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.
9
, No
.
3
,
J
un
e
201
9
, pp.
1944
~
19
56
IS
S
N:
20
88
-
8708
,
DOI: 10
.11
591/
ijece
.
v
9
i
3
.
pp1944
-
19
56
1944
Journ
al h
om
e
page
:
http:
//
ia
es
core
.c
om/
journa
ls
/i
ndex.
ph
p/IJECE
Impr
ovem
ent of
c
rank
shaft MA
C
p
r
oto
co
l
for
w
i
r
eless
s
ens
or
n
etworks
:
a
s
imu
lation
s
t
ud
y
Madina
H
ami
an
e
1
,
M
ak
i
ya
J.
Ah
me
d
2
1
Depa
rt
m
ent
of
Te
l
ec
om
m
unic
ation
Eng
ine
er
ing,
Ahlia
Univer
si
t
y
,
Bahr
ai
n
2
Depa
rtment
of
Inform
at
ion
T
echnolog
y
(pre
v
io
usl
y
)
,
Ahli
a
Uni
ver
sit
y
,
Bahr
ai
n
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
Ma
r
8
, 2
01
8
Re
vised
Oct
22
, 2
01
8
Accepte
d
Nov
20
, 201
8
Due
to
the
dr
amatic
growth
in
t
he
use
of
W
ire
l
ess
Sensor
Network
(W
S
N)
appl
i
ca
t
ions
ran
ging
from
envi
ronm
ent
and
habi
ta
t
m
onit
oring
to
tra
ck
ing
and
surveillance,
net
work
rese
arch
in
W
SN
protoc
ols
has
be
en
ve
r
y
ac
t
ive
in
the
la
st
d
ecade.
W
it
h
batter
y
-
powere
d
sensors
oper
at
ing
in
unat
te
nd
e
d
envi
ronm
ent
s,
e
ner
g
y
conse
rv
at
i
on
bec
om
es
the
ke
y
technique
fo
r
improving
W
SN
li
fet
imes.
W
SN
Medium
Acc
ess
Control
(MA
C)
protoc
ols
addr
ess
ene
rg
y
aw
are
ne
ss
and
red
uce
d
dut
y
c
y
c
le
s.
Th
e
foc
us
of
thi
s
stud
y
is
t
o
inve
stigate
,
thro
ugh
sim
ula
ti
on
,
t
he
eff
ect
of
v
ariati
ons
in
var
ious
factors
th
a
t
infl
uen
ce
th
e
pe
rform
anc
e
resul
t
s
of
W
S
Ns
.
Us
i
ng
MiXiM
fra
me
work
with
OM
NeT
++
sim
ula
tor
,
th
is
si
m
ula
ti
on
stud
y
proposes
m
odifi
c
at
ions
in
Cranksha
ft
MA
C
protoc
o
l
in
or
der
to
improve
i
ts
per
form
ance.
The
impact
of
dura
ti
on
and
num
ber
of
slots,
degr
e
e
of
con
nec
t
ivi
t
y
among
the
nodes
,
m
obil
ity
sp
ee
d
a
nd
m
obil
ity
upd
at
e
in
te
rv
al
and
al
so,
the
impact
of
sending
dat
a
p
ac
k
et
s
wi
thout
pre
ambles
are
inve
stig
at
e
d.
Based
on
the
sim
ula
ti
o
n
result
s,
a
n
imp
rove
d
ver
sion
of
the
Cranksh
aft
protoc
o
l
fo
r
W
SN
i
s
suggested
and
a
compara
ti
v
e
stud
y
of
th
e
per
fo
rm
anc
es
of
the
origi
nal
and
improved
protocol
is
pre
sente
d
.
The
resul
ts
cl
e
ar
l
y
indi
c
ate
the
s
uper
iorit
y
o
f
the
improved
pr
otoc
ol
over
it
s or
igi
nal ve
rsi
on
.
Ke
yw
or
d
s
:
Cran
kshaft M
AC pr
oto
c
ol
Energy
-
a
war
e
w
irel
ess
s
e
nsor
MAC
p
ro
t
oc
ol
Netw
orks
Copyright
©
201
9
Instit
ut
e
o
f Ad
vanc
ed
Engi
n
ee
r
ing
and
S
cienc
e
.
Al
l
rights re
serv
ed
.
Corres
pond
in
g
Aut
h
or
:
Ma
din
a
Ham
ian
e,
Dep
a
rtm
ent o
f Tel
ecom
m
un
ic
at
ion
E
nginee
rin
g,
Ah
li
a
Un
i
ver
si
ty
,
Ma
nam
a, Bahrai
n
.
Em
a
il
:
m
ha
m
ian
e@a
hlia.ed
u.bh
1.
INTROD
U
CTION
A
wireless
se
ns
or
net
wor
k
(
W
S
N)
is
a
wireless
net
work
c
onsist
ing
of
s
patia
ll
y
distribu
te
d
auto
no
m
ou
s
de
vices
usi
ng
s
ens
or
s
t
o
m
on
it
or
ph
ysi
cal
or
en
vironm
ental
co
nd
it
io
ns
.
A
ty
pical
WSN
s
yst
e
m
inco
rpor
at
es
a
gateway
that
pro
vid
es
wireles
s
connecti
vity
back
to
the
wi
r
ed
w
or
ld
a
nd
di
stribu
te
d
node
s
[1
]
.
Since
t
he
rad
i
o
is
the
WSN
c
om
po
ne
nt
that
m
os
t
con
s
um
e
s
ene
r
gy
an
d
is
us
e
d
m
or
e
fre
qu
e
ntly
w
he
n
higher
data
rates
are
t
o
be
m
on
it
or
e
d,
this
entai
ls
m
or
e
powe
r
co
nsum
ption
.
I
n
t
he
case
of
dense
W
S
N,
a
very
la
rg
e
nu
m
ber
of
se
nsor
nodes
c
omm
un
ic
at
e
with
each
ot
her
a
nd
sen
se
their
e
nv
i
ronm
ent
wirelessl
y.
The
t
ask
of
changin
g
t
he b
at
te
ry p
oweri
ng this
larg
e
nu
m
ber
o
f n
od
e
s
can th
e
refo
re
be
com
e cum
ber
so
m
e.
Wh
e
n
in
opera
ti
on
,
nodes
i
n
WSN
co
nsum
e
energy
w
he
n
the
ra
di
o
is
s
witc
hed
on
due
to
seve
ral
reasons
,
the
m
os
t
i
m
po
rtant
a
re
Id
le
li
ste
ning,
O
verhea
rin
g,
and
P
ro
t
oco
l
ov
e
r
head
[
2].
To
this
extent,
m
any
effor
ts
hav
e
be
en
dir
ect
ed
to
wards r
e
duci
ng
en
er
gy cons
um
pt
ion
which
r
esulte
d
in th
e de
velo
pm
ent o
f a
w
ide
range
of
ene
r
gy
eff
ic
ie
nt M
A
C protoc
ols
[3
]
-
[
12
]
.
Cran
kshaft
is
consi
der
e
d
to
be
one
of
the
best
ene
rg
y
-
ef
f
ic
ie
nt
MAC
protoc
ols
f
or
de
ns
e
wi
reless
sens
or
netw
or
ks
[13],
[
14]
.
Mi
ni
m
u
m
la
ten
cy
an
d
m
axim
u
m
ener
gy
eff
ic
ie
ncy
a
re
a
chieve
d
with
t
he
us
e
of
two
m
ulti
ple
a
ccess
protoc
ols
,
nam
el
y
TDMA
an
d
CSM
A.
This
protoc
ols
com
bin
at
ion
di
vid
es
the
tim
e
into
fr
am
es
wh
ere
each
f
ram
e
con
ta
ins
slots
a
nd
a
num
ber
of
slots
are
assig
ned
t
o
the
br
oa
dcast
traf
fics
and
t
he
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Impr
ovem
e
nt
of
crankshaft
M
AC pr
oto
c
ol fo
r wi
rel
ess sen
s
or
networks
: a
simu
l
ation
… (
Ma
dina
Hamia
ne
)
1945
rem
ai
nin
g
one
s
to
the
unic
ast
traff
ic
s.
Alth
ough
Cra
nksh
a
f
t
was
sho
wn
to
m
ini
m
iz
e
the
idle
li
ste
nin
g
,
hen
ce
energy
co
ns
um
ption,
i
t
su
ff
e
rs
howev
e
r
f
or
m
energy
con
s
um
pt
ion
due
to
the
sam
pling
of
the
broa
dcast
slots,
in addit
ion t
o t
he
li
m
it
ed
flexi
bili
ty
o
f
the
pr
oto
c
ol due t
o
t
he use
of T
DMA.
In
dense
W
S
N
s
,
each
n
ode
c
an
ha
ve
a
la
r
ge
nu
m
ber
of
ne
ighbou
rs
that
exceed
s
the
5
-
10
neig
hbors
that
are
assum
ed
in
ge
ne
ral
pur
po
se
M
AC
protoc
ols
[
15
]
.
Prob
le
m
s
can
there
f
or
e
of
t
en
occ
ur
w
hen
us
in
g
these
gen
e
ral
pur
po
se
proto
cols
in
den
se
WSNs,
the
m
os
t
im
po
rtant
of
these
pr
ob
l
e
m
s
are
Ov
e
r
hear
i
ng,
Com
m
un
ic
at
io
n
gro
upin
g,
O
ver
-
pr
ov
isi
oni
ng,
an
d
Neig
hbou
r
sta
te
[1
6]
,
[1
7].
Cra
nk
sh
aft
ad
dresse
s
these
pro
blem
s
by
l
et
ti
ng
the
node
s
tur
n
off
the
ir
rad
i
os
al
te
rnat
el
y
rather
th
an
sim
ultaneou
sly
with
the
aim
to
reduce the
ov
e
rh
ea
rin
g
an
d
c
omm
un
ic
at
ion
g
r
oupi
ng. A
ls
o, sche
du
li
ng th
e receivin
g
slot
s r
at
her
t
han
ke
epin
g
neig
hbor’s
stat
e re
du
ces
the
pro
visio
ning.
In
t
he
ori
gi
nal
i
m
ple
m
entat
io
n
of
Cra
nksh
a
f
t
[15],
two
t
raffic
patte
r
ns
,
flooding
a
nd
c
onve
r
ge
cast
,
wer
e
co
ns
i
dered
w
hen
ev
al
ua
ti
n
g
the
proto
col
perform
ance.
It
was
concl
ud
e
d
that
oth
er
pr
oto
c
ols
pro
vide
bette
r
delive
ry
rati
o
than
C
r
anksha
ft
prot
oc
ol,
but
at
th
e
exp
e
ns
e
of
higher
e
nergy
con
s
um
ption
wh
e
n
com
par
ed
to
C
ranks
haf
t.
It
w
as
al
so
dem
onstrat
ed
that
Cr
anksha
ft
co
uld
prov
i
de
lo
w
de
li
ver
y
rati
o
at
high
and
m
ediu
m
m
essage
rates
an
d
that
it
s
la
te
ncy
incr
eased
qu
ic
kly
when
t
he
l
oad
excee
de
d
0.1
5
m
essage/node/
sec.
Ther
e
f
or
e
sig
nificant im
pr
ove
m
ent in th
e
cra
nk
s
ha
ft pr
oto
c
ol p
e
rfo
rm
ance can
be o
btaine
d by:
a.
Im
pr
ov
i
ng the
delivery
rati
o
i
n broa
dcast fl
ood patt
er
n.
b.
Im
pr
ov
i
ng lat
ency in c
onve
r
ge
cast
p
at
te
r
n.
2.
OBJECTI
VE
S OF
THE ST
UDY
Fo
r
the
im
pr
ovem
ent
of
WSN
li
feti
m
e,
the
i
m
ple
m
ented
protoc
ols
s
houl
d
the
refor
e
overco
m
e
the
afore
m
entione
d
li
m
it
ation
s.
This
pap
e
r
pr
e
sents
a
sim
ulatio
n
st
ud
y
of
C
ranks
haf
t
prot
oco
l
with
t
he
aim
to
i
m
pr
ove
it
s
pe
r
form
ance
in
te
rm
s
of
ene
rg
y
eff
ic
ie
ncy,
la
te
ncy
an
d
sl
ot
co
ll
isi
on
s.
T
he
ef
fect
of
se
ver
al
oth
e
r
par
am
et
ers
affe
ct
ing
data
rate
an
d
t
hroug
hput,
cha
nn
el
r
el
ia
bili
ty
and
netw
ork
t
opol
og
y
is
in
vestigat
ed
.
Chan
ges
in
th
e
nu
m
ber
of
broa
dcast
slots,
total
nu
m
ber
of
slots,
slot
durati
on,
m
ob
il
it
y
sp
eed,
m
ob
il
it
y
update
inte
rv
al
an
d
play
gro
und
siz
e,
un
der
broa
dcast
a
nd
c
onve
rg
ecast
pa
tt
ern
s,
i
n
a
ddit
ion
to
t
he
num
ber
of
nodes
unde
r
the
br
oad
cast
pa
tt
ern
s
are
c
onside
red
a
nd
their
ef
fect
on
protoc
ol
pe
rfo
rm
ance
inv
est
igate
d.
Althou
gh
the
r
adio
com
m
un
ic
at
ion
is
the
hig
he
st
energy
con
s
um
ption
f
unct
ion
of
the
s
ens
or
no
des,
e
ach
of
the
ab
ov
e
par
a
m
et
ers
al
so
affe
ct
en
erg
y
e
ff
i
ci
ency.
The
pe
rfor
m
ance
of
t
he
im
pr
ov
e
d
C
ranks
haf
t
pr
op
t
oco
l
is
com
par
ed
t
o
th
at
o
f
the
or
i
gina
l protoc
ol im
ple
m
entat
ion
.
3.
METHO
DOL
OGY
In
t
his
stu
dy
c
ranks
haf
t
pro
tocol
was
im
ple
m
ented
us
i
ng
Mi
XiM2
.0
/
OMNeT++
4.1.
As
i
n
the
or
i
gin
al
stu
dy
,
and
to
al
lo
w
f
or
the
c
om
par
ison,
the
netw
ork
was
sim
ulate
d
us
in
g
96
node
s
in
90×5
0
m
et
ers
la
yout
with
ra
dio
range
of
25
m
et
ers
and
node
s
po
wer
e
d
at
3V.
T
he
bro
adcast
fl
ood
pa
tt
ern
with
5
di
ff
ere
nt
nu
m
ber
s
of
flo
od
pe
r
seco
nd,
and
t
he
co
nve
rg
ecast
patte
rn
with
8
dif
fer
e
nt
num
ber
s
of
(m
essages/nod
e/
sec)
wer
e
us
e
d
i
n
t
he
sim
ulate
d
WSN.
To
ens
ur
e
a
valid
c
om
par
ison
,
t
he
si
m
ulati
on
wa
s
car
ried
out
usi
ng
t
he
sam
e g
ener
al
s
et
ti
ng
s
of
t
he o
rigin
al
im
ple
m
entat
ion
.
As
m
entione
d
above,
the
pr
ot
oco
l
c
onsu
m
es
li
tt
le
ener
gy
bu
t
with
a
very
low
delive
ry
rati
o.
T
he
netw
ork
was
t
her
e
fore
sim
ulate
d
afte
r
c
ha
ng
i
ng
the
val
ue
s
of
th
e
af
fec
ti
ng
par
am
et
ers.
T
he
fo
ll
owi
ng
li
st
re
prese
nts the
t
est
ed param
et
e
rs
unde
r broad
cast
and c
onve
rg
ecast
patte
r
ns:
a.
Nu
m
ber
of br
oa
dcast slots
, (d
at
a t
ran
s
fer
rat
e ||
throug
hput)
b.
Total
num
ber
of slots,
(data
trans
fer
rate ||
thro
ughput)
c.
Slot du
rati
on, (data t
ra
ns
fe
r ra
te
||
thr
ou
ghput)
d.
Mob
il
it
y speed
, (netw
ork
t
opol
og
y)
e.
Mob
il
it
y updat
e inter
val
(n
et
work to
polo
gy)
f.
Play
groun
d
siz
e (net
work to
po
l
og
y
)
In ad
diti
on, th
e
num
b
er of nod
es (net
w
ork
t
opol
og
y
),
unde
r broa
dcast p
at
te
rn w
as
also
test
ed.
4.
SIMULATI
O
N RESULTS
AND A
NA
L
Y
SIS
The
e
ff
ect
s
of
the
ide
ntifie
d
pa
par
am
et
ers
unde
r
bo
t
h
broa
dcast
an
d
co
nv
e
recast
traf
fic
we
re
inv
est
igate
d
a
nd a
naly
sed
i
n
t
he follo
wing.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
9
, N
o.
3
,
June
201
9
:
1944
-
1956
1946
4.1.
Ef
fect o
f Dur
at
i
on
an
d
Nu
mber
of
Slot
s
To
exam
ine
the
i
m
pact
of
the
reserved
ti
m
e
for
each
node
to
sen
d
or
recei
ve
pac
kets,
the
nu
m
ber
of
slots in eac
h fram
e and
t
he d
urat
ion o
f
eac
h
s
lot we
re c
hang
ed
in
tu
rn.
4.1.1.
C
hangi
ng
th
e
t
otal n
umber
of slo
ts
, with s
ame
n
umber
of br
oadc
as
t
slo
ts
In
orde
r
to
e
xam
ine
the
i
m
pact
of
r
eser
ve
d
tim
e
fo
r
each
node
to
sen
d
or
receive
pac
kets
on
the
thr
oughput
,
th
e
protoc
ol
was
te
ste
d
by
cha
ngin
g
the
total
num
ber
of
slots in
each f
ram
e
unde
r
both b
r
oa
dcast
and
c
onve
rg
e
c
ast
traff
ic
and the
delive
ry r
at
io and en
e
r
gy co
nsum
ption
were exam
ined.
The
sim
ulati
on
s w
er
e
desig
ne
d
to
in
vestigat
e
this
eff
ect
by
dec
r
easi
ng
the
t
otal
nu
m
ber
of
sl
ots
by
3
(
7
slo
ts,
2
broa
dcast
and
5
un
ic
ast
)
a
nd
in
creasin
g
t
he
tot
al
num
ber
of
sl
ots
by
3
al
so
(
13
slots,
2
bro
adcast
a
nd
11
un
ic
ast
)
.
Fig
ure
1
an
d
Figure
2
cl
ea
rly
ind
ic
at
e
a
highe
r
delive
ry
rati
o
f
or
de
creasin
g
num
ber
of
sl
ots
a
nd
a
l
ow
e
r
e
nerg
y
consum
ption
f
or increa
sin
g n
um
ber
of sl
ots
unde
r broadca
s
t t
raf
fic
.
Figure
1. Deli
ve
ry Rat
io un
de
r
Br
oa
dcast tra
ff
ic
Figure
2. Ene
r
gy cons
um
ption
unde
r
Br
oa
dc
ast
traff
ic
It
can
al
so
be
no
ti
ced
t
hat
there
is
a
sat
ur
at
ion
poi
nt
in
the
gr
a
ph
at
flo
od
2,
for
w
hich
the
delive
r
y
rati
o
rem
ai
ns
const
ant.
Sim
ilarly
,
Fig
ur
e
3
and
Fig
ur
e
4
s
how
that
unde
r
co
nver
gecast
patte
rn,
Cra
nksha
ft
exh
i
bits
a
high
er
delivery
rati
o
for
decr
easi
ng
nu
m
ber
of
sl
ots
a
nd
an
ene
r
gy
co
nsum
ption
that
i
ncr
ea
se
s
with
a d
ec
rease i
n
t
otal n
um
ber
of
slots.
Figure
3. Deli
ve
ry r
at
io
unde
r C
onvergecast
t
raffic
Figure
4. Ene
r
gy Co
nsum
ption
unde
r
Co
nve
rg
ecast
traff
ic
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Impr
ovem
e
nt
of
crankshaft
M
AC pr
oto
c
ol fo
r wi
rel
ess sen
s
or
networks
: a
simu
l
ation
… (
Ma
dina
Hamia
ne
)
1947
The
la
te
ncy is
disp
la
ye
d
i
n
Fi
gure
5 wh
ic
h
ind
ic
at
es li
tt
le
eff
ect
of
t
otal n
um
ber
of slots
exp
ect
at t
he
m
idd
le
loa
ds
wh
e
re
t
he
pac
ket
e
xch
a
nge
i
s
m
or
e
act
ive
and
the
cha
nc
es
f
or
colli
sio
ns
a
nd
packet
loss
is
m
or
e, r
es
ulti
ng in
a
n
inc
rease
in lat
ency .
Figure
5. Lat
en
cy
u
nde
r
Co
nv
erg
ecast
t
raffic
4.1.2.
C
hangi
ng
number
of
broadc
as
t
slo
t
s
In
flo
od
i
ng
t
ra
ff
ic
,
broa
dcast
slots
are
m
or
e
act
ive
than
t
he
un
ic
ast
on
e
s.
The
nex
t
set
of
si
m
ulati
on
dem
on
strat
es
t
he
e
ff
ect
of
c
ha
ng
i
ng
the
nu
m
ber
of
broa
dc
ast
slots
from
1
t
o
3
on
delivery
rati
o
a
nd
energy
consum
ption
,
unde
r
bo
t
h
broad
ca
st
an
d
c
onve
rg
e
cast
pa
tt
ern
s.
Fig
ur
e
6
a
nd
Fi
gure
7
cl
early
i
nd
ic
at
e
that
Cran
kshaft
del
ivery
rati
o
dec
reases
with
t
he
nu
m
ber
of
br
oad
ca
st
slots
due
to
the
redu
ct
ion
of
the
e
f
fecti
ve
us
e
of
the
slot
s
durin
g
the
c
yc
le
.
Accord
i
ngly
,
ene
rg
y
c
onsu
m
ption
i
ncrea
sed
with
inc
rea
sin
g
deli
very
rati
o
and d
ec
reases
wh
e
n deli
ve
ry
rati
on d
ec
rease
s.
Figure
6. Deli
ve
ry r
at
io
unde
r B
ro
a
dcast tra
ffi
c
Figure
7. Ene
r
gy Co
nsum
ption
unde
r
Br
oa
dc
ast
traff
ic
Figure
8,
Fig
ur
e
9
an
d
Fig
ur
e
10
s
how
t
he
re
su
lt
s
of
va
ryi
ng
the
nu
m
ber
of
broad
c
ast
slot
s
un
der
t
he
conve
rg
ecast
pa
tt
ern
,
a
nd
in
di
cat
e
that
Cran
ks
ha
ft
e
xh
i
bits
a
lowe
r
deliv
ery
rati
o
with
si
m
ple
drop
i
n
the
energy c
on
s
umpti
on and
sli
ght i
ncr
ease i
n
t
he
la
te
ncy.
These
res
ults
are
un
e
xpect
e
d
a
nd
ar
e
du
e
to
t
he
pac
ket
loss
pro
blem
that
us
ua
ll
y
occurs
in
t
he
conve
rg
ecast
proces
s
for
wir
el
ess
sensor
ne
tworks
or
igi
na
ti
ng
from
the
congesti
on
a
nd
coll
isi
on
s
ne
ar
the
sink
node
[18].
Energy
co
nsu
m
pt
ion
an
d
la
te
ncy
increase
or
decr
ea
se
wi
th
increasi
ng
or
dec
reasin
g
de
li
ver
y
rati
o
re
sp
ect
ive
ly
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
9
, N
o.
3
,
June
201
9
:
1944
-
1956
1948
Figure
8.
Deli
ve
ry Rat
io un
de
r
Co
nver
gecast
traf
fic
Figure
9.
Ene
r
gy Co
nsum
ption
unde
r
Co
nve
rg
ecast
traff
ic
Figure
10.
Lat
e
ncy un
der Co
nver
gecast t
ra
ff
i
c
4.1.3
.
C
hangi
ng
sl
ot dur
at
i
on
Figure
11
an
d
Figure
12
di
sp
la
y
the
delivery
rati
o
a
nd
energy
consu
m
pt
ion
f
or
inc
reasin
g
slo
t
durati
on
under
broa
dcast
tra
f
fic.
F
or
lo
wer
loads
the
deli
ver
y
rati
o
is
l
ow
e
r,
w
hile
f
or
highe
r
loa
ds,
the
delivery
rati
o
and
c
onseq
ue
nt
ly
the
energy
con
s
um
ption
increase
be
cau
se
increasin
g
slot
durati
on
a
ll
ow
s
m
or
e tim
e to the
sen
ding a
nd re
cei
vin
g sl
ots
to
se
nd
or
recei
ve packets
.
Figure
11.
Deli
ver
y R
at
io
unde
r
Br
oa
dcast tr
aff
ic
Figure
12.
E
ne
rg
y C
on
s
um
ption u
nder
Br
oa
dcast
traff
ic
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Impr
ovem
e
nt
of
crankshaft
M
AC pr
oto
c
ol fo
r wi
rel
ess sen
s
or
networks
: a
simu
l
ation
… (
Ma
dina
Hamia
ne
)
1949
Unde
r
co
nver
ge
cast
patte
rn
a
nd
sm
aller
slot
durati
ons,
C
r
anksha
ft
ex
hibi
ts
a
higher
de
li
ver
y
rati
o
un
ti
l
the
rad
i
o
pack
et
s
sta
rt
colli
din
g
a
nd
the
sink
no
lo
ng
e
r
receiv
es
al
l
data
[1
9],
bu
t
a
highe
r
energ
y
con
su
m
ption
a
nd
le
ss
la
te
nc
y
are
exp
ect
e
d.
W
he
n
inc
rea
sing
the
sl
ot
durati
on
ho
wever,
the
deliv
er
y
rati
o
beco
m
es
lower
un
ti
l
the
rate
of
pack
et
s
inc
r
eases
in
the
ne
twork
a
nd
sl
ot
durati
on
is
ef
fici
ently
us
ed,
th
e
n
delivery
rati
o
i
ncr
ease
s.
T
his
is co
nf
irm
ed
b
u
the
p
l
ots in
F
igure
13, F
i
gur
e 14 a
nd Fig
ur
e 15.
Figure
13.
Deli
ver
y R
at
io
unde
r
Co
nver
geca
st t
raf
fic
Figure
14.
E
ne
rg
y C
on
s
um
ption
un
der
Conver
gecast t
raffic
Figure
15.
Lat
e
ncy
un
der Co
nver
gecast t
ra
ff
i
c
4.2.
Ef
fect o
f
degree
of
c
onnecti
vit
y am
ong
th
e
nodes
(
changin
g the
playgr
ou
n
d si
z
e)
To
te
st
the
e
ffec
t
of
t
he
c
onne
ct
ivit
y
of
no
de
s
on
the
deliv
ery
rati
o
of
t
he
netw
ork,
the
play
gro
und
siz
e,
as
on
e
of
the
par
am
et
ers
wh
ic
h
can
af
fect
the
de
gr
ee
of
co
nnect
ivit
y,
was
cha
ng
e
d
in
the
sim
ul
at
io
n.
Figure
16
an
d
Figure
17
sho
w
that
unde
r
broa
dcast
patte
r
n,
by
de
creasi
ng
or
increasi
ng
the
play
groun
d
siz
e,
that
co
ntains
t
he
nodes,
the
delivery
rati
o
beco
m
es
le
ss
in
the
l
ow
e
r
l
oa
ds
because
ea
ch
node
will
ha
ve
le
ss
nu
m
ber
of
ne
ighb
or
s
,
hen
c
e
the
c
olli
sion
s
will
be
le
ss,
so
t
her
e
is
on
ly
a
sli
gh
t
c
ha
ng
e
in
the
protoc
ol
perform
ance
with
no
cha
ng
e
in
the
ene
rgy
con
s
um
ption.
A
sim
il
ar
res
ult
is
obta
ine
d
w
hen
dec
reas
ing
t
he
play
gro
und
siz
e;
the d
el
ivery
rati
o
sli
gh
tl
y i
nc
reases w
it
hout
ch
an
ge
in
the
energy co
nsum
ption
beca
us
e
each
node
will
ha
ve
m
or
e
nieg
hbours
,
hen
ce
the
r
e
will
be
m
or
e
pack
et
colli
sio
ns
s
o
a
sli
ght
c
hange
i
n
t
he
de
li
ver
y
rati
o
ca
n be
noti
ced.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
9
, N
o.
3
,
June
201
9
:
1944
-
1956
1950
Figure
16. Del
ivery Rat
io
un
der Br
oad
cast
t
raffic
Figure
17. En
erg
y C
onsu
m
ptio
n u
nd
e
r
B
ro
a
dcast
traff
ic
Fo
r
the
sam
e
reasons
as
i
n
th
e
broa
dcast
patte
rn
,
m
od
ify
in
g
play
gro
und
s
iz
e
in
co
nver
ge
cast
patte
r
n
has
a
sli
ght
e
ff
ect
on
t
he
prot
oco
l
perf
orm
ance.
I
n
bo
t
h
cases
pe
rform
ance
i
m
pr
ov
e
m
ent
is
neg
a
ti
vely
aff
ect
ed
with
li
ttle
inc
rease
in
delive
ry
r
at
io
,which
le
ads
to
worse
la
te
ncy
and
e
nergy
co
nsum
ption.
Figure
18, Fi
gure
19, an
d Fi
gure
20 sh
ow th
is be
hv
i
our.
Figure
18. Deli
ver
y R
at
io
unde
r
Co
nver
geca
st t
raf
fic
Figure
19. En
erg
y C
onsu
m
ption
unde
r
Conver
gecast t
raffic
Figure
20. Late
ncy un
der Co
nver
gecast t
ra
ff
i
c
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Impr
ovem
e
nt
of
crankshaft
M
AC pr
oto
c
ol fo
r wi
rel
ess sen
s
or
networks
: a
simu
l
ation
… (
Ma
dina
Hamia
ne
)
1951
4.3.
Ef
fect o
f
degree
of c
onnecti
vit
y am
ong
th
e
nodes
(
changin
g the
number
of n
odes)
Ra
isi
ng
the nu
m
ber
o
f
node
s in
broa
dcast fl
ooding leads t
o i
ncr
easi
ng the
nu
m
ber
of n
ei
ghbo
rs
cl
os
e
to each
o
t
her,
hen
ce
the
pac
ke
t l
os
s
will
g
o dow
n
a
nd cons
equ
e
ntly
the
de
li
ver
y rati
o g
oe
s up,
als
o decr
easi
ng
nu
m
ber
of
node
s
decr
eases
th
e
chan
ces
f
or
colli
sion
s
an
d
hen
ce
inc
rease
s
the
delivery
rati
o
as
il
lustrate
d
i
n
Figure
21.
Whi
le
the
energy
consum
ption
i
n
both
ca
ses
does
no
t
inc
reas
e
as
show
n
i
n
Figure
22,
thi
s
m
eans
bette
r per
form
ance
for
the
pr
oto
c
ol.
Figure
2
1
. Del
ivery Rat
io
unde
r
Br
oa
dcast tr
aff
ic
Figure
22. En
erg
y C
onsu
m
ption
unde
r
B
ro
a
dcast
traff
ic
4.4.
Ef
fect o
f
degree
of m
obi
li
ty
(
ch
angin
g mobil
it
y
s
pee
d)
The
m
ob
il
it
y
sp
ee
d
par
am
eter
is
one
of
the
netw
ork
t
opol
og
y
facto
rs
.
U
nder
br
oadcast
traf
fic,
increasin
g
or
decr
easi
ng
t
he
m
ob
il
it
y
sp
eed
of
the
node
s
will
aff
ect
the
nodes
c
onnecti
vity
ei
ther
by
increasin
g
or
de
creasin
g
t
he
di
sta
nce
bet
wee
n
the
nodes,
th
us
inc
reasin
g
or
dec
reasin
g
t
he
c
ha
nces
of
pack
et
colli
sion
s,
a
nd
hen
ce
an
inc
r
ease
or
a
decre
ase
in
the
delivery
rati
o
or
energy
consu
m
pt
ion
.
T
he
re
su
lt
s
of
si
m
ulati
on
sho
wn
i
n
Fig
ur
e
23
a
nd
Fig
ur
e
24,
ind
ic
at
e
a
sli
gh
t
i
m
pr
ov
e
m
ent
in
the
prot
oco
l
pe
rfo
r
m
ance
wh
e
n
m
ob
il
it
y
sp
ee
d
of
the
nodes
is
var
ie
d,
wh
il
e
the
e
nergy
co
nsum
pti
on
rem
ai
ns
relat
ively
un
c
ha
ng
e
d
,
wh
ic
h
le
ad
s to
a b
et
te
r pe
rform
ance for
t
he pr
oto
c
ol.
Figure
23. Del
ivery Rat
io
unde
r
Br
oa
dcast t
r
aff
ic
Figure
24. E
ne
rg
y C
on
s
um
ption u
nder
Br
oa
dcast
traff
ic
Unde
r
c
onverg
ecast
traf
fic,
th
e
m
ob
il
it
y
sp
eed
par
am
et
er
ha
s
the
sam
e
eff
ect
as
the
play
gro
und
siz
e
.
Be
cause
the
m
ob
il
it
y
sp
ee
d
of
the
node
will
aff
ect
t
he
num
ber
of
neig
hbors;
decr
easi
ng
m
ob
il
it
y
sp
eed
will
hav
e
ne
gative
eff
ect
on
deliv
ery
rati
o,
w
he
reas
inc
reasin
g
m
ob
il
it
y
sp
eed
will
incre
as
e
the
pro
ba
bili
ty
of
colli
sion
s
that
hav
e
al
so
ne
ga
ti
ve
ef
fect
on
t
he
delivery
rati
o.
He
nce
li
tt
le
eff
ect
is
seen
on
la
te
ncy
a
nd
e
nerg
y
consum
ption
wh
e
n
the
m
ob
il
it
y
sp
eed
is
de
creased
or
in
creased
.
T
his
is
il
lustrate
d
in
Figure
25,
Fi
g
ure
26
and Fig
ure
27.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
9
, N
o.
3
,
June
201
9
:
1944
-
1956
1952
Figure
25.
Deli
ver
y R
at
io
unde
r
Co
nver
geca
st
traff
ic
Figure
26.
E
ne
rg
y C
on
s
um
ption u
nder
Conver
gecast t
raffic
Figure
27.
Lat
e
ncy un
der Co
nver
gecast t
ra
ff
i
c
4.4.
Ef
fect o
f
degree
of m
obi
li
ty
(
ch
angin
g mobil
it
y
u
pd
at
e
inter
va
l
)
Unde
r
broad
ca
st
traff
ic
a
nd
i
n
high
loa
ds,
t
he
routers
nee
d
to
be
up
date
d
ver
y
of
te
n
to
i
m
pr
ove
t
he
perform
ance
[
20
]
,
s
o
Cr
an
kshaft
s
hows
higher
delive
ry
ra
ti
o
in
the
high
er
loa
ds
by
shorte
ning
the
m
ob
il
it
y
update
inter
val
.
Wh
il
e
in
low
loads
the
delivery
rati
o
decre
ases,
beca
us
e
the
protoc
ol
over
hea
d
will
increas
e
by
increasi
ng
nu
m
ber
of
up
dating
ti
m
es
with
sm
all
num
ber
of
pac
ke
ts
in
the
netw
ork,
as
il
lustra
te
d
in
Figure
28.
On
the
oth
e
r
ha
nd
increasin
g
m
ob
il
it
y
interval
t
i
m
e
raises
the
delivery
rati
o,
because
the
pr
oto
c
ol
ov
e
r
head will
be decrease
d.
The
e
nergy c
onsu
m
ption wil
l n
ot ch
a
nge i
n bo
t
h
cases
as i
nd
ic
at
ed
in
Fig
ur
e
29.
Fig
ure
28: Deli
ver
y R
at
io
unde
r
Br
oa
dcast tr
aff
ic
Figure
29. E
ne
rg
y C
on
s
um
ption u
nder
Br
oa
dcast
traff
ic
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Impr
ovem
e
nt
of
crankshaft
M
AC pr
oto
c
ol fo
r wi
rel
ess sen
s
or
networks
: a
simu
l
ation
… (
Ma
dina
Hamia
ne
)
1953
Var
yi
ng
m
ob
il
it
y
up
date
interval
in
co
nv
e
r
gecast
patte
rn
has
an
ef
fect
si
m
i
la
r
to
that
of
br
oad
cast
flo
od
i
ng
patte
r
n
on
the
pr
otoc
ol
perf
or
m
ance.
Chan
ging
m
ob
il
it
y
update
inter
val
decr
ea
s
es
the
delivery
rati
o
in
the
low
lo
a
ds
.
I
n
c
on
tra
st,
in
high
loa
ds
var
yi
ng
m
ob
il
ity
up
date
in
te
rv
al
increas
e
d
the
delive
ry
rati
o,
because
in
hi
gh
e
r
loa
ds
th
e
routers
nee
d
to
be
upda
te
d
ver
y
of
t
en
to
im
pr
ove
the
perf
orm
ance.
Fr
om
Figure
30, F
i
gure
31 and Figure 3
2,
it
is o
bs
e
rv
e
d
tha
t t
her
e is no
c
ha
ng
e i
n
ene
rg
y
co
nsum
ption
a
nd
no
sign
ific
a
nt c
ha
ng
e
in
t
he pr
ot
oco
l l
at
e
ncy.
Figure
30. Del
ivery Rat
io
unde
r
Co
nver
geca
st t
raf
fic
Figure
31. E
ne
rg
y C
on
s
um
ption u
nder
Conver
gecast t
raffic
Figure
32. Late
ncy un
der Co
nver
gecast t
ra
ff
i
c
4.5.
Ef
fect
of
degree
of m
obi
li
ty
(
simul
at
i
on
w
ith
m
ob
il
e sink)
In
wireless
ad
ho
c
netw
orks
,
the
m
a
in
interact
ion
s
be
twee
n
no
des
are
th
e
interact
ion
s
betwee
n
t
he
sink
node
an
d
oth
e
r
no
des
in
the
net
work
.
I
n
us
ual
WSN
scenari
os
,
t
he
sink
node
is
pl
aced
in
t
he
c
orner
of
the
netw
ork.
T
her
e
fore
us
i
ng
m
ob
il
e
sink
node
instea
d
of
fi
xed
on
e
,
m
ean
s
changin
g
the
m
ob
il
i
ty
of
the
sink
node
a
nd
will
hav
e
the
s
am
e
i
m
pact
as
cha
ngin
g
the
m
ob
il
it
y
of
ot
her
sen
so
r
no
des.
I
n
t
his
stu
dy,
t
he
s
ta
ble
sink
is
re
place
d
with
a
m
ob
il
e
sink
.
A
m
ob
il
e
sink
ha
s
an
eff
ect
on
the
c
onnecti
vity
between
t
he
sin
k
and
t
he
oth
e
r nodes,
and
hen
ce
r
e
placi
ng the
s
ta
ble
si
nk w
it
h m
ob
il
e one ca
n
a
ff
ect
the d
el
iv
ery
rati
o.
Figure
33
to
F
igure
37
sho
w
that
by
rep
la
c
ing
the
s
ta
ble
sink
with
a
m
ob
il
e
one
in
bot
h
broad
ca
st
and
co
nver
gec
ast
patte
rn
s
,
th
ere
is
in
deed
a
no
ti
ceable
im
pr
ovem
ent
in
th
e
delivery
rati
on
wit
h
no
si
gn
i
ficant
i
m
pact o
n
ene
r
gy cons
um
ption
a
nd lat
ency
.
Evaluation Warning : The document was created with Spire.PDF for Python.