Internati
o
nal
Journal of Ele
c
trical
and Computer
Engineering
(IJE
CE)
V
o
l.
6, N
o
. 5
,
O
c
tob
e
r
201
6, p
p
. 2
345
~235
1
I
S
SN
: 208
8-8
7
0
8
,
D
O
I
:
10.115
91
/ij
ece.v6
i
5.1
078
4
2
345
Jo
urn
a
l
h
o
me
pa
ge
: h
ttp
://iaesjo
u
r
na
l.com/
o
n
lin
e/ind
e
x.ph
p
/
IJECE
EETA: An Energy Efficient Tr
an
smission
Alignment for
Wireless Sensor Ne
twork Applications
San
g
s
o
on Lim
1
, Ha
yo
ung Oh
2
1
Software R&D
Center, Samsung
Electron
i
cs, Seoul, South
Korea
2
School of
Electronic
and
Engin
eering
,
SoongSil University
, Seo
u
l, South
Korea
Article Info
A
B
STRAC
T
Article histo
r
y:
Received Apr 10, 2016
Rev
i
sed
Jun
1
,
2
016
Accepted
Jun 17, 2016
Energ
y
cons
erv
i
ng MAC protocols performing adaptive d
u
ty
-
c
y
c
ling
m
echanis
m
hav
e
been
widel
y
s
t
udied
to im
prove the
energ
y
e
ffici
enc
y
in
Wireless Sensor Networks (WSNs). In
particular, sever
a
l
as
y
n
chronous low
power
listening
(LPL) MAC
protoc
ols suc
h
a
s
B-M
A
C,
X-MAC
a
n
d
ContikiMAC transmit a long pr
eamble or
consecutive data p
a
ckets for an
efficient
rendezvous between
senders
and
receivers. However,
th
e
rendezvous r
e
sults in
the
cha
llenging
probl
e
m
of unnecess
a
r
y
ch
anne
l
utili
zat
ion since
the senders occup
y
a larg
e
portion of the m
e
dium
.
F
u
rtherm
ore,
when a tra
ffi
c ge
ner
a
tion
tim
e overl
aps with other
neighbouring nodes,
they
frequently
encounter spatially
-
co
rrelated
conten
tion incur
r
ing excessive channel co
n
t
en
tio
n. In this paper
,
we propose
a novel traff
i
c d
i
stribution scheme called an En
erg
y
Ef
ficient Transmission
Alignm
ent (EE
T
A), that shifts
a traffic gen
e
r
a
tion tim
e of th
e appli
cat
ion
lay
e
r
.
B
y
using
a MAC lay
e
r
feedback
includin
g
conten
tion inf
o
rmation, the
cross-lay
e
r fr
amework determin
es whet
her
th
e node
delay
s
its transmission
or not.
EETA is
robust from
the heav
y
cont
end
i
ng environm
ent
due to
its
traffic distr
i
bution feature. We
evalu
a
te the per
f
ormance of
EETA throug
h
diverse
experim
e
nts on th
e TelosB plat
form.
The results show that EETA
im
proves
the ov
eral
l en
erg
y
eff
i
cien
c
y
b
y
up
to
35%, and r
e
duc
e
s
the l
a
ten
c
y
b
y
up to
48%
co
m
p
ared to
the
ex
isting sch
e
m
e
.
Keyword:
App
licatio
n traffic
Energy e
fficie
n
cy
M
A
C
p
r
ot
oc
ol
W
i
rel
e
ss se
ns
o
r
net
w
or
ks
Copyright ©
201
6 Institut
e
o
f
Ad
vanced
Engin
eer
ing and S
c
i
e
nce.
All rights re
se
rve
d
.
Co
rresp
ond
i
ng
Autho
r
:
H
a
you
ng
Oh
,
Scho
o
l
of
Electr
on
ic and
Engin
eer
ing
,
Soo
n
g
S
il Un
i
v
ersity,
Seo
u
l
,
S
o
ut
h
K
o
rea
.
Em
a
il: h
y
o
h
79@g
m
ail.co
m
1.
INTRODUCTION
W
i
rel
e
ss Se
ns
or
Net
w
or
k
(
W
S
N
) i
s
a
n
e
m
ergi
n
g
t
echno
log
y
th
at su
pp
orts low-po
wer and
reliab
l
e
wi
rel
e
ss c
o
m
m
uni
cat
i
on o
f
vari
ous
real
-l
i
f
e ap
pl
i
cat
i
ons
[1
].
It
h
a
s a
g
r
eat po
ten
tial to
en
ab
le nov
el sm
art
services i
n
the
Inte
rnet
of
Th
ings
(I
oT
) e
n
v
i
ro
nm
ent
[2]
.
Si
nce
WS
Ns a
r
e m
a
de up
o
f
t
i
n
y
bat
t
e
ry
-p
o
w
er
e
d
sens
or
n
o
d
e
s,
o
n
e
of
t
h
e m
o
st
cri
t
i
cal
chal
l
e
n
g
es i
s
e
n
e
r
gy
c
ons
um
pt
i
on
pr
obl
em
[3]
.
Seve
ral
cl
ever
schem
e
s have
been
pr
o
pos
e
d
[
4
]
-
[
8
] in
MAC layer to
deal with
th
e prob
lem
.
Th
e
schem
e
s can
be cate
g
orized into two types accord
ing
to duty-cycling m
ech
anism
s
: synchronous
and
asy
n
ch
ro
n
o
u
s
app
r
oaches
. I
n
t
h
e sy
nch
r
o
n
ous a
p
pr
oac
h
e
s
[4]
,
[5]
,
al
l
n
ode
s de
pl
oy
ed
i
n
t
h
e sam
e
n
e
t
w
o
r
k
d
o
m
ain
sim
u
lt
an
eou
s
ly turn th
e RF t
r
ansceiv
e
r
s
on
and
o
ff to
synchr
on
ize th
eir
tr
an
sm
i
ssio
n
schedu
les wh
ile
p
e
rf
or
m
i
n
g
lo
w
po
w
e
r
op
eratio
n
.
A
l
t
h
ough
it p
r
eserv
e
s
m
o
r
e
en
erg
y
in
a sp
ecif
i
c en
v
i
ron
m
en
t, clo
c
k
syn
c
hron
ization
li
m
its th
e p
e
rfo
r
m
a
n
ce i
m
p
r
ov
em
en
ts in
th
e wireless
en
v
i
ron
m
en
t. In
th
e asy
n
chro
nou
s
approaches
[6]
-
[8], a receive
r chec
ks
the on-going trans
m
issions every
wake
-up interval. It m
a
kes
use
of
Clear Cha
nnel
Assessm
ent (C
CA) m
echan
is
m
to identify Z
i
gBee data
pac
k
ets.
If t
h
e inte
nde
d
receive
r s
e
nses
a busy cha
n
nel
,
it kee
p
s the
wake
-up
state for the
po
ten
t
i
a
l p
ack
et. A sen
d
e
r con
tin
uou
sly transm
i
t
s a lo
ng
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
I
J
ECE
Vo
l. 6
,
N
o
. 5
,
O
c
tob
e
r
20
16
:
234
5
–
23
51
2
346
pream
ble or
da
ta packets
until receiving
a
n
a
c
knowledge
pa
cket or reac
hing
the e
n
d of the wake-up i
n
te
rval.
The asy
n
ch
r
o
n
ous
sc
hem
e
s have
been
i
m
plem
ent
e
d o
n
Ti
ny
OS
[
9
]
an
d
C
ont
i
k
i
OS
[
1
0]
, a
nd
wi
del
y
use
d
.
These
protoc
ol
s ena
b
le a ZigBee transceive
r to ope
r
at
e at lo
w-du
ty cycle with
ou
t an
y time syn
c
h
r
o
n
i
zatio
n
and provi
de si
m
p
licity and effectivenes
s.
Ho
we
ver
,
t
h
e
asy
n
ch
ro
n
o
u
s
Lo
w P
o
we
r Li
st
eni
n
g (L
PL)
M
A
C
pr
ot
ocol
s suc
h
as B
-
M
A
C
[
6
]
,
X
-
M
A
C
[7]
an
d
C
ont
i
k
i
M
AC
[
8
]
occu
py
a l
a
r
g
e p
o
rt
i
o
n
of t
h
e m
e
di
u
m
dur
i
ng t
h
e
ren
d
ez
vo
us
bet
w
ee
n
sen
d
er
s
and recei
vers
. It significantl
y aff
ects the
unnecessa
ry c
h
annel c
onte
n
ti
ons i
n
m
o
st of
WSN a
p
plications.
M
a
ny
WSN
a
ppl
i
cat
i
o
ns
rep
o
rt
t
h
ei
r i
n
f
o
r
m
at
i
on suc
h
a
s
eve
n
t
dat
a
,
m
oni
t
o
ri
ng
dat
a
an
d c
o
nt
rol
dat
a
by
peri
ods
[11]. In a
ddition, a
c
e
ntral
e
n
tity, called sink
node, gat
h
ers
thos
e inform
ation
from
all nodes
in the
sam
e
dom
ai
n n
e
t
w
o
r
k
.
Si
nce t
h
e t
r
a
n
sm
i
ssi
ons
ha
ve a
sp
at
i
a
l
cor
r
el
at
i
o
n
,
excessi
ve
dat
a
t
r
ansm
i
ssi
ons
occu
r
at th
e sam
e
ti
me in
th
e same in
terferen
c
e reg
i
o
n
. In
this case, the
prolifera
tion of
data packets
cause
s
additional cha
nnel c
onte
n
tions and res
u
lts
in seve
re pa
cket
collisions
am
ong neighbourhood nodes.
It
aggra
v
ates the
energy effi
ciency, end-t
o
-e
nd
delay and
pac
k
et recep
tion rat
i
o
in
t
h
e real WSN depl
oyment.
In
th
is
p
a
p
e
r,
we propo
se a
n
o
v
e
l traffic alig
n
m
en
t sch
e
me, called
EETA, to
allev
i
ate th
e sp
atially-
correlated cont
ention
proble
m in
W
S
Ns. E
E
TA analyses a
backoff
patte
rn
i
n
M
A
C
l
a
y
e
r and
pr
o
v
i
d
es
usef
ul
feedbac
k
s to t
h
e application layer. It m
easure
s
a busy
p
e
riod wh
ile sen
s
ing
th
e ch
ann
e
l statu
s
to
tran
sm
it
d
a
ta
packet
s
.
B
a
se
d
on
t
h
e
m
e
asure
d
b
u
sy
peri
od
f
r
om
t
r
ansm
i
ssi
ons
of
ot
her
nei
g
hb
o
u
r
h
oo
d
n
o
d
es,
t
h
e
application layer shifts its initial
tr
ansm
ission
poi
nt of the periodic traf
fic. EETA aims at slightly av
oiding
ove
rlapping of the channel usage pa
ttern
between
n
o
d
e
s
lo
cated
in
th
e
sam
e
in
terferen
ce reg
i
on
. Thu
s
, it
considera
b
ly reduce
s
c
h
annel
conten
tion del
a
y
and unneces
sary
e
n
ergy
drain in the
cha
n
nel c
onte
n
tion
pha
se.
We im
p
l
e
m
en
t
e
d
EETA
o
n
t
h
e Telo
sB
senso
r
no
d
e
[1
2
]
to
v
e
rify th
e feasib
ility o
f
the traffic alignmen
t
schem
e
. Furt
h
e
rm
ore, we e
v
al
uat
e
t
h
e pe
rf
orm
a
nce of
o
u
r
sch
e
m
e
in
term
s
o
f
th
e
e
n
ergy efficiency,
end-to-
end
del
a
y
a
n
d
pac
k
et
rece
pt
i
on
rat
i
o
.
T
h
e
res
u
l
t
s
sh
o
w
t
h
at
ou
r
pr
o
pos
ed
schem
e
re
duces
t
h
e
ener
gy
con
s
um
pt
i
on
b
y
up t
o
3
5
%, e
n
d
-
t
o
-en
d
del
a
y
by
up t
o
4
8
%
. The
pe
rf
or
m
a
nce im
pro
v
e
m
e
nt
i
n
crease
s
as t
h
e
num
ber of
nei
g
hb
o
u
r
h
oo
d n
o
d
e
s
i
n
crea
se.
The
rest
of
t
h
i
s
pa
per
pr
ocee
ds as
f
o
l
l
o
ws.
Sect
i
o
n 2 in
trod
u
c
es th
e relat
e
d
work
.
We th
en descri
b
e
th
e app
licatio
n traffic align
m
en
t sch
e
m
e
in
Sectio
n
3.
Se
ct
i
on
4 eval
ua
t
e
s t
h
e p
r
o
p
o
s
e
d sc
hem
e
. Fi
nal
l
y
,
Sect
i
on
5 c
o
nc
l
udes
t
h
e
pa
per
.
2.
RELATED WORK
2.
1.
Wireless
Sens
or Netw
ork Applicati
o
ns
WSNs a
r
e a
p
plied in
dive
rse
areas s
u
c
h
as
s
m
art
buildi
n
g,
sm
art hom
e, smar
t healthcare, and
so on
[13
]
. Sm
art b
u
i
ld
in
gs try to
red
u
c
e en
erg
y
con
s
u
m
p
tio
n
b
y
ap
pro
p
riate humid
ity, v
e
n
tilat
i
o
n
,
air co
nd
itio
n
i
n
g
(HVAC
) con
t
ro
l an
d
m
easu
r
e roo
m
o
ccu
pan
c
y, tem
p
erat
u
r
e, and
v
a
riou
s features of th
e facilities.
S
m
art
h
o
m
e ap
p
licatio
n
in
tellig
en
t
l
y ch
eck
s env
i
ro
n
m
en
tal s
t
atu
s
o
r
hu
man
activ
ities
an
d
co
n
t
ro
ls
h
o
m
e
envi
ro
nm
ent
s
wi
t
h
vari
ous
act
uat
o
rs
. Sm
art
heal
t
h
ca
re
f
o
cuse
s
on
m
oni
t
o
ri
ng
pe
opl
e
’
s
dai
l
y
l
i
f
e a
n
d
p
r
ov
id
ing
m
e
d
i
cal alar
m u
n
d
er certain
m
e
d
i
cal co
n
d
itio
ns. In
th
ese scen
ari
o
s, m
u
ltip
l
e
sen
s
or no
d
e
s are
lo
cated
in th
e
sam
e
g
e
o
g
r
aph
i
c area
fo
r t
h
e pu
rp
ose of
reliab
ility. In
ad
d
ition
,
wh
en
an
in
terestin
g
ev
en
t
hap
p
e
n
s, m
u
l
t
i
pl
e n
odes
ge
ne
rat
e
a peri
odi
c
t
r
affi
c t
o
c
o
nve
y
t
h
ei
r
m
easur
ed dat
a
or c
ont
rol
m
e
ssages. I
t
t
hus
lead
s to
sp
atially-co
r
related
ch
ann
e
l con
t
en
t
i
o
n
.
In
th
is
cas
e, a l
o
t
of sen
s
or
no
des c
onc
u
rre
nt
l
y
t
r
ansm
it
dat
a
p
ack
ets in
resp
on
se to
th
e
sam
e
ev
en
t. Esp
ecially, th
ese p
h
e
n
o
m
en
a d
e
grad
e ch
an
nel q
u
a
lity, th
ereb
y
agg
r
a
v
at
i
ng t
h
e overal
l
pe
rf
o
r
m
a
nce of
W
S
N ap
pl
i
cat
i
ons
. Ou
r pr
o
p
o
s
e
d
schem
e
can
avoi
d t
h
e spat
i
a
l
l
y
-
correlated c
h
a
n
nel contenti
on
via an ef
ficient
application tra
ffic alignm
ent.
2.
2.
Sensor
MA
C Prot
oc
ol
s
S-
MA
C [4
] is o
n
e
o
f
the m
o
s
t
w
e
ll-
k
now
n
en
erg
y
con
s
erv
i
n
g
M
A
C p
r
o
t
oco
l
s in
W
S
N
s
. Each
node
em
pl
oy
i
ng
S-
M
A
C
si
m
u
l
t
a
n
e
ou
sl
y
t
u
r
n
s i
t
s
R
F
t
r
an
scei
v
e
r o
f
f
t
o
c
o
n
s
e
r
ve t
h
e e
n
er
gy
co
nsum
pt
i
on
du
ri
n
g
i
d
l
e
l
i
s
t
e
ni
ng
peri
od
. Se
n
d
ers a
n
d rec
e
i
v
ers
peri
odi
cal
l
y
share t
h
e sy
nc
hr
o
n
i
z
at
i
on i
n
fo
rm
at
i
on t
o
comm
unicate
each othe
r during
active pe
riod.
In [5], T-M
A
C
dynam
i
cally
adjusts active
a
n
d sleep pe
riods by
con
s
i
d
eri
ng c
o
m
m
uni
cat
i
on p
a
t
t
e
rns an
d ha
r
d
wa
re l
i
m
i
t
a
ti
o
n
. Altho
ugh
th
ese so
l
u
tio
ns ad
dress th
e
p
r
o
b
l
em
of e
n
er
gy
was
t
e due t
o
i
d
l
e
l
i
s
t
e
ni
ng, t
h
e
net
w
or
ks se
v
e
rel
y
suf
f
er
fr
om
sy
nchro
n
i
zat
i
on o
v
e
r
hea
d
s t
o
m
a
i
n
t
a
i
n
wa
ke
-u
p sc
he
dul
es
.
Anothe
r approach to pre
s
erve en
ergy wastes is asynchronous m
echani
s
m
prop
ose
d
i
n
B
-
M
A
C
[6]
.
B-MAC m
a
kes use
of a long
pre
a
m
b
le
conce
p
t instea
d of static time synchroni
zation. Eac
h
receive
r
peri
odi
cal
l
y
w
a
kes
up
eve
r
y
wake
-
up i
n
t
e
r
v
al
. A se
n
d
er
t
r
ansm
i
t
s
a l
ong
pream
bl
e d
u
ri
n
g
t
h
e
ent
i
r
e
wa
ke-
u
p
in
terv
al wh
en
th
e no
de h
a
s
a
packet to se
nd. Th
ere
f
ore, the
y
can success t
h
eir
tran
sm
issi
o
n
s
with
in
the wak
e
-
up
i
n
t
e
r
v
al
. T
h
i
s
schem
e
does
n
o
t
nee
d
t
o
e
x
chan
ge
sy
nc
hr
oni
zat
i
o
n i
n
f
o
r
m
at
i
on. T
o
a
d
di
t
i
onal
l
y
re
du
ce t
h
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
EETA: An
En
erg
y
Efficien
t
Tra
n
smissi
o
n
Al
ig
hmen
t f
o
r Wi
reless S
e
n
s
o
r
Netwo
r
k .... (Sa
n
g
s
o
o
n
Lim)
2
347
h
eav
y burd
e
n
s
o
f
th
e lon
g
p
r
ea
m
b
le, X-MAC [7
] co
n
s
ecu
t
iv
ely tran
sm
i
t
s
sh
ort p
r
eam
b
l
es u
n
til th
e in
ten
d
ed
receiver
receives the short prea
m
b
le and se
nds a
n
early-a
ck. In pa
rticular, ContikiM
AC [8], widely use
d
in
real
W
S
N test
beds, com
b
ine
s
the adva
ntages of va
riou
s asyn
chr
ono
us ap
proaches s
u
c
h
as B-MAC,
X-M
A
C,
and
B
o
X-M
A
C
[1
4]
.
H
o
we
v
e
r, t
h
ese s
o
l
u
t
i
ons
i
n
c
u
r
e
x
ce
ssi
ve c
h
an
nel
c
ont
e
n
t
i
o
n
due
t
o
a l
o
ng
p
r
eam
bl
e o
r
m
a
ny
successi
ve dat
a
pac
k
et
s
of se
nde
rs.
Ou
r pr
o
p
o
s
ed sc
h
e
m
e
deal
s wi
t
h
t
h
e chan
nel
co
nt
ent
i
o
n p
r
o
b
l
e
m
of
t
h
e asy
n
c
h
r
o
no
us L
P
L M
A
C
pr
ot
oc
ol
s
by
p
r
ovi
di
n
g
M
A
C
l
a
y
e
r co
nt
ent
i
o
n
feed
bac
k
t
o
t
h
e a
ppl
i
cat
i
o
n
l
a
y
e
r.
3.
ENERGY EFFICIE
N
T
TR
AFFI
C ALIG
NME
N
T
In t
h
i
s
sect
i
o
n,
we fi
rst
p
o
i
n
t
out
a cha
n
nel
cont
e
n
t
i
on
pr
o
b
l
em
of cur
r
ent
WS
N M
A
C
p
r
ot
oc
ol
s an
d
then
propose e
n
ergy efficient
traffic a
lig
nm
ent (E
ETA
)
fra
m
e
wor
k
a
n
d al
go
rithm
.
3.
1.
Ch
annel Contention
O
v
erhe
a
d of
en
er
g
y
c
o
nser
vi
n
g
MA
C pr
ot
oc
ol
s
In a low
duty cycle environmen
t
of
WSN
s
, t
h
e
ene
r
gy
con
s
um
pt
i
on
o
f
t
h
e
rece
pt
i
o
n
ope
rat
i
o
n
su
fferi
n
g fro
m
id
le listen
i
ng
an
d ov
e
r
hea
r
i
n
g i
s
m
o
re se
ve
re
pr
obl
em
t
h
an t
h
e
t
r
a
n
sm
i
s
si
on
o
p
erat
i
o
n
[
12]
.
There
f
ore,
m
o
st
of
t
h
e e
x
i
s
t
i
n
g asy
n
ch
ro
n
o
u
s
l
o
w
po
wer
M
A
C
pr
ot
oc
ol
s i
n
WS
Ns a
d
opt
a l
a
r
g
e
pream
bl
e o
r
consecutive
da
ta transm
issio
n
s to
re
ndez
vous
betwe
e
n a
sende
r a
nd a
r
eceiver. Howeve
r, t
h
is approac
h
cau
ses h
e
av
y ch
ann
e
l con
t
en
tio
n
o
v
e
rh
ead
.
Figu
re
1
illu
strates th
e ch
ann
e
l con
t
en
t
i
o
n
ov
erh
e
ad
o
f
th
e
asy
n
ch
ro
n
o
u
s
l
o
w
p
o
we
r M
A
C
pr
ot
oc
ol
em
pl
oy
ed i
n
C
o
nt
i
k
i
O
S.
A se
nde
r nee
d
s t
o
t
r
ans
m
i
t
consecut
i
v
e dat
a
packets
duri
ng the e
n
tire
wake-up i
n
terv
al
. If a tra
ffic a
rri
val of the a
p
pli
catio
n
layer
o
c
cu
rs at th
e
sam
e
ti
m
e
am
ong
nei
g
h
b
o
u
r
ho
o
d
no
des
,
t
h
ey
se
ver
e
l
y
suf
f
er
f
r
om
ex
cessi
ve c
h
an
ne
l
cont
e
n
t
i
o
n
du
e t
o
c
o
n
s
ecut
i
v
e dat
a
p
ack
ets. During
th
e ch
ann
e
l
co
n
t
en
tio
n, every no
d
e
m
u
st k
e
ep
id
le listen
i
ng
to ch
eck a ch
ann
e
l statu
s
.
It
drastically degrades
the
overa
ll perf
orm
a
nce
of
t
h
e
net
w
or
k. T
o
o
v
erc
o
m
e
t
h
e i
n
e
ffi
ci
en
cy
, we
de
vi
se
EET
A
to
ev
en
ly d
i
strib
u
t
e ap
p
licatio
n traffics amo
n
g
no
d
e
s.
We will exp
l
ain
th
e core
fram
e
work of
EETA in
the
next
su
b-
sect
i
o
n.
Fi
gu
re
1.
C
h
a
n
nel
co
nt
ent
i
on
ove
r
h
ead
o
f
as
y
n
ch
ro
n
ous
l
o
w
po
wer
M
A
C
p
r
ot
ocol
3.
2.
Cro
s
s-lay
e
r
Fra
mewo
r
k o
f
EETA
EETA is
designed
for a
n
efficient transm
iss
i
on
di
stribution to
reduce the
energy cons
um
ption
of
each se
ns
or node.
Figure
2 s
h
ows
the
fram
e
work
of
EET
A.
In case
of
MAC laye
r, t
h
e contention
m
onitor
m
odul
e chec
ks
a
back
o
f
f
del
a
y
aft
e
r t
h
e
n
o
d
e
ge
nerat
e
s
i
t
s
dat
a
packet
.
T
h
e
bac
k
o
f
f
del
a
y
st
at
i
s
t
i
c
s are t
h
e
n
pr
ocesse
d
by
a cont
e
n
t
i
o
n a
n
al
y
s
i
ng m
o
d
u
l
e t
h
at
deci
de
s whet
her i
t
i
s
no
rm
al
cont
ent
i
on
o
r
n
o
t
i
n
t
h
e
cont
e
n
t
i
on a
n
a
l
y
s
e proces
s. T
h
e deci
si
o
n
re
s
u
l
t
i
s
t
h
en p
r
oc
essed
by
a cro
ss-l
a
y
e
r fee
dba
ck m
odul
e t
o
n
o
t
i
f
y
i
t
t
o
t
h
e appl
i
cat
i
on l
a
y
e
r. I
n
case of ap
pl
i
cat
i
on l
a
y
e
r, t
h
e t
r
affi
c al
i
g
n
m
ent
m
odul
e shi
f
t
s
i
t
s
t
r
ans
m
i
ssi
on
start po
in
t
of ev
ery
p
e
ri
od
ic data p
a
ck
et b
a
sed
o
n
th
e
m
ac la
yer fee
d
backs
.
A
key a
dva
nta
g
e
of traffic s
h
ifting
is th
at it can
min
i
mize th
e ch
ann
e
l con
t
en
t
i
o
n
co
st, th
ereb
y i
m
p
r
ov
ing
ch
ann
e
l u
tilizatio
n
and
redu
cin
g
id
le
l
i
s
t
e
ni
ng o
f
t
h
e no
de. The f
eedbac
k
i
n
fo
r
m
at
i
on i
n
cl
ude
s t
h
e num
ber of t
h
e co
nsec
u
t
i
v
e fai
l
u
res o
f
C
l
ear
C
h
an
nel
Asses
s
m
e
nt (C
C
A
),
whi
c
h per
f
o
rm
s ener
gy
-ba
s
ed channel sensi
ng to c
h
eck
whethe
r a channel is
busy
or i
d
l
e
.
Whe
n
ot
he
r s
e
ns
or
no
des
o
ccupy
a l
a
r
g
e
po
rt
i
o
n o
f
t
h
e
m
e
di
um
due t
o
t
h
ei
r l
o
w
po
we
r
o
p
e
ration
,
t
h
e
CCA m
a
y su
ccessiv
e
ly fail un
til th
e
o
t
h
e
r
pa
irs of
no
d
e
s are
d
o
n
e
with
t
h
eir tran
sm
issi
o
n
s
.
In
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
I
J
ECE
Vo
l. 6
,
N
o
. 5
,
O
c
tob
e
r
20
16
:
234
5
–
23
51
2
348
ad
d
ition
,
a wak
e
-up
in
terv
al
o
f
MAC
p
r
o
t
o
c
o
l
is ano
t
h
e
r p
a
ram
e
ter to
sch
e
d
u
l
e applicatio
n
traffics. To
support t
h
ese
feature
s
,
we
design a
cross
-
layer soft
ware fram
e
wo
rk
b
e
tween
MAC
and
ap
p
licatio
n
layers.
Wh
en th
e M
A
C layer an
alyses con
t
en
tio
n pattern
s,
t
h
e
fram
e
work immediately adjusts the
packe
t
transm
ission sc
hedule.
Fi
gu
re
2.
C
r
oss
-
l
a
y
e
r f
r
am
ework
o
f
e
n
ergy efficient
traffic a
lignm
ent
3.
3.
Traffic
Alignment Algorithm
C
u
r
r
ent
l
o
n
g
pream
bl
e ove
r
l
o
w
po
wer
o
p
erat
i
o
n c
onst
i
t
u
t
e
s a m
a
jor part
of e
x
cess
i
ve cha
nnel
conte
n
tion i
n
WSNs. T
h
ere
f
ore
,
it is necessary that each
node avoids concurre
nt
trans
m
issions of a
peri
odic
d
a
ta p
ack
et th
at o
v
e
rlap
s
with
th
e activ
ities o
f
o
t
h
e
r n
o
d
e
s. Accord
ing
to
ch
an
nel u
s
ag
e p
a
ttern
s
of
n
e
igh
bou
rho
od no
d
e
s, th
ere is a stab
le
p
a
ttern
for th
e
ev
en
t-driv
en
WSN
ap
p
lication
.
Eac
h
no
de m
easur
es i
t
s
b
ackof
f
d
e
lay
T
D
p
e
r on
e
ap
p
lication
p
a
ck
et in
a Carrier Sen
s
ing
Mu
ltip
le Access (CSMA) su
b-layer
per
f
o
r
m
i
ng ch
annel
c
ont
e
n
t
i
on
pr
ocess
,
o
n
ce t
h
e ori
g
i
n
al
packet
i
s
ge
n
e
rat
e
d by
t
h
e
appl
i
cat
i
o
n l
a
y
e
r. I
n
ad
d
ition
,
it ch
eck
s
a wak
e
-up in
terv
al T
w
f
r
o
m
a R
a
di
o Dut
y
C
y
cl
e (R
DC
) m
odul
e t
o
cal
cul
a
t
e
t
h
e avera
g
e
tran
sm
issio
n
delay o
f
lo
w
p
o
wer
o
p
e
ration
.
Th
e algo
rith
m
fo
r traffic alig
n
m
en
t is g
i
v
e
n in
Algo
rith
m
1
.
In
every run, t
h
e
inputs
to
th
e alg
o
rith
m
are T
D
and T
w
and
th
e ou
tpu
t
is traffic sh
ifting
o
f
fset, T
S
. First,
th
e
sender com
put
es the cu
rren
t
MAC layer b
ack
off
d
e
lay and
add
s
it to
T
D
. If T
D
is larg
er th
an
th
e av
erag
e
w
a
k
e
-u
p
in
ter
v
al an
d
th
e node su
ff
er
s
f
r
o
m
co
nsecu
tiv
e
b
a
ck
of
f, th
e node in
cr
eases th
e n
u
m
b
e
r
of
b
a
ck
of
f
trial, N
B
. And then, t
h
e node com
putes
th
e cu
lu
m
a
tiv
e
MAC layer b
a
ck
off
d
e
lay to calcu
late th
e traffi
c
sh
ifting
o
f
fset wh
en
t
h
e N
B
e
x
ceeds a s
p
eci
fic thre
shold
. E
v
e
n
tua
lly, although t
h
ere exist se
veral
nodes
atte
m
p
tin
g
to
t
r
an
stm
it at th
e sam
e
ti
me in
th
e in
terfer
i
n
g
reg
i
o
n
, th
ey
slig
h
tly sh
ifts th
eir traffic
g
e
n
e
ratio
n
tim
e
s and ca
n
avoi
d
unnecess
a
ry cont
ention
tim
e
causing e
n
ergy
wastes.
Algo
rith
m
1
:
Traffic Align
m
en
t
Inpu
t : T
D
-
ba
cko
f
f
dal
a
y
per
o
n
e a
ppl
i
cat
i
o
n
packet
;
T
w
-
wak
e
-u
p in
terval o
f
RDC;
Out
put
:
T
S
- traffic sh
ifting
offset;
1:
C
o
m
put
e
M
A
C
l
a
y
e
r
ba
cko
f
f
del
a
y
2:
if
T
D
> T
w
/2
an
d con
s
ecu
tiv
e
b
ackoff i
s
tru
e
th
en
3:
N
B
inc
r
ease
4:
else
5:
N
B
rese
t
6:
C
o
m
put
e c
u
m
u
l
a
t
i
v
e M
A
C
l
a
y
e
r bac
k
o
f
f
del
a
y
T
CD
7:
if
N
B
>
th
e
n
8:
T
S
= T
W
/2
–
T
CD
+ (ra
nd
()
%T
I
)
9:
else
10:
T
S
reset
11:
R
e
t
u
r
n
T
S
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
EETA: An
En
erg
y
Efficien
t
Tra
n
smissi
o
n
Al
ig
hmen
t f
o
r Wi
reless S
e
n
s
o
r
Netwo
r
k .... (Sa
n
g
s
o
o
n
Lim)
2
349
4.
PERFO
R
MA
NCE E
V
ALU
A
TIO
N
4.
1.
Experimental Setup
We ha
ve i
m
pl
em
ent
e
d o
n
Tel
o
sB
sens
o
r
n
o
d
es [
12]
r
u
nni
n
g
C
o
nt
i
k
i
OS 2
.
7. Tel
o
sB
i
s
m
a
de u
p
o
f
a
M
SP4
30 M
i
cr
o C
o
nt
r
o
l
U
n
i
t
(M
C
U
) a
nd
a C
C
2
4
20
rad
i
o chi
p
set
.
Th
e C
C
2
4
20 c
h
i
p
set
i
s
an
80
2
.
1
5
.
4
com
p
l
i
a
nt
R
F
transcei
ver
ope
rat
i
ng i
n
t
h
e
2.
4G
Hz I
S
M
ba
nd a
nd s
u
pp
o
r
t
s
a dat
a
rat
e
of
25
0k
b
p
s.
We m
a
ke
use
of t
h
e C
o
nt
i
k
i
M
A
C
m
odu
l
e
devel
ope
d
b
y
du
nkel
s
[
10]
and a
d
d M
A
C
l
a
y
e
r feed
bac
k
m
echani
s
m
.
To t
h
e
sender side,
we e
m
bedde
d traffic alignm
ent function fo
r traffic
distri
bution. W
e
set one receive
r and nine
send
er
s t
o
evalu
a
te ou
r pr
oposed
sch
e
m
e
. To
exp
l
o
r
e t
h
e ef
f
ect
o
f
t
h
e w
a
k
e
-up
in
terv
al
d
u
r
a
tion
,
w
e
var
y
th
e
wake
-
up i
n
t
e
r
v
al
fr
om
62.
5
m
s t
o
25
0m
s in t
h
e sam
e
di
st
ance. I
n
ou
r
expe
ri
m
e
nt
, we anal
y
s
e t
h
e ener
gy
con
s
um
pt
i
on
b
y
com
p
ari
n
g
t
h
e s
o
jo
ur
n
time at active a
n
d sleep state.
We
al
so m
easure
en
d-t
o
-e
nd
de
l
a
y
o
f
application tra
f
fics to
ve
rify the feasibility of
EETA.
4.
2.
Energy Consu
m
ption
Fi
gu
re
3 s
h
ow
s t
h
e e
x
peri
m
e
nt
al
res
u
l
t
f
o
r t
h
e a
v
era
g
e e
n
e
r
gy
c
o
nsum
pt
i
o
n
pe
r
u
n
i
t
t
i
m
e
o
f
pai
r
e
d
sen
d
er a
n
d
rec
e
i
v
er a
s
a
f
u
n
c
t
i
on
of
t
h
e
w
a
ke-
u
p i
n
t
e
rva
l
s. I
n
b
o
t
h
bas
i
c LPL M
A
C
and
LPL
M
A
C
wi
t
h
EETA, the
rec
e
iver s
a
ves
m
o
re
ene
r
gy as
the wa
ke
-up
i
n
terval inc
r
eas
es due t
o
a
large
porti
o
n of slee
p
du
rat
i
o
n.
Whe
n
a no
de pe
rf
o
r
m
s
dut
y
-
cy
cl
ing m
echani
s
m
wi
t
h
l
o
n
g
wa
ke-
u
p i
n
t
e
rval
,
t
h
e no
de kee
p
s sl
eep
state avoi
ding unneces
sary
energy dr
ain
during idle
state, there
b
y th
e sc
hem
e
ca
n im
prove
the
energy
efficiency.
When seve
ral senders tra
n
sm
it
their pac
k
et
at
t
h
e sam
e
t
i
m
e
, onl
y
one se
n
d
e
r get
s
t
h
e m
e
di
um
du
ri
n
g
t
h
e c
h
a
nnel
c
ont
e
n
t
i
o
n p
h
ase. T
h
us
, ot
he
r n
o
d
es
wai
t
fo
r t
h
e n
e
xt
o
p
p
o
rt
uni
t
y
whi
l
e
per
f
o
r
m
i
ng
ad
d
ition
a
l ch
an
n
e
l
b
acko
f
f.
In
th
at case, th
e b
a
si
c LPL MAC sev
e
rel
y
su
ffers fro
m ch
ann
e
l con
t
en
tio
n
o
v
e
rh
ead, th
ereb
y it lead
s to
in
crease i
d
le listen
i
ng
du
ra
ti
on
fo
r an
add
itio
n
a
l ch
ann
e
l
back
off. However,
o
u
r
p
r
op
o
s
ed
sch
e
me slig
h
tly sh
i
f
ts its traffic
gen
e
ratio
n
tim
e
to
m
itig
ate sp
atially-co
rrelated
co
n
t
en
tion
a
m
o
n
g
nei
g
hb
o
u
ri
ng
no
des
.
Aft
e
r a no
de sense
s
an excessi
ve ch
annel
c
ont
e
n
t
i
o
n
,
i
t
adapt
i
v
e
l
y
adjust
s i
t
s
p
e
ri
o
d
i
c
traffic
p
a
ttern
. As a
resu
lt, LPL MAC
wit
h
EETA con
s
erves t
h
e a
v
era
g
e ene
r
gy
cons
um
ption c
o
m
p
ared to
th
e
ex
isting
LPL
MAC b
y
u
p
to
35
%
du
e
t
o
t
h
e e
ffect
of tra
ffic alignm
ent.
Fi
gu
re
3.
A
v
er
age e
n
er
gy
c
o
n
s
um
pt
i
on
with
resp
ect t
o
the
wak
e
-u
p in
tervals
4.
3.
End-To-End
Delay
Fi
gu
re 4
descri
bes t
h
e e
xpe
ri
m
e
nt
al
resul
t
for t
h
e l
a
t
e
ncy
as a fu
nct
i
on
o
f
wa
ke-
u
p i
n
t
e
rval
s.
In L
P
L
M
A
C
,
b
o
t
h
a s
e
nde
r a
n
d
a re
cei
ver
do
n
o
t
k
n
o
w
t
h
e exact
wake
-
up
p
o
i
n
t
of eac
h
ot
he
r.
It
i
n
cu
rs
ren
d
e
z
vo
u
s
d
e
lay b
e
tween
th
e send
er and
th
e in
te
nde
d
re
ceiver.
The
r
efore, as t
h
e wa
ke
-up interval increases, t
h
e average
delay also inc
r
eases.
Whe
n
t
h
e nodes
pe
rform
EETA to av
oid c
o
ncurre
nt
transm
issions
,
they are c
o
nsiderably
ab
le to
redu
ce th
e laten
c
y
com
p
ared to the basic LPL MAC by up to
48%. Since the application traffics are
ev
en
ly
d
i
stributed
, sim
u
ltan
e
o
u
s
tran
sm
issi
o
n
trials can
be reduce
d.
It i
ndicates t
h
at a
node
can ha
ve m
o
re
o
ppo
rt
u
n
ity to g
r
ab
t
h
e m
e
d
i
u
m
with
less ch
an
nel conten
tio
n
and
min
i
m
i
ze id
le li
sten
ing
d
u
ratio
n fo
r
un
necessa
ry
c
h
an
nel
c
ont
e
n
t
i
ons
. I
n
case
o
f
LPL M
A
C
wi
t
h
EET
A, t
h
e del
a
y
vari
at
i
on i
s
sm
al
l
e
r t
h
an t
h
e
LPL
M
A
C. Since EETA a
d
a
p
tively cont
rol
s
the contention environm
ent,
it effectively
reduces the ra
ndom
property of a
n
additional c
h
annel bac
k
off a
n
d
unnecessa
ry
waiting tim
e for a tra
n
sm
ission.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
I
J
ECE
Vo
l. 6
,
N
o
. 5
,
O
c
tob
e
r
20
16
:
234
5
–
23
51
2
350
Fig
u
re
4
.
Latency with
resp
ect
to
th
e wak
e
-up
in
terv
als
4.
4.
Packe
t
Reception Ratio
Fi
gu
re 5
dem
onst
r
at
es t
h
e
pa
cket
rece
pt
i
on
rat
e
as a fu
nct
i
on
of c
o
ncu
rre
nt
t
r
ansm
i
ssi
on n
o
d
es.
I
n
WSNs,
pe
riodic data
packet i
s
one
of the m
a
in feat
ur
es
for m
o
st of t
h
e
applications.
For e
x
am
ple, se
veral
sens
or
n
ode
s
p
e
ri
o
d
i
cal
l
y
rep
o
rt
di
ve
rse i
n
f
o
rm
at
i
on t
o
t
h
e cent
r
al
c
o
or
d
i
nat
o
r
,
cal
l
e
d s
i
nk
n
ode
. I
n
a
d
di
t
i
on,
som
e
no
des l
o
cat
ed i
n
t
h
e s
a
m
e
cove
rage
t
r
ansm
i
t
peri
od
ic inform
atio
n
after sen
s
ing
a sp
ecial even
t. To
v
e
r
i
f
y
ou
r
p
r
opo
sed
schem
e
, a
ll n
e
ig
hbo
ur
ing
nod
es sim
u
lt
an
eou
s
ly g
e
n
e
rate p
e
r
i
od
ic tr
af
f
i
c ev
er
y
5
seco
nd
s.
As the num
b
er of c
onc
urre
nt packet trans
m
issions
incre
a
ses, the pac
k
et reception
ra
te of LPL MAC is
si
gni
fi
ca
nt
l
y
d
e
gra
d
e
d
. T
h
e
r
easo
n
i
s
t
h
at
8
0
2
.
1
5
.
4
has i
t
s
ow
n l
i
m
it
at
i
on
of c
h
a
nnel
c
ont
e
n
t
i
o
n
.
I
f
a
n
ode
fails to
g
r
ab
th
e m
e
d
i
u
m
d
u
r
ing
th
e li
m
i
tatio
n
,
th
e tr
an
smissio
n
trial a
l
so
fails. As a resu
lt, it ag
g
r
av
ates
packet recepti
on
rate of LPL MAC.
However,
our sc
hem
e
doe
s not e
x
ce
ed the c
h
a
nnel
conte
n
tion limitation
o
f
802
.1
5.4.
Wh
en
t
h
e ch
an
n
e
l con
t
en
tion
d
u
ration
ex
ceed
s a sp
ecific th
resho
l
d
,
it in
tellig
en
tly avo
i
d
s
the
channel c
onte
n
tion. The
r
efore
,
LPL M
A
C with EETA
ou
tperform
s the va
nilla LPL
MAC. This shows that
EETA is m
o
re
practical sol
u
tion in a
real envi
ronm
ent
that consists of
m
a
ny se
nsor nodes
.
Our propose
d
sch
e
m
e
is well-
m
a
tch
e
d
with
th
e
h
e
tero
gen
e
ou
s
wireless n
e
two
r
k
env
i
ro
n
m
en
t d
u
e
to
th
e ad
ap
t
a
tio
n
alg
o
rith
m
.
Figure
5. Pac
k
et reception
rat
e
with
res
p
ect
t
o
num
ber of
concu
rre
nt tra
n
s
m
ission nodes
5.
CO
NCL
USI
O
N
In
WS
Ns, di
ve
rse ene
r
gy
co
n
s
ervi
ng
pr
ot
oc
ol
s base
d o
n
d
u
t
y
-cy
c
l
i
ng m
e
chani
s
m
have
been
wi
del
y
use
d
in MAC layer because of its efficiency. Howeve
r, t
h
e heavy
burde
n
s
of low power
ope
rations in
MAC
l
a
y
e
r seve
rel
y
deg
r
a
d
e t
h
e
o
v
e
ral
l
per
f
o
rm
ance
of
l
o
w
po
wer
p
r
ot
ocol
s
.
I
n
part
i
c
ul
ar
,
whe
n
a
t
r
a
n
sm
i
ssi
on
occurs at t
h
e
sam
e
tim
e, it
incurs
unneces
sary ene
r
gy
waste due to t
h
e excessi
ve c
h
annel c
o
ntenti
on of
peri
odic data transm
issions.
We propose
d a novel tr
affic alignm
ent sc
hem
e
, na
m
e
d EETA, which
evenl
y
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
EETA: An
En
erg
y
Efficien
t
Tra
n
smissi
o
n
Al
ig
hmen
t f
o
r Wi
reless S
e
n
s
o
r
Netwo
r
k .... (Sa
n
g
s
o
o
n
Lim)
2
351
d
i
stribu
tes ap
plicatio
n
traffics.
W
e
d
e
si
g
n
e
d
cro
s
s-la
yer
framewo
rk
to fu
lly u
tilize MAC layer feed
b
a
ck
and
sh
ift app
lication
traffics. To
v
e
rify its feasi
b
ility, we i
m
p
l
e
m
en
ted
ou
r
propo
sed
sch
e
m
e
o
n
Telo
sB sen
s
o
r
no
des
.
O
u
r ex
peri
m
e
nt
al
resul
t
s
sho
w
e
d
t
h
at
our sc
hem
e
im
pro
v
es ene
r
gy
effi
ci
ency
and e
n
d
-
t
o
-en
d
del
a
y
com
p
ared t
o
the existing sc
he
me.
ACKNOWLE
DGE
M
ENTS
Th
is r
e
sear
ch
w
a
s supp
or
ted b
y
Basic Scien
ce
Research Prog
ram
th
rou
g
h
th
e
Nation
a
l Research
Fo
un
dat
i
o
n o
f
Korea
(NR
F
) fu
n
d
ed
by
t
h
e M
i
ni
st
ry of Sci
e
nce,
IC
T & Fut
u
re Pl
an
ni
ng
(NR
F
-
2
014
R1A1
A1
00
356
2)
.
REFERE
NC
ES
[1]
R. Khadim,
et al.
, “Performance
Stud
y
of IEEE 802.15. 4 under
OP
NET Modeler for Wireless Sensor Networks,”
TELKOMNIKA Indonesian Journ
a
l
of Electrical
Engineering
, vol/issue: 16(1)
, pp
. 98-105, 2015.
[2]
N. Dinh
and S.
Lim, “Analy
sis
of IEEE 802
.15.4 Beac
on-En
abled MAC Protocol,”
IJECE
, vol/issue: 6(3), 2016.
[3]
Ak
y
ild
iz
,
et al.
,
“A survey
on sensor networks,”
IEEE Communi
cations Magazin
e
, vol/issue: 40(
8), pp. 102-114
,
2002.
[4]
W.
Ye
,
et
al
.
,
“
A
n energ
y
effi
ci
ent m
a
c
proto
c
o
l
for wir
e
less sensor networks,”
in
21st In
ternational Annua
l Jo
int
Conference of th
e IEEE Com
puter and Communications Societies
, 2002.
[5]
T. van D
a
m and K. Lang
endoen, “An ad
aptiv
e energ
y
-effcient MAC protocol
for wireless sen
s
or networks,” in
Pr
oc. o
f
the
Fir
s
t
ACM
Confer
en
ce on
Embedd
ed
Networked S
e
nsor Systems
, 2003
.
[6]
J.
Pola
stre
,
et al.
, “
V
ers
a
ti
le lo
w power m
e
dia acces
s
for wire
les
s
s
e
ns
or networks
,” in
Proc.
of the 2nd AC
M
SenSys Conferen
ce
, 2004.
[7]
M. Buettner
,
et al.
, “X-MAC: A Short Preamble MAC
Protocol for
Duty
-C
y
c
led
Wireless Se
nso
r
Networks,” in
Proc. o
f
the 4th
ACM SenS
ys Co
nference
, 2006.
[8]
A. Dunkels, “The contikim
ac rad
i
o dut
y
cy
cling
protocol
,” Tech
nical Report
5128, Swedish Institute of Com
puter
Science, 2011.
[9]
Tiny
OS, http://www.
tiny
o
s.
net.
[10]
ContikiOS,
http://
www.
contikios.
org.
[11]
U. Prathap,
et al.
, “Wireless sensor networks applic
ations and rou
ting protoco
l
s: su
rve
y
and res
e
ar
ch chal
leng
es
,”
Cloud and S
e
rvices Computing
(
I
SCOS)
,
2012 International Sym
posium on. I
EEE
, 2012
.
[12]
TelosB, http
://w
ww.m
e
m
s
ic.com
/userfil
e
s/fi
l
e
s/Datasheets/WSN/telosb_dat
a
sheet
.pdf
[13]
A. Al-Fuqaha,
et al.
, “Internet of things: A
survey
on enab
ling
technologies, protocols,
an
d applications,”
Communications
Sur
veys
&
T
u
tor
i
als
,
IE
EE
, vol/issue: 17(4), pp.
2347-2376, 201
5.
[14]
D. Moss and P. Levis, “BoX
-
M
ACs: Exploiting ph
y
s
ical
and
link lay
e
r boun
daries in
low-power networking
,”
Computer Systems Laborat
ory Stanford University
, pp. 116-119,
2008.
BIOGRAP
HI
ES OF
AUTH
ORS
Sangsoon Lim r
eceived Ph
. D. d
e
gree in
the Sch
ool of Computer
Scien
ce
and
En
gineer
ing from
Seoul National
University
in 2
013. Since Oct
ober 2013, he
works as a senior engineer at
S
o
ftware R&D Center
, S
a
m
s
ung Elect
ronics
. H
i
s
current res
ear
ch inter
e
s
t
s
are in the are
a
of
wireless networks includ
ing wir
e
less LAN, wireless sensor Netw
orks, cognitiv
e r
a
dio networks.
Hay
oung Oh received th
e B.S. d
e
gree
in Com
puter Scien
ce from Duksung Wo
ma
ns University
and the M
.
S. degree in
the Scho
ol of Comput
er
Science and
En
gineer
ing from Ewha Womans
University
in
2
002 and 2006
r
e
spectively
.
An
d she receiv
e
d
the Ph.D. degr
ee in Computer
Science from Seoul Nation
a
l
University
in
2013. From 200
2 to 2004, she joined Shinhan
Financial Group
as a d
e
veloper
in applied r
e
sear
ch. In
2010, she was with U.C
.
Berkeley
as
a
research
er. Since 2013, she h
a
s
been with Soon
gs
il University
as a pro
f
essor in the School of
Ele
c
troni
c Engi
neering
.
Her re
s
earch int
e
res
t
s
include s
o
ci
al
and co
mputer
networks, and
s
ecurit
y
.
Evaluation Warning : The document was created with Spire.PDF for Python.