TELKOM
NIKA
, Vol.14, No
.1, March 2
0
1
6
, pp. 136~1
4
3
ISSN: 1693-6
930,
accredited
A
by DIKTI, De
cree No: 58/DIK
T
I/Kep/2013
DOI
:
10.12928/TELKOMNIKA.v14i1.2744
136
Re
cei
v
ed O
c
t
ober 6, 20
15;
Revi
se
d Feb
r
ua
ry 2, 20
16
; Accepte
d
Febru
a
ry 17, 2
016
Energy Management of Context-aware Cold Chain
Vehicle Nodes
Ronling Hu
*
, XiangLi, Ya
ngLou, Wei
Yan
F
a
cult
y
of Com
puter an
d Softw
a
r
e En
gin
eer
i
ng, Hua
i
yin Ins
t
itute of
T
e
chnolo
g
y
,
Huai
’a
n 22
300
3, Jiangs
u, Chi
n
a
*e-mai
l
: 223
77
976
05@
qq.co
m
A
b
st
r
a
ct
In view
of the
questi
on that t
he en
ergy-co
n
s
traine
d senso
r
nodes
affect the nor
ma
l op
eratio
n o
f
the w
hol
e system du
e to th
e
short su
rviv
al
peri
od, i
n
this
pap
er w
e
pr
e
s
ent a
new
sol
u
tion
of a
dapti
v
e
ener
gy mana
g
e
ment from th
e ang
le of con
t
ext-aw
are
comp
utin
g. In different scenar
io
mod
e
s, w
e
take
different env
iro
n
ment acq
u
isiti
on an
d co
mmu
n
icati
on strate
gy accord
i
ng t
o
scenar
io par
ameters, and t
hus
buil
d
the co
nte
x
t-aw
are energ
y
man
a
g
e
m
ent
mod
e
l. Fi
na
lly
, w
e
extend the w
hole syste
m
life by l
o
w
e
r
i
ng
ener
gy cons
u
m
pti
on of sin
g
l
e
intern
et of thin
gs v
ehic
l
e
nod
e. Experi
m
ental d
a
ta sho
w
that the energy
ma
na
ge
me
nt p
l
an of co
ntext-aw
are w
i
reless
sensor no
des
effectively re
d
u
c
es the en
ergy
consu
m
pti
on
of
nod
es an
d extends the syste
m
life u
n
d
e
r the
pre
m
is
e of providi
ng re
lia
bl
e
services.
Ke
y
w
ords
: co
ntext-aw
are, in
ternet of things
, co
ld chai
n lo
g
i
stics, energy
ma
na
ge
me
nt
Copy
right
©
2016 Un
ive
r
sita
s Ah
mad
Dah
l
an
. All rig
h
t
s r
ese
rved
.
1. Introduc
tion
For the pro
d
u
ction, proce
ssi
ng,
tran
sp
ortation, sal
e
s of cold
ch
ai
n logistics, the normal
operation
of t
he
whol
e
system ne
ed
s to
monito
r
th
e
temperature,
humidity an
d
othe
r lo
gisti
c
s
environ
ment [
1
], while the t
r
an
spo
r
tation
environ
ment
monitori
ng i
s
the mo
st difficult an
d critical,
becau
se
thi
s
contin
uou
s m
onitorin
g
nee
ds an
eq
uipm
ent
[2]
to co
n
t
inue wo
rki
n
g
.
Ho
weve
r,
d
u
e
to the limitation of sp
eci
a
l con
d
ition
s
, the envir
o
n
men
t
al monitorin
g
node
s in
side
refrige
r
ato
r
van
usu
a
lly u
s
e t
he b
a
ttery po
wer that i
s
n
o
t
easy to
cha
nge
or
refu
el. This limited
energy of n
o
des
will seriously
affect the servic
e life of
the whole
cold chain
logistics monitoring system [
3
].
Therefore, th
e low-po
we
r desi
gn is p
a
rticularly
imp
o
r
tant for the
wirel
e
ss sen
s
or
enviro
n
m
ent
monitori
ng sy
stem [4]. At present, the research
on
sensor node energy ma
nagement mostly
still
stays in th
e
basi
c
routing
proto
c
ol
sel
e
ction
or
al
g
o
rithm o
p
timization, fe
w st
udying the
d
e
sig
n
optimizatio
n
of ene
rgy
manag
eme
n
t from th
e
perspe
c
tive
of the wh
ole sy
stem.
The
developm
ent
of co
mpute
r
scien
c
e
and
com
m
uni
cati
on te
chnol
og
y make
s th
e
cont
ext-awa
r
e
techn
o
logy a
r
ise
at the hi
stori
c
mom
e
nt. In
1994,
Schilit and
T
heime
r have
put forward
the
con
c
e
p
t of co
ntext, and def
ined it a
s
: location, nea
rby
peopl
e and
o
b
ject id
entification a
s
well
as
the ch
ange
of these o
b
j
e
cts [5]. Co
n
t
ext aware
n
e
ss [6, 7] is
a cal
c
ulatio
n
model u
s
ing
the
environ
ment context information to hel
p the deci
s
io
n optimizatio
n, which
generally involves the
informatio
n a
c
qui
sition, co
ntext modelin
g and intellig
ent pro
c
e
ssi
n
g
asp
e
ct
s [8].
At prese
n
t, in view of the energy co
nsu
m
pt
ion optimi
z
ation p
r
o
b
le
m of wirel
e
ss sen
s
o
r
node
s, the rese
arche
r
s mainly study
from two
a
s
pe
cts of
system ar
chite
c
ture and ro
uting
algorith
m
. Reasona
ble system archit
ecture can
redu
ce the n
ode en
ergy
con
s
um
ption
to a
certai
n extent
, but if we ne
ed to save
en
ergy co
n
s
um
ption in a d
e
e
per level,
we
also
need to
do
in-de
p
th re
se
arch on routi
ng algo
rithm.
We so
lve th
e energy pro
b
lem from ro
uting algo
rith
m,
su
ch a
s
in lit
eratu
r
e [9] we pro
p
o
s
e mi
nimum h
o
p
s
energy-a
dapt
ed protocol (MHEP) throu
gh
the an
alysi
s
of the mi
nimu
m cost
ro
utin
g alg
o
rithm.
The
algo
rith
m u
s
e
s
the
m
i
nimum
path
node
resi
dual
en
ergy and
the m
i
nimum
hop
cou
n
t of
sin
k
nod
es a
s
th
e sele
ction
routing m
e
tri
c
s to
compl
e
te da
ta forwa
r
din
g
; In literature [10] we
use en
erg
y
-saving o
p
timization e
n
e
rgy
manag
eme
n
t strate
gy of slee
ping
an
d awakeni
ng
strate
gy, to tran
sform t
he sl
eepi
ng
and
awa
k
e
n
ing
probabili
stic certainty probl
e
m
into
a
ba
rg
aining
gam
e,
and th
us a
c
hi
eve the
bala
n
c
e
of ene
rgy
an
d pe
rforman
c
e; In
literature
[11] we
pre
s
ent a ne
w d
i
stribute
d
po
sitioning t
r
a
cki
ng
strategy ba
sed
on dyna
mic
p
o
wer manag
eme
n
t, whi
c
h u
s
e
s
the wavelet de-noi
sing
and
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Energ
y
Man
a
gem
ent of Context-a
w
a
r
e
Cold
Chain V
ehicl
e No
des
(Ro
n
ling
Hu)
137
autore
g
ressiv
e predi
ction
algorith
m
, to
reali
z
e
the
node
dynami
c
a
w
a
k
e
n
ing
in the
do
rm
ant
state, extend
the sle
ep tim
e
,
and redu
ce the ene
rgy
con
s
um
ption.
These meth
o
d
s a
nalyze f
r
o
m
the routing al
gorithm a
nd other pe
rspe
ctives, and
thus red
u
ce the energy con
s
u
m
ption of nod
e
s
to a ce
rtain
extent, but all node
state
s
are sw
itch
ed from the
perspe
c
tive
of prob
ability or
forecast, not rigo
rou
s
and
pre
c
ise, and
redu
ce the
q
uality of service to a ce
rtain extent, so we
need to
find
a mo
re
rea
s
onabl
e an
d e
ffective mech
anism
for
en
ergy o
p
timiza
tion.In literat
ure
[12] we a
pply the co
ntext-awa
r
e m
e
tho
d
to
monito
r the environm
ent of medi
ci
ne man
agem
ent
place, and
re
alize th
e di
stributed m
u
ltiple so
urce
env
ironm
ent con
t
ext awarene
ss. At p
r
e
s
en
t,
some
re
se
arche
r
s
also a
pply the con
t
ext awar
e
n
e
ss to th
e en
ergy ma
nag
e
m
ent of sen
s
or
node
s.
Woo
d
et al. [13]
usi
ng the
bidi
re
ctional d
a
ta flo
w
, an
alyze
a
nd
study the
resi
dent
activ
i
ty
pattern
and f
eed b
a
ck to t
he net
wo
rk, t
o
re
alize the
context-a
w
a
r
e po
we
r ma
n
ageme
n
t; For the
study on e
n
e
r
gy co
nsumpt
ion opt
imi
z
ation problem o
f
indoor e
n
vironmental
qu
ality monitori
ng
sen
s
o
r
s, V Jelici
c
et al. [14] put forward
the co
ntext-aware nod
e
ener
gy mana
gement meth
od,
to decrea
s
e the activities o
f
nodes a
nd redu
ce en
ergy
demand.
The a
bove
work
ha
s ve
rified the
effect of
conte
x
t awarene
ss o
n
saving
ene
rgy
con
s
um
ption
of the wirele
ss se
nso
r
nod
es, but t
he cu
rre
nt study is only lim
ited to the netwo
rk of
fixed place.
Consi
d
e
r
in
g the goo
d perform
a
n
ce of co
ntext aware
ness in en
ergy
manag
eme
n
t.This pap
er h
a
s
pro
p
o
s
ed
the e
n
e
r
gy
manag
eme
n
t plan
of
cont
ext-awa
r
e
co
ld
chai
n logi
stics van
sen
s
or nod
es. A
c
co
rding
to di
ffe
rent ap
plication contexts, it
take
s
differe
nt
informatio
n a
c
qui
sition a
n
d
data tra
n
smissi
on
strat
egie
s
, to effectively red
u
c
e the e
n
e
r
g
y
con
s
um
ption
of the node
s and ext
end
the life cycle
of the whole
system
witho
u
t affecting the
norm
a
l ope
ra
tion of syste
m
.
2. The Sy
ste
m
Archite
c
tu
re of Energy
Manag
e
ment
With the ad
vancem
ent i
n
netwo
rkin
g and multi
m
edia te
chn
o
logie
s
en
a
b
les the
distrib
u
tion. A
l
though
en
cry
p
tion can
pro
v
ide multim
e
d
ia content
o
n
ce
a pi
ece o
f
digital conte
n
t
is de
crypted,
the dish
one
st
custo
m
er
ca
n redi
strib
u
te it arbitrarily
s.
The sy
stem architectu
re
of cold chain
logistic
s inte
rnet of thing
s
is sho
w
n in
Figure 1.
The system con
s
i
s
ts
of
th
ree part
s
:
informatio
n
a
c
q
u
isition network, rem
o
te monitori
ng ce
nter
and the Q
R
code tra
c
ea
bili
ty system.
Figure 1. The
system archi
t
ecture
of co
l
d
chai
n logi
stics inte
rnet of
things
The i
n
form
ation a
c
q
u
isitio
n net
wo
rk in
t
he
refri
gera
t
ed van
in
clu
d
ing
ra
dio freque
ncy
comm
uni
cati
on mod
u
le,
gatheri
ng
node
and
some no
de
s of environ
mental mo
ni
toring,
impleme
n
t th
e data
a
c
qui
sition
of car environme
n
t, and i
n
teract data
with
vehicle
ma
ste
r
terminal; Veh
i
cle control t
e
rmin
al, inclu
d
ing the mai
n
cont
rolle
r, GPS module,
GPRS mod
u
le,
power supply
module, etc.
, realize the
real
-tim
e env
ironm
ental m
onitorin
g
of refrige
r
ated
car
and i
n
tera
ct
data
with t
he
system
backg
rou
nd;
System b
a
c
kgro
und
is the
cont
rol
and
disp
atchi
ng
center of the
whol
e
cold
chain l
ogi
stics sy
stem, respon
sible
for the
storage
a
n
d
analytical
cal
c
ulatio
n of main data info
rmation.
At the sam
e
time, the system b
a
ckgroun
d is
the
main
sou
r
ce
of co
ntext informatio
n. The fr
a
m
e
w
o
r
k of
co
ntext-aw
are en
ergy mana
gem
ent
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 14, No. 1, March 2
016 : 136 – 1
4
3
138
system is
sh
own in Fig
u
re 2. The system fr
ame
w
ork
con
s
i
s
ts of the context information
acq
u
isitio
n, context-aware
mode
ling a
n
d
context-a
w
are ap
plicatio
n, and is cen
t
ralize
d
co
ntext
informatio
n manag
ement a
r
chite
c
tu
re.
Figure 2. The
framework of
contex
t-a
w
a
r
e energy man
ageme
n
t syst
em
Context info
rmation a
c
qui
sition i
s
mai
n
ly res
pon
sibl
e for a
c
q
u
irin
g all ki
nd
s of
context
informatio
n, and convertin
g
into approp
riate data fo
rmat after pret
reatme
nt. In this sy
stem, the
acq
u
ire
d
cont
ext informatio
n incl
ude
s th
ree a
s
pe
ct
s of
conte
n
ts: the
physi
cal
sce
ne info
rmatio
n,
the en
d-user
operating
info
rmation
an
d t
he b
a
ck
g
r
ou
n
d
ma
nag
eme
n
t ope
ratio
n
i
n
formatio
n. T
he
context information provid
es a data
b
a
sis for context-awa
r
e mo
deli
ng.
C
o
n
t
e
x
t-
aw
are
mo
de
lin
g p
l
a
y
s
a cr
uc
ia
l
r
o
le
in
the
wh
ole
system. T
he
system
efficiently
extract
s
mea
n
ingful info
rmation fro
m
vast c
onte
x
t informatio
n that is
acquire
d from
the
unde
rlying, a
nd ca
n effecti
v
ely organi
ze
and pre
s
e
n
t informatio
n, namely contex
t repre
s
e
n
tation
and tran
sformation. Data
base is u
s
ed
for h
e
lpin
g
the sy
stem
re
aso
n
out
th
e
cu
rrent cont
ext
types, storag
e context and
context usa
g
e
.
Context-a
w
a
r
e ap
plication
is the
final
pu
rpo
s
e
of
cont
ext awa
r
en
ess. Th
e e
s
tabli
s
hme
n
t
of co
ntext-a
w
are m
odel
is to p
r
ovi
de u
s
e
r
s
wi
th more i
n
telligent
se
rvice, a
n
d
pro
v
ide
person
a
lized servi
c
e
s
acco
rding
to
different contexts.
In this pa
per
,
the
context-a
w
are system
is
mainly eng
a
ged in lo
w-p
o
we
r optimi
z
ation proble
m
of sen
s
o
r
node
s. Accordin
g to different
appli
c
ation
co
ntexts, it tries to redu
ce th
e power
con
s
umption of ea
ch no
de
without affecting t
he
data accu
ra
cy and efficie
n
cy, and extend the life cyc
le of the system. Of co
urse, the ene
rgy
manag
eme
n
t is ju
st one
a
s
pe
ct of cont
ext-awa
r
e
ap
plicatio
n, and
it also
can
b
e
used in ve
h
i
cle
sched
uling,
crisi
s
m
ana
ge
ment a
nd
oth
e
r
aspe
cts. In
this pap
er we a
r
e
com
m
itted to
loo
k
ing
for
effective mea
n
s of sen
s
o
r
node en
ergy manag
em
e
n
t, and the cont
ext-awa
r
e ap
plicatio
n is the
energy mana
gement.
3. The Sy
ste
m
Archite
c
tu
re of Energy
Manag
e
ment
3.1. Energ
y
Cons
umptio
n Analy
s
is o
f
Wireles
s
Sensor
Node
s
The st
ru
cture
of wirel
e
ss
sen
s
o
r
no
de
s u
s
ually
co
nsi
s
ts of p
o
w
er
man
age
ment unit,
sen
s
o
r
, micro
c
ontrolle
r and
radio fre
que
ncy modul
e, as sho
w
n in
Figure 3.
Figure 3. Structure of
wirel
e
ss se
nsor n
ode
s
The ene
rgy consumption o
f
nodes mai
n
l
y
come
s from
the following
asp
e
ct
s:
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Energ
y
Man
a
gem
ent of Context-a
w
a
r
e
Cold
Chain V
ehicl
e No
des
(Ro
n
ling
Hu)
139
1) Sen
s
or m
o
dule
The
sen
s
o
r
module i
s
d
a
ta acqui
sition mod
u
le,
mainly re
sp
o
n
sibl
e for
sampling,
recupe
ratin
g
, conve
r
ting v
a
riou
s p
h
ysi
c
al env
iro
n
me
nts, and it
s
energy
co
nsumption can
be
s
h
ow
n
as
fo
llo
w
s
:
*
PW
N
E
(
1
)
Whe
r
ein,
W is the en
ergy con
s
um
ption fo
r singl
e
sam
p
ling, N
is
the nu
mber of
sampli
ng, E
is the
e
nerg
y
con
s
u
m
pti
on d
u
ri
ng i
d
l
e
time. F
r
o
m
the fo
rmu
l
a, the
sam
p
ling
freque
ncy of lowe
r nod
es
can red
u
ce en
ergy co
nsum
ption.
2) Micro
c
ont
roller
The mi
crocontroller as data processi
ng un
it, is
responsible f
o
r
cont
rolling signal
acq
u
isitio
n, data processing a
nd wirele
ss
com
m
unication
of the whol
e node
s. T
h
e
microcontroll
er typically has va
rio
u
s work
states, in
cludi
ng no
rm
al, idle, low powe
r
co
nsum
ption
and othe
r mo
des.
Figure 4. Value com
p
a
r
iso
n
s of wi
re
le
ss sen
s
or n
ode
power con
s
u
m
ption
3) Ra
dio fre
q
uen
cy modul
e
Radi
o fre
que
ncy mo
dule
as th
e wi
rel
e
ss commu
nication inte
rface of
sen
s
o
r
node
s, i
s
respon
sibl
e for the wirele
ss com
m
uni
ca
tion task
with
other node
s
or gate
w
ays. It can be see
n
from the
figu
re 4,
se
nso
r
s
and
proce
s
so
rs con
s
um
e
v
e
ry little e
nergy, while
the
radio
fre
que
n
cy
module
con
s
umes a
la
rg
e pe
rcenta
g
e
of
ene
rgy
in th
e n
o
d
e
en
ergy co
nsum
ption.
Radi
o
freque
ncy
mo
dule
co
nsum
es
ro
ughly th
e same
en
erg
y
in
sendi
ng, receiving and
idle state, wh
ile
the po
we
r
consumption
i
s
lo
w i
n
d
o
rmant
stat
e.
Therefore,
th
e no
de
s
sho
u
ld ma
ke
ra
dio
freque
ncy mo
dule keep in
dorm
ant state
without com
m
unication.
3.2. The Desi
gn of Energ
y
Manageme
n
t
Plan
The acqui
sition of system
context information
incl
ud
es the physi
cal context informatio
n,
the en
d-u
s
e
r
ope
ration
in
formation,
an
d the
ba
ckground
man
a
g
e
ment
ope
rat
i
on info
rmati
on.
The si
nk n
o
d
e
s conve
r
ge,
pretre
atment
and pa
ck
the gathe
red t
e
mpe
r
at
ure, physi
cal cont
ext
data, and
send to the gateway
nod
es
in driving
cab, the gate
way will
upload to the rem
o
te
monitori
ng ce
nter after furt
her p
r
o
c
e
ssi
n
g
. The vehicl
e sched
uling
and man
age
ment ope
rati
on
informatio
n in
the backgrou
nd are
sent to
the monitorin
g
cente
r
thro
ugh the Internet.
The m
onitori
ng
cente
r
re
pre
s
ent
s a
n
d
co
nver
ts th
e input
co
ntext informati
on, an
d
descri
b
e
s
the
s
e context informatio
n in the same
fo
rm, establi
s
hi
ng the co
ntext-a
w
are mod
e
l.
Acco
rdi
ng to
the esta
blish
ed context m
odel,
the
system output
s
a sp
ecifi
c
ap
plicatio
n
strategy, n
a
m
ely the en
ergy ma
nag
e
m
ent strate
g
y
. The gate
w
ay no
de
s receive i
n
st
ru
ction
s
from the m
o
nitoring
ce
nter, an
d send
to the si
nk
node
s a
nd a
c
qui
sition
no
des th
ro
ugh
the
wirel
e
ss
way. Thro
ugh th
e related
pa
ra
meters of
the
control software, it intellig
ently cha
nge
s the
freque
ncy of
sin
k
no
de
s a
nd colle
cted
sign
als,
the
numbe
r of d
a
ta sto
r
age t
he freq
uen
cy
of
radio
fre
que
ncy
comm
un
ication, th
e
volume of
wirel
e
ss
co
mmuni
cation
data, a
nd
thu
s
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 14, No. 1, March 2
016 : 136 – 1
4
3
140
effectively re
duces th
e en
ergy con
s
um
ption of
nod
e
s
, and
exten
d
s the life
cycle of the
wh
ole
system. In t
he p
r
a
c
tical
appli
c
ation,
we al
so
nee
d to con
s
ide
r
differe
nt control
strate
gie
s
according
to
different tran
sported
g
ood
s,
re
aliz
in
g th
e
balan
ce
of int
e
lligen
ce
an
d
ene
rgy
savin
g
.
Table 1 sho
w
s the co
ntext-awa
r
e en
ergy manage
ment
strategy.
Table 1. The
c
ontext-aware
energy mana
gement st
rate
gy
Context
model
Microcontroller
Sensor module
Radio freque
nc
y
communication
Loading
Activity
alternate
s
w
i
th
dormanc
y
Acquisit
i
on interval is t1
Each acquisition
n1 times
In transport
a
tion
Activity
alternate
s
w
i
th
dormanc
y
Acquisit
i
on interval is t2
Each acquisition
n2 times
Unloading
Activity
alternate
s
w
i
th
dormanc
y
Acquisit
i
on interval is t3
Each acquisition
n3 times
Empt
y
idle
Dormanc
y
Closed
Closed
Parking rest
Activity
alternate
s
w
i
th
dormanc
y
Acquisit
i
on interval is t4
Each acquisition
n4 times
Breakdo
w
n
maintenance
Activity
alternate
s
w
i
th
dormanc
y
Acquisit
i
on interval is t5
Each acquisition
n5 times
Acco
rdi
ng to
differe
nt co
ntexts, we
m
a
ke
corre
s
po
nding
adj
ust
m
ents for th
e working
mode
s of
sen
s
or no
de
s, su
ch a
s
adju
s
ting the tem
p
e
r
ature a
c
qui
si
tion, the time
interval of
rad
i
o
freque
ncy
co
mmuni
cation,
etc. In the stage of
loadi
n
g
, unloadi
ng,
parking rest
and b
r
ea
kdo
w
n
maintena
nce, the value
s
t1 and
n1
sh
ould b
e
re
du
ced
acco
rdin
gly as it is e
a
sy for th
e ca
r
temperature
t
o
have
an
om
alies,
we
sho
u
ld u
p
loa
d
th
e van
tempe
r
ature
to the
monitori
ng
ce
nter
at a hi
ghe
r f
r
equ
en
cy. Bu
t in tran
sp
ort
a
tion,
the ab
norm
a
l
temp
eratu
r
e ha
s small occu
rri
ng
prob
ability, we can in
crea
se the t1 and n
1
corre
s
p
ondi
ngly.
The working
state of nod
e
s
ca
n be d
e
scrib
ed a
s
an
orde
rly state
seq
uen
ce, su
ch a
s
[S1
→
S2
→
S3
→
S4
→
…
→
Sn], n wo
rk
states
ci
rculate co
nt
inuo
usly. We
assume
that th
e
corre
s
p
ondin
g
po
wer
co
nsumption of
e
a
c
h
state is
re
spe
c
tively P1
、
P2…Pn, the wo
rki
ng tim
e
of
each
state i
s
re
spe
c
tively T1
、
T2
…Tn.
The
ene
rgy
co
nsumptio
n of n
ode
s i
s
cal
c
ulate
d
as
follows (ign
ori
ng state switch time).
1
(*
)
n
ii
i
EP
T
(
2
)
In the process of actual u
s
e,
the workin
g state of se
nso
r
nod
es i
s
con
s
tantly switchi
ng.
Whe
n
n
ode
s
do n
o
t nee
d t
he temp
eratu
r
e a
c
q
u
isitio
n
and
wi
rele
ss com
m
uni
cati
on, the
se
nsor
module
an
d
radio f
r
eq
uen
cy commu
nication m
odule
sh
ould
be
cl
ose
d
, to m
a
ke no
de
s
kee
p
i
n
the state of
low po
we
r
con
s
um
ption
as far a
s
possibl
e. Un
der e
a
ch co
ntext of loading,
transpo
rtation
,
etc, the
working
state
of
node
s i
s
ch
a
nging
con
s
ta
ntly, and th
e
circulate
d
tim
e
interval of this state is al
so
chan
ging a
s
the context chang
es.
4. Experimental De
sign and Simulation Analy
s
is
4.1. Experimental Env
i
ro
nment Desig
n
The vehi
cle t
e
rmin
al sy
ste
m
desi
gne
d i
n
this
p
ape
r
con
s
i
s
ts of 1
gateway nod
e, 1 sin
k
node a
nd 4 a
c
qui
sition n
o
d
e
s.
This pap
er make
s
th
e
p
o
we
r con
s
um
ption
contrast experim
ent
on a
c
qui
sitio
n
nod
es.
The voltage
of acqui
sition
nodes i
s
3.3
V,
the
wirel
e
ss tran
smi
s
sion fre
que
ncy is 2.454 GHz,
and the
outp
u
t power of
radio fre
que
n
c
y is 1
DB
M.
We me
asure the current
of experim
e
n
ta
l
node
s in
different
wo
rki
n
g co
ndition
s,
and th
e
Ta
ble 2
sh
ows the test
re
sults.
With t
he
advan
ceme
nt in networkin
g and
multim
edia te
chn
o
lo
gies ena
ble
s
the
dist
ributio
n
and sh
arin
g
of
multimedia
content wid
e
ly. In the me
antime,
pira
cy become
s
i
n
crea
singly rampant a
s
the
cu
stome
r
s can ea
sily d
u
p
licate
and
redistri
bute th
e re
ceive
d
multimedia
content to a
l
a
rge
audie
n
ce. critical [1]. Althoug
h en
cry
p
tion c
an p
r
ovide multim
edia
conte
n
t once a
piece of
digital co
nten
t is decrypted
, the dishon
e
s
t
cu
stome
r
can redi
stri
but
e it arbitrarily.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Energ
y
Man
a
gem
ent of Context-a
w
a
r
e
Cold
Chain V
ehicl
e No
des
(Ro
n
ling
Hu)
141
Table 2. The
measured pa
ramete
rs of di
fferent state
s
of node
s
State Parameters
Node ene
rg
y
consumption
Condition
S1 Activation
3.515mA
The internal cr
yst
a
l oscillat
or as system clock, only
CPU w
o
rks.
S2
MCU+ Sensor m
odule
4.525mA
Sensors
acquire, convert and sen
d
data to MC
U.
S3
MCU+ Radio fr
e
quenc
y
transmission
29.015mA
External 32
M clock, in the sending state, the po
w
e
r is
1dBm.
S4
MCU+Radio fre
q
uenc
y
r
e
ceiving
25.846mA
External 32
M clock, in the receiving state.
S5
Dormanc
y
0.109mA
It can enter the
n
o
rmal
w
o
rking m
ode b
y
inte
rrupti
on
Figure 5. Wo
rking
state tra
n
sition di
agra
m
The no
de
s i
mmediately
enter a
do
rm
ant st
ate
aft
e
r
ea
ch acq
u
isition data and
radio
freque
ncy co
mmuni
cation,
and the next roun
d of state
circulate
s
au
tomatically when it’s the time
for the
set sl
e
ep time
r. Accordin
g to
the
state
se
tting
of Tabl
e 2, th
e Fig
u
re
5
sh
ows the
worki
n
g
state tran
sitio
n
model of the desi
gn no
d
e
s.
The A nod
e acq
u
ire
s
tem
peratu
r
e a
n
d
upload
s d
a
ta on 6 min
u
tes a
s
a
cycl
e. The B
node u
s
e
s
th
e context-aware e
ner
gy manag
ement pl
an, intelligent
ly controlli
ng
the tempe
r
at
ure
acq
u
isitio
n a
nd the f
r
eq
ue
ncy of
data u
p
loadi
ng, na
mely the valu
es
of t and
n
in Ta
ble
1.
The
node e
nergy con
s
um
ption
can b
e
cal
c
ul
ated as follo
ws:
1
234
EE
E
E
E
(
3
)
In the above formula, E1, E2, E3, E4 a
r
e re
sp
ectivel
y
loading, tra
n
sp
ortation, rest, the
unloa
ding
ph
ase
of th
e
energy
con
s
umption.
T
h
e Ta
ble
3 li
sts the
co
ntext-awa
r
e
en
ergy
manag
eme
n
t strategy of the B node for ea
ch sta
g
e
, in table Tw expre
s
se
s the interval o
f
temperature
acq
u
isitio
n, and Ct
i
s
th
e
tempe
r
at
u
r
e a
c
qui
sition
interval
of
radio
fre
que
ncy
comm
uni
cati
on.
4.2. Data
An
aly
s
is and Result
Acco
rdi
ng to
the expe
rim
ental conditi
on
setting,
we u
s
e
MAT
L
AB to gen
e
r
ate the
simulatio
n
da
ta, and get 6
grou
ps of ra
n
dom ex
peri
m
ental data as
sho
w
n in Tab
l
e 4. Combin
ed
with the
actu
al produ
ction
co
ndition
s, t
he n
ode
s n
e
ed 2
s
fo
r a
d
a
ta a
c
qui
sitio
n
, and
both t
he
radio frequ
en
cy re
ceiving a
nd se
ndin
g
n
eed 30 m
s
.
Table 3. Power co
nsumpti
on optimizi
n
g
strategy of B node for ea
ch stage
Time interval
T
w
/min
Ct/ time
Loading
1
1
Transpo
rtation
5
2
Stopping to rest
3
1
Unloading 1
1
To
save e
n
e
r
gy, the no
de
s imm
ediatel
y enter
a d
o
rmant stat
e wi
thout data
a
c
quisitio
n
and
radio
fre
quen
cy tra
n
smissi
on a
nd
receiving ta
sk. According
to the po
wer
con
s
um
ption
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 14, No. 1, March 2
016 : 136 – 1
4
3
142
para
m
eters shown
in
the Table 2,
com
b
ined
with
th
e form
ula
2 a
nd formula
3,
we
calculate
the
power
co
nsu
m
ption of th
e nod
e B. We also
ca
l
c
ul
ate the p
o
we
r con
s
umptio
n of A no
de
by
formula 2
an
d formula
3, and set ea
ch pa
ram
e
te
r acco
rdin
g to the co
ntext-awa
r
e e
n
e
r
gy
manag
eme
n
t strategy in th
e Table 4.
Table 4. Each time perio
d distrib
u
tion of
rando
mly ge
nerate
d
tran
sportation ta
sk
Experimental
group
Total
transportation
duration /h
Loading
duration
/min
Rest
duration
/min
Unloading
duration
/min
Cy
cle
count of A
node
activity
Acquisit
i
on
times of B
node
temperatu
r
e
Ti
mes
of B
node radio
frequenc
y
communication
1 11.97
13.45
22.84
24.50
120
140
85
2 23.67
11.81
51.71
14.60
237
263
151
3 10.92
15.11
26.37
21.52
109
130
81
4 28.94
27.17
54.18
26.21
289
325
189
5 20.51
28.22
54.77
18.08
205
239
145
6 20.01
23.99
49.37
29.77
200
234
143
For the no
d
e
s of usi
n
g
context-a
w
a
r
e ene
rgy m
anag
ement
strategy, in
orde
r to
cal
c
ulate
co
n
v
eniently, we
ca
rry o
n
the
statistics
at variou
s
stag
es of tran
sp
orta
tion. We
sele
ct
grou
p 3 and
6 grou
ps of
experim
ental
data, to
list
the energy consumpt
ion curves.
From th
e
picture, we
can see m
o
re
intu
itively the co
ntext-aware e
n
e
r
gy
m
anag
ement
strategy ha
s v
e
ry
obviou
s
ene
rgy saving effect. The Figu
re 6 and Fi
gu
re 7 sh
ow the
simulation
re
sults.
Figure 6. Co
mpari
s
o
n
of two ru
nnin
g
re
sults in
the 3rd g
r
ou
p
Figure 7. Co
mpari
s
o
n
of two ru
nnin
g
re
sults
in the 6th gro
u
p
For the
sim
u
lation results of 6 g
r
ou
ps
of expe
ri
ments, we
analyze the
energy
con
s
um
ption,
as
sh
own in
Table
5. Th
roug
h t
he
an
alysis
co
ntra
st, the conte
x
t-awa
r
e e
n
e
r
gy
manag
eme
n
t plan ma
ke th
e averag
e en
ergy
co
nsum
ption of node
s red
u
ce by about 21%.
Table 5. Co
m
parative an
al
ysis of
all 6 group
s experi
m
ental re
sults
Experimental g
r
o
up
number
Node ene
rg
y
con
s
umption
under ope
ration
w
i
thout
str
a
tegy
Node ene
rg
y
con
s
umption
under ope
ration
of context
-
aware strat
e
g
y
Saved energ
y
consumption
Energ
y
saving
proportions /
%
1 8384.72
6731.99
1652.73
19.71
2 16574.50
12226.61
4347.90
26.23
3 7650.30
6335.53
1314.77
17.18
4 20264.18
15258.33
5005.85
24.70
5 14362.09
11452.20
2909.89
20.26
6 14011.94
11289.43
2722.52
19
.
43
Mean value
21.25
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Energ
y
Man
a
gem
ent of Context-a
w
a
r
e
Cold
Chain V
ehicl
e No
des
(Ro
n
ling
Hu)
143
The expe
rim
ental re
sult
s have achieve
d
t
he expe
cted effect, an
d verified the
context-
awa
r
e
ene
rg
y manag
eme
n
t method
ca
n more effe
ctively redu
ce
the po
we
r consumption
and
extend the life of the limited ene
rgy no
des. Mo
reov
er, as th
e tra
n
sp
ortation
di
stan
ce (or ti
me)
increa
se
s,
th
e
no
des, whi
c
h use
the context-aware stra
tegy
fo
r e
nergy mana
g
e
ment,
will
h
a
ve
more o
b
viou
s advantage
s
of low po
wer
con
s
um
ption.
5. Conclusio
n
In view
of tha
t
the en
ergy-con
strai
ned
n
ode
s in
the
system affe
ct the
se
rvice
life of th
e
whol
e sy
ste
m
due
to the
sh
ort life
cycle, in th
is pa
per
we
propo
se
a
meth
od of
co
ntext-aware
energy ma
na
gement. T
he
data a
c
q
u
isiti
on, p
r
o
c
es
sin
g
, sto
r
ag
e, transpor
t
of n
o
des chan
ge
as
the co
ntext chang
es, an
d
also
gre
a
tly redu
ce
the e
n
e
rgy con
s
um
ption of nod
e
s
when m
eeti
n
g
the intellige
n
t wo
rk. In
this pap
er, the
method i
s
ap
plied to th
e
vehicle
syst
e
m
of cold
ch
ain
logisti
cs inte
rnet of things,
which ha
s well solv
ed the
defect that the vehicl
e no
des h
a
s a
sh
ort
cycle, imp
r
ov
ed the efficie
n
cy of
the lo
gistics tra
n
sp
ortation, an
d greatly re
du
ced the logi
stics
co
st. In this pape
r, our fu
rther
work is to
explore a more
sci
e
n
tific context-awa
r
e mo
del
ing
method o
n
th
e ba
sis of the
existing sy
stem, and
furth
e
r expa
nd th
e appli
c
ation
rang
e of cont
ext
awa
r
en
ess.
Ackn
o
w
l
e
dg
ements
This
work
wa
s su
ppo
rt
ed by the Scien
c
e a
n
d
Tech
nolo
g
y Project
s
of
Huaia
n
(HAG
201
402
8) and th
e
University Science
Re
sea
r
ch Proje
c
t of Ji
ang
su Provi
n
ce
(15K
JB520
00
4, 13K
JB510
004) an
d the
Scie
nce
a
n
d
Te
chn
o
logy Proje
c
t
of Jia
ngsu
Province
(BE2015
127
) and the S
c
ientific F
o
u
ndation P
r
oj
ect of Huai
yin Institute of Tech
nol
ogy
(HGC1
412
)
a
nd the
Intern
et of Thin
gs
Develo
pment
Speci
a
l F
u
n
d
of
Chin
a in
201
3 a
nd th
e
Small and Me
dium Enterp
ri
se Te
chn
o
log
y
Innovation Fund of Chi
n
a in 2013.
Referen
ces
[1]
Defraey
e
T
h
ijs, Cronje Paul, Berry
T
a
rl.
T
o
w
a
r
d
s Integr
ated Perform
a
nce Eva
l
uati
o
n of F
u
ture
Packag
i
ng for
F
r
esh Prod
uce
in the C
o
ld
Ch
ain.
T
r
ends in F
ood
Sci
ence &
T
e
chno
lo
gy
. 201
5; 44(2)
:
201-
225.
[2]
García LR, Elorza PB, Rodríguez-Berm
e
jo J, et
al. Monitori
ng the Intermodal, Refrig
erat
ed
T
r
ansport
of F
r
uit Using
Sensor N
e
t
w
or
ks.
Spanish Jo
urna
l of
Agric
u
l
t
ural Res
earch
.
2007; 5(2): 1
4
2-15
6.
[3]
Bachma
nn C,
Ashou
ei M, Pop V
.
Lo
w
-
po
w
e
r W
i
reless
Sensor N
odes
for Ubiqu
i
tou
s
Long-T
e
r
m
Biome
d
ica
l
Sig
nal Mo
nitori
ng
.
IEEE Comm
unications Magaz
ine
. 20
12; 50(
1): 20-27.
[4]
Sari, L
y
di
a. Routing Ra
ptor code for energ
y
-ef
f
icient
w
i
re
le
ss bod
y
ar
ea net
w
o
rk data transmissi
o
n
.
T
E
LKOMNIKA
T
e
leco
mmunic
a
tion, Co
mputi
ng, Electron
ics
and Co
ntrol
. 2
015; 13(
1): 277
-283.
[5]
Gamecho Borj
a, Silva Hugo, Guerreir
o Jose
.
A
C
ontext-Aw
are
Ap
plic
atio
n
to
Increase El
derl
y
User
s
Compl
i
a
n
ce
w
i
th Ph
y
s
ic
al Re
hab
ilitati
on Exerci
ses at Home via
Anim
atro
nic Biofee
db
ac
k.
Journal of
m
e
dic
a
l system
s
. 20
15; 39(1
1
): 296-3
02.
[6]
Perera C, Z
a
slavsk
y
A, Christ
en P
,
Georgak
opo
ulos
D. Co
nte
x
t
Aw
ar
e Comput
i
ng for the Internet of
T
h
ings: A
Surve
y
.
IEEE Comm
unic
ations S
u
rveys & T
u
torials
. 201
4; 6(1):
414-4
54.
[7]
T
ong En-do
ng,
Shen Qian
g, Lei Ju
n, Liu
Y
u
. Stud
y
o
n
C
onte
x
t-a
w
a
r
e
T
e
chn
o
lo
gi
es for Internet of
T
h
ing.
Comput
er Scienc
e
. 20
1
1
; 38(4): 9-1
4
.
[8
]
R
o
sma
w
arni
N
e
ny
.
A
mo
b
i
l
e
e
c
o
t
ou
ri
sm re
co
mme
n
dati
o
n
s
sy
ste
m
u
s
i
n
g
ca
rs-con
tex
t
aw
are
appr
oach
e
s.
T
E
LKOMNIKA
T
e
lec
o
mmun
icat
ion,
Co
mputin
g, Electronics
and C
ontrol
. 2
013; 1
1
(
4
):
845-
852.
[9]
Y
u
Lei-le
i
, CHAI Qiao-lin, W
A
NG Chun-l
e
i. Energ
y
A
d
a
p
tive
Routing
A
l
g
o
rithm for W
S
N
based on Mi
n
Imum Hops.
A
pplic
atio
n Res
earch of Co
mp
uters
. 2007; 2
4
(
1
1
): 236-
238.
[10]
Y
unfen
g Lin
g
, Hon
g
tao Z
h
a
n
g
, Guiru Gu.
A
Prin
tab
l
e CNT
-Based F
M
Pas
s
ive W
i
rel
e
ssS
ensor
T
ag o
n
a F
l
exib
le Sub
s
trateW
ith Enh
ance
d
Sens
itivi
t
y
.
IEEE Sensors Journal
. 201
4; 14(4): 1
1
9
3
-
1
19
7.
[11]
Lin Ji
n-zh
ao,
Li Guo-
jun,
Z
h
ou
Xi
ao-n
a
. D
P
M-base
d
Met
hod for
T
r
acking Ma
neuv
eri
n
g
T
a
rgets in
Wireless Sens
or Net
w
orks.
Journ
a
l on C
o
mmu
n
ic
ations
. 2
010; (12): 9
0
-9
6.
[12]
Li Z
hen
g, Ma
Y
o
n
g
-tao, Li
u Kai-h
ua. D
e
sig
n
an
d Implem
ent
atio
n
of Environm
e
n
tal Co
nte
x
t
Aw
are
ness S
ystem Based o
n
Internet of
T
h
i
ngs.
Co
mp
uter
Engin
eeri
n
g
. 2
012; 38(
17): 1-
4.
[13]
W
ood
A, Stan
kovic JA, V
i
r
o
ne G, Sel
a
vo
L, He
Z
,
Ca
o
Q, et al. Cont
ext-a
w
a
r
e W
i
r
e
less S
ens
or
Net
w
orks for
Assisted Liv
i
ng
and R
e
sid
enti
a
l Monitori
ng.
IEEE Network
. 2008; 22(
4): 26-
33.
[14]
Jelicic V
,
Mag
no M, Brunel
li
D, Paci G. C
ont
e
x
t-ad
aptiv
e Multimo
dal
W
i
reless Sens
or Net
w
ork for
Energy
-Ef
f
icient Gas Monitoring.
IEEE Sensors Journal
. 201
3; 13(1): 32
8-3
38.
Evaluation Warning : The document was created with Spire.PDF for Python.