
neig
hbori
ng
node
i
n
it
s
reg
i
on.
F
or
e
xam
ple,
a
node
fou
nd
i
n
the
sec
ond
re
gion
w
ou
ld
connect
t
o
it
s
nea
rest
node
in
t
he
first
r
egio
n.
The
la
st
ste
p
w
ou
l
d
be
em
plo
yi
ng
the
An
t
li
on
op
ti
m
iz
ation
on
the
m
od
ifie
d
PE
GA
S
IS.
Ov
e
rall
,
this
would
j
ust
ify
the
connecti
on
of
no
des
an
d
i
m
pr
oves li
feti
m
e o
f
t
he node
s.
Su
m
m
ary
fo
r
t
he pr
opos
e
d n
ode
a)
All sensi
ng no
des
a
re
posit
ion
ed
r
a
ndom
ly
and co
ntin
ue b
ei
ng
sta
ti
c eve
n
a
fter
bein
g
l
oc
at
ed.
b)
Each se
nsor
no
de possesse
s
a
m
at
chless ID to
be di
sti
nguis
hed f
ro
m
o
the
r
nod
e
s.
c)
It can
not re
pla
ce the
batte
ry of all
the se
ns
i
ng
nodes
that
possess the
sam
e p
rim
ary energy.
d)
The
m
ap
is cl
ust
ered
i
nto
ci
rc
ular regi
ons aroun
d
the
b
a
se s
ta
ti
on
(
BS
).
e)
Each
node
is c
onnected
to
it
s
m
os
t neighboring n
ode
within
t
he
a
rea that is
precede
d.
f)
The A
nt li
on optim
iz
at
ion
is app
li
ed
on t
he m
od
ifie
d
PE
G
AS
I
S
wh
e
re
no
des
a
re success
fu
ll
y
interl
ink
e
d, an
d
thei
r
li
feti
m
e
is m
assivel
y dev
el
oped
.
g)
It is possi
ble to
adjust
transm
i
ssion p
ower
of
sens
or
s
d
e
pe
nding
on c
omm
u
nicat
ion
d
ist
an
ce.
3.1.
An
t
li
on
op
timi
z
at
i
on
(
ALO)
On
ce
the
m
odel
of
m
od
ifie
d
PEGAS
IS
is
pr
e
par
e
d
s
ucce
ssfu
ll
y,
the
ant
li
on
is
a
ppli
ed
as
ca
n
be
seen
in
al
gori
thm
(1
).
To
de
velo
p
the
li
f
et
i
m
e
of
energy,
any
ope
ra
tors
rec
omm
e
nd
e
d
in
the
previ
ou
s
su
bse
ct
ions
we
re
us
e
d.
The
refor
e
,
it
is
po
ssi
ble
no
t
to
de
fine
the
AL
O
op
tim
iz
at
ion
al
gorithm
.
Ho
wev
e
r,
the
ALO
al
gorith
m
is
con
sider
ed
as
a
func
ti
on
ha
ving
three
tu
ples
f
un
ct
io
ns
that
est
i
m
a
te
wo
rld
wide
op
ti
m
iz
ation
i
s
su
es.
T
his is
outl
ined f
ur
t
her
,
Com
pre
hendi
ng
ALO a
lgorithm
:
MP
EGAS
I
S re
fe
rs t
o
th
e
st
art po
pula
ti
on
Mea
sure
th
e
f
it
n
ess of
ant l
ions
Choose
the m
ost suit
ab
le
ant l
ion
(assum
e
as
th
e e
li
te
)
W
hil
e
th
e end cr
it
eri
on
in
unassured y
e
t
For e
ac
h
an
t
l
ion
Emplo
y
the Rou
le
tte
whe
el for th
e
sel
ec
t
ed ant li
o
n
Update
p
ara
m
ete
rs c
and
d
in acc
orda
nce wit
h
(
16
)
(
17
)
.
Gene
rate an
un
a
rra
nged
w
al
k
then norm
al
ize it in
accorda
n
ce wit
h
(
7
)
and
(
13
)
.
Inform
the
n
ew
l
oca
t
ion
of
the a
n
t
li
n
e
End
loop
for
Com
pute
all
ant
’
s li
on
f
it
ness
Exc
hang
e an ant
li
on
with
i
ts c
or
r
el
a
te
ant t
urn
into fi
t
te
r
(18)
Upgrade
th
e el
i
t
e
in
th
e
c
ase
wh
en
an
an
tl
ion
is f
it
te
r
th
an the
elit
e
End
whil
e
Ret
urn elite
4.
RESU
LT
S
AND DI
SCUS
S
ION
Im
po
rtant
ste
ps
to
be
ta
ke
n
a
re
to
both
te
st
and
e
valuate
t
he
prese
nted
w
ork
in
this
st
udy.
This
is
in
te
rm
s
of
ene
r
gy
co
nsum
ption
in
PE
GASI
S
prot
oco
l
.
Al
l
te
sts
li
nk
ed
to
the
perf
orm
ed
ex
per
im
e
nts
an
d
si
m
ulati
on
s
we
r
e
car
ried
out
us
in
g
Ma
tl
ab
2019
.
As
a
rem
i
nd,
this
pro
po
s
ed
w
ork
f
ocu
s
ed
on
hybri
d
a
nt
li
on
al
gorithm
wit
h
m
od
ifie
d
PEG
A
SIS.
The
pr
oce
dure
be
gan
by
deter
m
ining
the
m
ap
siz
e
(200*
200
m
2
)
fo
ll
owe
d
by
ra
ndom
distribu
t
ion
of
nodes
a
rou
nd
the
base
sta
ti
on
.
T
he
posit
io
n
was
10
0,100
a
s
no
ti
c
ed
in
Figure
4. F
or
t
he param
et
ers
of this
wor
k,
T
able 1 ca
n b
e
use
d.
To
cl
us
te
r
the
m
ap
into
ci
rcul
ar
reg
i
on
s
,
co
m
pu
ti
ng
the
E
uclidean
distan
ce
betwee
n
the
nodes
a
nd
the
BS
is
require
d
to
be
ca
rr
ie
d
out.
A
fterw
a
rd
s
,
the
ci
r
cular
re
gions
are
di
vid
e
d
w
her
e
t
he
first
regi
on
con
ta
in
s all
no
des wit
hin a
distance
of
less
than 2
0m
f
ro
m
the BS,
t
he
sec
ond re
gion c
onta
ins
al
l n
od
es
withi
n
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
23
, N
o.
1
,
Ju
ly
2021
:
273
-
2
8
4
280
a
distance
of
20
to
40m
fr
om
the
BS
an
d
s
o
on
the
rest
of
the
re
gions.
Figure
5
cl
ea
rly
sh
ows
this
a
sp
ect
.
Subseque
ntly
,
li
nk
in
g
the
node
s
occurs
w
here
var
io
us
no
de
s
betwee
n
reg
i
on
s
wer
e
inte
rlink
e
d.
T
his
is
wh
e
r
e
each
no
de
wa
s
co
nn
ect
e
d
t
o
it
s
near
est
no
de
in
th
e
area
that
pr
ece
de
d
it
.
Th
is
can
be
see
n
certai
nly
in
Fig
ure
6.
On
c
e
nodes
a
re
li
nk
e
d,
t
he
ant
l
ion
op
ti
m
iz
ati
on
is
em
plo
ye
d
to
reduce
e
ne
rg
y
c
on
s
um
ption
vi
a
j
ust
ify
ing
t
he
li
nk no
de.
Fig
ure
7
pro
ves
t
hat
this co
nce
pt to be c
orrect.
Fig
ure
4. Ra
nd
om
d
ist
ribu
ti
on
of
n
odes
Table
1.
WSN
si
m
ulati
on
p
a
r
a
m
et
ers
for
th
e
prop
os
ed
pr
oto
cols
P
a
ram
et
er
s
V
a
l
u
e
s
T
h
e
A
re
a
200*200
2
I
n
i
t
i
al
e
n
er
g
y
0
.
0
0
3
N
u
m
b
er
o
f
n
o
d
e
s
200
Bit
_
r
a
t
e
1
M
b
/
s
e
c
P
a
c
k
e
t
Le
n
g
t
h
500
(
B
y
t
e
)
M
e
d
i
a
Ac
c
e
ss
C
o
n
t
r
o
l
L
ay
er
I
EE
E
8
0
2
.
1
1
No
te
: Packet siz
e r
ep
resents
the a
m
o
u
n
t of
data th
at collected
f
ro
m
the sen
so
rs
Fig
ure
5. Cl
us
t
erin
g of
t
he
m
ap
int
o
ci
rc
ular
reg
i
on
s
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c
En
g
&
Co
m
p
Sci
IS
S
N:
25
02
-
4752
Ener
gy
ef
fi
ci
en
t WSN
us
in
g hy
br
id
mod
if
ic
ati
on PE
GA
SIS
w
it
h
ant l
io
n…
(
Ah
med Ab
dul
Azeez As
m
ael)
281
Fig
ure
6. Lin
kin
g n
odes
with
oth
e
r nodes
Fig
ure
7. A
ppli
cat
ion
of A
L
O i
n
hy
br
i
d
MPE
GASIS
5.
PERFO
R
MANC
E
METR
I
CS
T
he
r
e
a
s
o
n
be
hi
nd
us
i
ng
pe
r
f
or
m
a
nc
e
m
et
r
i
c
s
i
s
t
o
a
na
l
y
se
a
nd
t
e
s
t
ne
t
w
o
r
ks
a
c
c
or
di
ng
t
o
s
pe
c
i
f
i
c
c
r
i
t
e
r
i
a
t
o
ju
s
t
i
f
y
t
he
s
im
ul
a
te
d
s
c
e
na
r
i
os
'
be
ha
vi
or
.
I
n
pa
r
t
i
c
ul
a
r
,
w
he
n
pe
r
f
or
m
a
nc
e
m
e
t
ri
c
s
a
r
e
us
e
d,
t
he
im
pr
ov
e
m
e
nt
le
ve
l
of
t
he
pr
e
sented
p
r
ot
oc
o
l
s
i
n
c
om
pa
r
i
s
on
t
o
ot
he
r
on
e
s
i
s
i
ll
us
t
r
a
t
e
d.
A
ny
of
t
he
c
l
us
t
e
r
i
ng
pr
ot
oc
ol
s
t
ha
t
a
im
t
o
r
e
vi
e
w
a
nd
e
va
l
ua
t
i
ng
t
e
s
t
s
of
m
os
t
W
S
N
s
hi
gh
l
i
g
ht
t
he
im
po
r
t
an
c
e
of
e
va
l
ua
t
i
ng
H
N
D
(
ha
l
f
n
od
e
s
di
e
)
a
nd
F
N
D
(
f
i
r
s
t
no
de
di
e
)
i
n
t
he
ne
t
w
or
k.
B
e
s
i
de
s
,
t
he
a
ve
r
a
ge
di
s
s
i
pa
t
e
d
e
ne
r
gy
w
ou
l
d
be
a
l
s
o
i
nc
l
ud
e
d.
I
t
i
s
a
gr
e
e
d
o
n
t
ha
t
no
s
e
ns
or
a
c
t
i
vi
t
y
w
it
hi
n
t
he
ne
t
w
or
k
i
s
pe
r
c
e
i
ve
d.
T
he
s
t
a
bi
l
it
y
pe
r
i
od
in
t
he
c
l
us
t
e
r
i
ng
pr
ot
oc
ol
i
s
ve
r
i
f
i
e
d
f
or
f
ol
l
ow
i
n
g:
a)
F
i
r
s
t
Node
to
Die
:
The
t
i
m
e
that
the 1
st
t
dead
node
app
ea
rs.
b)
H
a
l
f
Nodes
to
D
i
e
:
T
he
t
im
e
be
t
w
e
e
n
ne
t
w
or
k
op
e
r
a
t
i
on
s
t
a
r
t
i
ng
t
o
t
he
t
im
e
of
ha
l
f
no
de
s
be
i
ng
de
a
d.
A
s
t
he
ne
t
w
or
k
l
i
f
e
t
im
e
of
th
e
pr
op
os
e
d
p
r
ot
oc
ol
w
a
s
c
o
nf
i
r
m
e
d
i
n
c
om
pa
ri
s
on
t
o
ot
he
r
o
ne
s
,
a
nu
m
be
r
of
pe
rfor
m
ance
m
e
a
sur
e
s
,
c
om
pr
i
s
in
g
a
ve
r
a
ge
e
n
e
r
gy
c
on
s
u
m
p
t
i
on
,
p
r
od
u
c
t
i
vi
t
y
,
p
ac
k
e
t
d
e
l
i
ve
r
y
r
at
i
o
a
n
d
e
n
d
-
to
-
e
n
d
d
e
l
ay
,
c
a
n
be
di
s
c
ov
e
r
ed
,
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
23
, N
o.
1
,
Ju
ly
2021
:
273
-
2
8
4
282
1)
A
ve
r
a
ge
T
h
r
o
u
gh
p
u
t
:
T
hr
ou
gh
pu
t
i
s
us
e
d
t
o
m
e
a
s
ur
e
data
pack
et
s
that
are
receive
d
suc
cessf
ully
,
p
e
r
un
it
t
im
e
vi
a
(
19
)
.
Avg
.
t
hroug
hput
=
(
z
/
t
o
tal
ti
me
)
…
.
.
(19)
Wh
e
re
z
repres
ents the
n
et
of
su
ccess
fu
ll
y re
cei
ved
pac
kets.
2)
E
n
d
to
E
n
d
D
e
l
a
y
:
Re
fer
s
to
t
h
e
t
i
m
e
ta
ken
be
t
w
e
e
n
r
e
c
e
p
t
i
o
n
a
nd
transm
issi
on
of
data
pack
et
s
.
It
can
be
cal
c
ulate
d from
(
20
)
.
Ave
rag
e
en
d
to
en
d
delay
=
∑
(
T
i
r
−
T
i
s
)
m
m
i
=
1
…
(20)
Wh
e
re
(
−
)
in
dic
at
es
the
va
riat
ion
in
rece
ptio
n
a
nd
tra
ns
m
is
sion
tim
e
of
th
e
pac
ket
a
nd
m
in
the total
nu
m
ber
of
t
ran
sm
it
ted
pack
et
s
.
3)
Av
er
age
ener
gy
c
onsumed
:
Energy
c
on
s
um
pt
ion
happe
ns
in
tra
ns
m
i
tt
ing
,
receivin
g,
a
n
d
proces
sin
g
of
the
pac
ket
by
each
node
th
r
ough
the
operati
on
of
the
netw
ork.
E
ne
rg
y
c
onsu
m
ption
i
n
e
ach
no
de
(i
)
is
worked
out f
r
om
this form
ula:
Eci
=
ET
+
EPi
+
Eri
(21)
Wh
e
re
c
i
po
i
nt
s
ou
t
t
he
t
ot
a
l
c
on
s
um
e
d
en
e
r
gy
,
r
e
f
e
r
s
t
o
t
he
c
on
s
um
e
d
e
ne
r
gy
th
r
o
ug
h
t
he
pr
oc
e
s
s
of
t
r
a
n
sm
it
ti
ng
,
i
s
t
he
c
o
ns
um
e
d
e
ne
r
gy
t
hr
ou
gh
pr
oc
e
s
s
i
ng
of
pa
c
ke
t
,
i
s
t
he
c
o
ns
um
e
d
e
ne
r
g
y
thr
ough
t
he
pr
oc
e
s
s
of
r
e
c
e
pt
i
on
a
n
d
f
i
na
l
l
y
(
i
)
i
s
vi
e
w
e
d
a
s
qu
a
nt
i
t
y
of
t
he
no
de
s
.
F
i
n
di
ng
ou
t
t
he
a
ve
r
a
ge
di
s
s
i
pa
t
e
d
e
ne
r
gy
a
l
s
o m
e
a
n
s
f
i
nd
i
ng
t
he
pe
r
c
e
nt
a
ge
of
t
he
t
ot
a
l
ne
t
w
or
k
e
ne
r
gy
-
c
o
ns
um
pt
i
on
t
o
t
he
(
n)
no
de
s
.
Ave
rag
e
Ene
rg
y
Co
nsump
t
io
n
=
∑
E
ci
n
n
i
=
1
…
.
(22)
W
he
r
e
i
s
t
he
c
on
s
um
e
d
e
ne
r
gy
of
t
he
i
-
t
h
no
de
a
n
d
n
r
e
pr
e
s
e
nt
s
t
he
w
ho
l
e
n
od
e
of
w
i
r
e
l
e
s
s
s
e
ns
o
r
ne
t
w
or
ks
.
4)
P
ac
k
e
t
d
e
l
i
ve
r
y
r
at
i
o
:
Re
f
ers
to
the
r
at
e
of c
orrect pac
kets
bein
g deli
ver
e
d
PDR
=
(
z
m
)
∗
100
…
.
(
23
)
Table
2
pro
vide
s
per
ce
ntage
enh
a
ncem
ent
reg
ar
ding
H
D
N
for
both
pro
pose
d
prot
oco
ls
MPEG
AS
I
S
and
MPE
GASI
S
with
A
LO
a
nd
with
ot
her
protoc
ols
for
t
he
sce
na
ri
o
(B
S
bei
ng
at
m
idd
le
).
The
us
e
of
HDN
ind
ic
at
es
that
consi
der
a
ble
a
m
ou
nt
of
at
te
nt
ion
is
giv
e
n
as
the
appr
ox
i
m
at
ed
value
f
or
the
a
ver
a
ge
of
both
li
fetim
e
and
per
f
or
m
ance
of
t
he
net
work.
H
ence,
Ta
ble
2
on
ly
f
ocuses
on
H
DN
as
pect
s.
Id
e
al
ly
,
com
par
i
ng
the
pe
rce
ntage
enh
a
ncem
ent
f
or
t
he
pro
po
se
d
MPE
GASI
S w
it
h
ALO
an
d
MPEG
AS
I
S wi
tho
ut both
sce
nar
i
os
is necessa
ry.
In Table
3, the
de
ad perce
ntage
of the
prese
nted work
are
s
ourced.
As
can
be
vie
wed
from
Table
2,
the
am
ou
nt
of
im
pr
ove
m
ent
fo
r
eac
h
case
in
M
P
E
C
A
S
IS
-
AL
O,
wh
e
re
HND
ha
s
i
m
pr
oved
i
n
M
P
E
C
A
S
I
S
-
A
L
O
ov
e
r
M
P
E
C
A
S
I
S
w
i
t
ho
ut
t
he
us
e
of
A
L
O
by
5.
2
0%
.
I
n
t
e
rm
s
of
Ne
t
w
o
r
ks
L
i
f
e
t
im
e
,
t
he
e
nh
a
nc
em
e
nt
wa
s
by
a
n e
s
ti
m
a
t
e
of
8.
66
%
.
M
or
e
ov
e
r
,
t
he
T
hr
o
ug
hp
ut
e
nh
a
nc
em
e
nt
w
a
s
a
bo
ut
4.
8
0
%
i
n c
om
pa
r
i
s
on
t
o E
n
d T
o E
nd
De
l
a
y
e
nh
a
nc
em
e
nt
wh
i
c
h w
a
s
a
bo
ut
5.
2
9%
.
O
ve
r
a
l
l
,
t
he
Av
g.
C
on
s
um
e
d
(
r
e
m
a
i
ne
d)
e
nh
a
n
c
em
e
nt
w
a
s
a
bo
ut
3.
60
%
w
he
r
e
t
he
E
ne
r
gy
E
nd
t
o
E
n
d
D
e
l
a
y
e
nh
a
n
c
em
ent
r
e
s
ul
t
e
d
i
n
3.
60
%
.
F
or
t
hi
s
s
t
ud
y
,
a
ny
im
pa
c
t
of
s
c
a
l
a
bi
l
it
y
im
po
s
e
d
on
M
P
E
G
A
S
I
S
a
n
d
M
P
E
G
A
S
I
S
w
i
t
h
A
L
O
i
s
a
na
l
y
z
e
d
reli
a
bly
.
T
hi
s
s
t
ud
y
ha
s
em
ph
a
s
iz
e
d
l
a
r
ge
-
s
c
a
l
e
ne
t
w
or
ks
.
T
h
e
r
e
f
or
e
,
o
nl
y
W
S
N
-
2
m
od
e
l
i
s
f
ul
ly
c
on
s
i
de
r
e
d
.
In
t
e
rm
s
of
s
im
ul
at
i
on
re
s
ul
t
s
,
T
ab
l
e
3 pr
e
s
e
nt
s
t
he
fi
nd
i
ng
s
w
h
e
r
e
i
t
ou
t
li
ne
s
th
e
nu
m
be
r
of
r
ou
nd
s
ob
s
e
r
ve
d
f
o
r
1
%
,
2
5%
,
50
%
a
nd
70
%
of
n
o
de
de
a
t
h.
I
t
i
s
ob
s
e
r
ve
d
t
ha
t
a
l
t
e
r
i
ng
t
he
nu
m
be
r
of
t
he
di
f
f
us
e
d
no
de
s
ha
s
no
i
m
pa
c
t
on
t
he
n
e
t
w
or
k
’
s
l
i
f
e
t
im
e
w
hi
c
h
i
s
de
s
i
r
a
bl
e
.
Table
2.
Perce
ntage
e
nhan
ce
m
ent f
or
MPE
CASIS
-
AL
O
M
e
tr
i
c
M
PE
C
AS
I
S
-
AL
O
M
PE
C
AS
I
S
HNDs
65
.30
%
6
0
.10
%
N
e
t
w
o
r
k
s
Li
fe
t
i
m
e
5
9
.86
%
5
1
.20
%
T
h
r
o
u
g
h
p
u
t
8
7
.30
%
8
2
.50
%
E
n
d
To
E
n
d
Dela
y
2
9
.20
%
2
3
.91
%
A
v
g
.
Co
n
su
m
ed
(
r
em
ai
n
e
d
)
E
n
e
r
g
y
4
4
.10
%
4
0
.50
%
E
n
d
To
E
n
d
Dela
y
5
3
.00
%
4
9
.40
%
Evaluation Warning : The document was created with Spire.PDF for Python.