TELKOM
NIKA
, Vol.11, No
.3, Septembe
r 2013, pp. 5
63~574
ISSN: 1693-6
930,
accredited
A
by DIKTI, De
cree No: 58/DIK
T
I/Kep/2013
DOI
:
10.12928/TELKOMNIKA.v11i3.1056
563
Re
cei
v
ed Ma
rch 1
8
, 2013;
Re
vised July
16, 2013; Accepted July 2
9
,
2013
Wireless Sensor Network for Forest Fire Detection
M. Y. Hariy
a
w
a
n
1
, A. Guna
w
a
n
2
, E. H. Putra*
3
Dep
a
rtment of Electrical E
ngi
neer
ing, Po
lite
k
nik Calt
e
x
Ria
u
Jala
n Umba
n Sari No 1 Pek
a
nbar
u, T
e
lp. (0
761) 5
3
9
3
9
*Corres
p
o
ndi
n
g
author, e-ma
i
l
:
y
a
nuar
@pcr.
a
c.id, agu
n@
p
c
r.ac.id, emans
a@pcr.ac.i
d
Abs
t
rak
Kebak
ara
n
hut
an
meru
pak
an
salah s
a
tu
masal
ah ya
ng
me
ng
anca
m
k
e
lestar
ian
huta
n
. Siste
m
penc
eg
aha
n d
i
ni u
n
tuk i
n
d
i
ka
si keb
a
kar
an
hutan
san
gat
dip
e
rluk
an. L
u
a
snya
huta
n
me
nj
adi s
a
l
ah
satu
mas
a
l
ah ya
ng
dih
ada
pi
dal
a
m
p
e
m
a
n
tau
a
n
kond
isi h
u
t
an.
Untuk
me
ng
atasi
mas
a
la
h i
n
i
,
diranc
ang
su
atu
sistem
d
e
teksi kebak
aran
h
u
tan den
ga
n me
nga
do
psi
J
a
ri
n
gan
Sens
or N
i
r
k
abe
l (W
irel
es
s Sens
or N
e
tw
ork)
me
ng
gun
aka
n
beb
erap
a n
o
d
e
sens
or. Seti
ap n
o
d
e
se
ns
or me
mil
i
ki
mi
krokontro
ler, p
e
manc
ar/pe
ner
ima
dan ti
ga se
nso
r
. Metode pe
ng
ukura
n
di
lakuk
an d
eng
an
me
nguk
ur suh
u
, a
p
i, tingk
at met
ana, h
i
drok
arb
o
n
,
dan
CO2
di
b
e
bera
pa k
a
w
a
s
an
huta
n
d
a
n
me
ng
ukur
pe
mbakar
an
ga
mb
ut di
se
bua
h s
i
mu
lator.
Dari
h
a
sil
pen
guk
uran su
hu, kadar
met
ana, gas
h
i
dro
k
arbo
n dan C
O
2 di daer
ah
terbuka
me
nu
njukk
an tidak
ad
a
tanda-ta
nd
a k
ebak
aran
kar
ena
ni
lai
su
h
u
, metan
a
, g
a
s hi
drok
arbo
n, da
n CO2
ada
lah
di
b
a
w
ah
pen
guk
uran d
i
ruan
g si
mul
a
to
r.
Ka
ta
k
unc
i:
de
teksi kebak
ara
n
hutan, j
a
rin
g
an sens
or nirk
abe
l, nod
e sen
s
or
A
b
st
r
a
ct
Forest fires ar
e one of
problem
s
that threaten sustainability of
the for
e
st. Early pr
evention syst
em
for indications
of fore
st fires is absolutely necessary
. T
h
e
extent of the f
o
rest
to b
e
o
n
e of the
pro
b
le
ms
enco
unter
ed i
n
the forest co
n
d
itio
n
mo
nitori
ng. T
o
ov
erco
me
the
prob
le
ms
of forest e
x
tent, desi
gne
d
a
system
of fore
st fire detecti
o
n
syste
m
by
a
dopti
ng
th
e Wi
reless S
ens
or
Netw
ork (WSN) usin
g
mu
ltipl
e
sensor n
odes.
Each sensor
node h
a
s a
microc
ontro
l
l
e
r, transmitter
/
receiver a
n
d
three sensor
s.
Measur
e
m
ent meth
od is
p
e
r
forme
d
by measur
ing
t
he
temp
eratur
e, fla
m
e, t
he lev
e
ls
of meth
an
e,
hydroc
arbo
ns, and CO2 i
n
so
me forest
are
a
and the co
mb
ustion of pe
at in
a si
mul
a
tor. F
r
om results of
me
asur
e
m
ents
of temp
eratur
e, levels
of me
thane, a
hy
dro
c
arbo
n gas a
n
d
CO2 in a
n
o
pen ar
ea i
n
d
i
c
a
tes
there
are
no
si
gns
of fires
d
u
e
to t
he v
a
l
ue
of th
e
te
mper
a
t
ure,
metha
ne,
hydr
ocarb
o
n
gas, a
n
d
CO2
is
bel
ow
the me
a
s
ure
m
e
n
t in the space si
mula
tor.
Ke
y
w
ords
: forest fire detectio
n
, w
i
reless sen
s
or netw
o
rk, sensor no
de
1. Introduc
tion
Fores
t
s cons
titute a
s
o
urc
e
of life for all livi
ng things
.
With the fores
t, s
y
mbios
i
s
,
and the
chai
n
can
run
ou
r life.
Wildf
i
res a
r
e
a
se
ri
ous p
r
obl
e
m
f
a
cin
g
tod
a
y. It has be
en
fel
t
by the
publi
c
and g
o
vern
m
ent. Land
an
d fore
st fire
s in Ria
u
p
r
o
v
ince h
a
s a
con
s
id
era
b
le
effect on th
e
occurre
n
ce o
f
haze p
o
llution acro
ss st
ate boun
dar
i
e
s. In gen
eral, fires in Riau province
in
peatlan
d
s do
minate the
re
gion
by 60%.
The
r
efo
r
e smog con
s
titutes natural
p
h
enome
non
th
at
comm
only occurs du
rin
g
fire se
ason a
nd have an i
m
pact on n
e
i
ghbo
ring
cou
n
tries
su
ch a
s
Malaysia a
n
d
Singapo
re.
To redu
ce th
is p
r
obl
em, the gove
r
nm
e
n
t ha
s lau
n
ched
seve
ral
prog
ram
s
ori
ented to
a
n
appe
al or
a l
egal
san
c
tion
again
s
t any
action th
at threate
n
s t
h
e
su
staina
bility of the fore
st.
Ho
wever,
althoug
h the
en
tire work p
r
o
g
ram
ha
s b
e
en de
sig
ned,
the forest
s still sho
w
s a p
r
etty
poor [1].
In the term of the current techn
o
logi
cal
developme
n
ts, su
staina
bility program
s tend t
o
requi
re
a
sy
stem
cap
abl
e of a
nalyzi
ng a
nd m
o
n
i
toring i
ndi
ca
tion of forest fires.
Wirel
e
ss
techn
o
logy th
at is able to
send d
a
ta without
u
s
ing
wire
s is exp
e
cted to be
come one of the
developm
ent
of a
pplied
tech
nolo
g
y that
can
s
uppo
rt forest
co
nservatio
n
p
r
og
ram.
Th
e
monitori
ng sy
stem is expe
cted to pre
s
e
n
t data in t
he form of a fire indication for the vast la
nd
even.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 11, No. 3, September 20
13: 56
3 – 574
564
1.1. Wireless
Sensor Ne
tw
o
r
k (WSN)
Wirel
e
ss Sen
s
or
Network
(WSN) i
s
a u
n
i
t of
the mea
s
urem
ent
p
r
o
c
ess, comp
uting,
an
d
comm
uni
cati
ons that pro
v
ide admini
s
trative c
apa
b
ilities to a device, observation, and any
treatment for any events and ph
eno
mena that
o
c
cur in th
e environ
ment
using
wirel
e
ss
t
e
chn
o
logy
.
This
sy
st
em
is mu
ch mo
r
e
ef
f
i
ci
ent th
an the u
s
e o
f
cable
s
. Thi
s
sy
stem ha
s a
function fo
r v
a
riou
s type
s
of appli
c
ation
s
. WS
N
te
ch
nology p
r
ovid
es th
e foun
d
a
tion to
con
d
u
ct
experim
ents
on the
environment. F
o
r example,
bi
ologi
sts
want
to monito
r t
he b
ehavio
r
of
animal
s
in th
eir habitat, e
n
vironm
ental
research
ers requi
re a system ca
pabl
e of monitori
ng
environmental pollution, farmers
can increase yields
by examinin
g soil fertility, geologi
sts
need a
system to monitor sei
s
mi
c activity, even
in the m
ilitary also needs a system that able to monit
o
r
area
s that are difficult to achi
eve. Overall human a
c
t
i
vities r
equi
re
monitorin
g
WSN
system
s.
In som
e
of t
he literature
related
to th
e
implem
ent
ation of
WS
N that u
s
e
s
a multi-
sen
s
o
r
y, dev
elope
d ma
n
y
robot
s [4]
[5]. In 198
9 [6], su
gge
sts th
e imp
o
rtan
ce
of the
developm
ent
of multisen
so
r to impr
ove i
n
telligent
syst
ems. Multi
s
e
n
so
r ap
plications i
n
ro
botics,
biomedi
cal
sy
stem
s, monit
o
ring
eq
uipm
ent, rem
o
te sensi
ng and
transport
sy
ste
m
s de
scribe
d
[7]
in 200
2. David L. Hall an
d
Jame
s
Lliná
s
explain th
eoretical int
r
od
u
c
tion to m
u
ltisen
so
r fu
sion
of
data [8]. They provide tuto
rials on
data fu
sion, t
he d
a
ta
fusion a
pplications, p
r
o
c
e
ss m
odel
s, a
nd
the ide
n
tifica
tion of a
ppli
c
ation
en
gin
eerin
g,
which aim
s
to
show ho
w th
e sen
s
or fu
ssion
measure to g
e
t results. Th
ey also
sho
w
a flow
chart
to explain th
e different
ways to co
nne
ct
multiple
sensors i
n
one device. In [
9
] multis
en
so
r
is
us
e
d
fo
r
mon
i
to
r
i
ng
w
e
ld
in
g in th
e
automotive in
dustry, with a
mult
i-se
nsory, it is possib
l
e to
measure cu
rre
nt, voltage and
po
we
r
weldi
ng. The
use of multisen
so
r also can b
e
se
en
in [10], who desig
ned a l
i
ght-ad
d
ressa
b
le
potentiomet
r
i
c
sen
s
or
(lap
s) a
s
the
real
ization
of p
o
rtable multise
n
so
r dev
i
c
e. The
light sou
r
ce
and ele
c
tro
n
i
c
oscillato
r in
clud
e multiple
xer, pre
-
ampli
f
ier and a hig
h
-pa
s
s filter.
WSN comp
o
nents in
clud
e
sen
s
ors, wireless
mod
u
le
s, and PC. All compo
nent
s will form
a monitori
ng
system that i
s
able to di
splay the
data
in the form of sen
s
or
ch
ara
c
teri
stics
are
use
d
to ta
ke
advantag
e of
wirele
ss
me
dia. Becau
s
e
it can
be
used to
sha
r
e
appli
c
ation
s
,
use
sen
s
o
r
type selecte
d
ba
se
d on appli
c
ati
on.
Table 1. Type
of Sensors
Ty
pe
o
f
Senso
r
s
Senso
r
Temper
ature
Thermistor,
th
er
mocouple
Pressure
Pressure gauge,
baromete
r
, ionization gauge
Optic Photodiodes,
ph
ototransis
tors, infrared sensors,
C
CD
sensors
Acoustic
Piezoelectric
resonators, microph
ones
Mechanical
Strain gauges, ta
ctile sensors, ca
pacitive diaphragms,
piezoresistive cells
Movement and V
i
bration
Accelerometers, g
y
roscopes, phot
o sensors
Position
GPS, ultrasound
-
based s
ensors, infrared
-based se
nsors,
inclinometers
Humidit
y
Capacitive and resistiv
e sensors, h
y
gr
ometers, ME
MS-
based humidit
y
sensors
Radiation
Ionization detecto
rs, Geiger
–Muel
ler counters
1.2.
Wireless Se
nsor Net
w
o
r
k Archi
t
ec
tu
re
Each no
de WSN system
s
gene
rally co
n
s
ist of
sen
s
in
g, processin
g
,
communi
cat
i
on and
power that
can be illust
rated as in
Figure 1.
How to com
b
i
ne this is the thi
n
g to watch
out
for
whe
n
we d
o
the de
signin
g
. The system
pro
c
e
s
sor
i
s
the most imp
o
r
tant
part of the WSN
syst
em
that can affect the perform
ance or en
ergy con
s
umpti
on. Several o
p
tions fo
r the
pro
c
esso
r m
a
y
cho
o
s
e
incl
ud
e:
Microcontroll
er
Digital
si
gnal
pro
c
e
s
sor
Applicatio
n-specifi
c
IC
Field progra
mmable g
a
te array
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
9
30
Wirel
e
s
s
Sen
s
or
Netw
or
k for Fo
re
st
Fire Detec
t
ion
(M.Y. Hariy
a
wan)
565
Figure 1. Gen
e
ral a
r
chitect
u
re of a WS
N
A.
Uvt
r
on Flam
e
Detecto
r
Flame
se
nso
r
i
s
u
s
e
d
to
detect th
e p
r
ese
n
ce of fi
re that in
dicate a fi
re. In
pl
annin
g
to
use fi
re
sen
s
or mad
e
Hamamat
s
u type R286
8.
T
h
is fo
rm of
comm
uni
cati
ng sen
s
or ci
rcuit
C37
04 kit
tha
t
is
o
nboa
rd, as sh
own
in Figure
2.
T
h
i
s
sen
s
o
r
will detect UV
lig
ht
in
the 185
-260
nm interval d
e
tected a
s
UV light from the fire.
The
s
e sen
s
o
r
s a
r
e not able to detect the si
ze of
the fire, be
ca
use
ba
sically the cig
a
rette at a dist
a
n
ce up to 5
met
e
rs can b
e
d
e
tected
by these
sen
s
o
r
s. In realization, flame sen
s
o
r
is activa
ted wh
en the rated
voltage of 350 Vdc, so it is
necessa
ry to
activate
C37
04
kit. The i
n
put volt
age
required fo
r th
is
kit C3
704
9-30
Vd
c ran
ge,
but if we have a fixed voltage of 5 volts, we si
mply
mengin
putka
nnya to terminal 'O' on the
kit
becau
se ba
si
cally the input
voltage of 9-30
Vdc ea
rlie
r will be di
reg
u
lator be 5 vo
lts.
Figure 2. Block
circuit C3
7
04.
The wo
rki
ng prin
ciple
of
th
is seri
es
of kit
C3
704
is cha
nging
the
sup
p
ly voltage
of 5 volt
s
to 350 volts
DC at the
Hi
gh Voltage
DC to DC c
o
n
v
erter to a
c
tivate the sen
s
or.
While Si
gnal
Processin
g
Circuit
se
rve
s
t
o
regul
ate
the
amou
nt of i
n
comin
g
p
u
ls
e
s
from the
UV
Tron
sen
s
or f
o
r
2 se
co
nd
s
wil
l
be respon
de
d by the
C37
04 a
pul
se
wi
dth of 10
ms.
Output p
u
lse
s
of 1
0
m
s
ca
n
be co
uplin
g
the output capa
citor to p
r
odu
ce
a wi
der p
e
rio
d
. In planni
ng, i
t
is use
d
by
1
microfara
d
s
capa
citor to produ
ce an o
u
tput of 1 seco
nd peri
o
d on t
e
rmin
al Cx C3704
kit.
B.
MQ-2 Gas Sensor
MQ
-
2
ga
s
s
e
n
s
or
is
h
i
gh
ly s
e
ns
itive
to
s
u
bs
ta
nc
es
Sn
O
2
, w
h
ic
h
is
lo
w
c
o
nd
uctivity
in
clea
n air. When flamma
ble ga
s is
detecte
d,
the sen
s
o
r
wil
l
be highe
r con
d
u
c
tivity
and
con
c
e
n
tration
of the gas
also in
crea
se
d. By us
ing simple el
ect
r
ocircuit it will
be conve
r
te
d
according to the co
ndu
ctivity chang
e of t
he output si
g
nal of gas
co
nce
n
tration.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 11, No. 3, September 20
13: 56
3 – 574
566
MQ-2
sen
s
o
r
also h
a
s hi
g
h
sen
s
itivity to LP
G, Prop
ane an
d Hyd
r
ogen, an
d u
s
ually are
u
se
d
for
methane a
n
d
flammable vapors an
d more. MQ2 is
suitabl
e for smoke a
nd ga
s dete
c
tion, also
relatively inexpen
sive. The physi
cal form
of the MQ-2 sen
s
o
r
is sho
w
n in Figu
re
3.
Figure 3. MQ
-2 Ga
s Sen
s
o
r
C.
LM35 Tem
p
e
r
ature Sensor
LM35 tem
p
e
r
ature
se
nsor i
s
a tran
sdu
c
e
r
that
ha
s a
functio
n
to co
nvert the tem
peratu
r
e
scale into
el
ectri
c
al
quan
tities in the f
o
rm of
volta
ge. LM3
5
ha
s hig
h
a
c
curacy an
d ea
se of
desi
gn
wh
en
comp
ared to
t
he oth
e
r tem
peratu
r
e
sen
s
or, LM
35
also
ha
s
a lo
w
ou
tput impe
dan
ce
and
high
line
a
rity so that
it ca
n
be
ea
sil
y
con
n
e
c
t
ed with
a
seri
es of
sp
eci
a
l co
ntrol and
do
e
s
n
o
t
requi
re fu
rthe
r adj
ustme
n
t. The tem
perature
ca
n be
m
easure
d
by th
e IC i
s
q
u
ite
wide, b
e
twe
e
n -
55
C to 15
0
C. Figu
re 4 shows the ba
si
c ci
rcuit LM
3
5
.
Figure 4. The
Basic
Circuit LM35
D.
Nod
e
Sensor
System se
nsor no
de
con
s
ists of 3 pi
eces
of
sen
s
o
r
s, micro
c
ontrollers an
d RF modul
e
KYL 500s, a
s
sho
w
n in Fig
u
re 5.
Figure 5. Sensor
Nod
e
To be abl
e to sen
d
serial
data over th
e air at lea
s
t
need a d
e
vice that can
make th
e
pro
c
e
ss of la
ying a digital serial d
a
ta to the carrie
r freque
ncy to a highe
r freq
uen
cy and th
en
emitted into t
he air.
One e
x
ample of th
e RF m
odul
e
that can
do
this is th
e m
odule KYL
5
00S
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Wirel
e
s
s
Sen
s
or
Netw
or
k for Fo
re
st Fire
Detectio
n (M
.Y. Hariya
wa
n)
567
Wirel
e
ss
Dat
a
Tra
n
sceive
r. Wirele
ss
Data Mo
dule
KYL 500S tran
sceivers can tran
smit
and
receive
seri
al
data th
rou
g
h
the me
dium
of air,
with
4
33/868/9
15 MHz
fre
quen
cy
and
ba
ud rate
up to 38,40
0 bps. Use
of these mod
u
le
s is q
u
ite pr
a
c
tical b
e
cau
s
e of the fairly small in
size
and
can
be
conn
e
c
ted
dire
ctly t
o
the
RS23
2. The
mod
u
le
works
with
su
pply bet
wee
n
3.3 to
5 V
D
C.
In one modul
e can b
e
use
d
as the sen
der an
d re
cei
v
er. Serial da
ta to be tran
smitted thro
u
gh
the RF
mod
u
l
e
KYL 50
0S fed to the
microcontroller serially. Simila
rly, the data
received will
be
taken by a
serial mi
croco
n
trolle
r. The
dist
an
ce
coul
d be go
ne th
roug
h ab
out 100 m - 5 K
m
.
Baudrate in
the air 12
00
bp
s, 240
0
bps,
480
0
b
p
s, 9
600
bp
s, 19
200
bp
s, or 38
400
bps.
Modulatio
n G
F
SK (Gau
ssi
an Frequ
en
cy Shift Keyin
g
) is
used
so
it is anti-inte
rf
eren
ce
and B
E
R
(Bit Erro
r
Rat
e
) i
s
lo
w. Th
e phy
sical
form of
the
RF t
r
an
sceiver K
Y
L 500S
mo
dule i
s
sho
w
n in
Figure 6.
Figure 6. KYL 500S Data T
r
an
sceiver M
odule
ATMEGA853
5 micro
c
o
n
troller serve
s
as the ce
ntra
l data processor te
m
perature, sm
oke an
d
flame whi
c
h
will then b
e
sent to com
p
u
t
er
thro
ugh
wirele
ss
com
m
unication cha
nnel
s. At each
node th
ere
are
3 pie
c
e
s
of se
nsors
are tem
p
e
r
at
ure
se
nsors, smo
k
e
and
flame an
d
RF
transceive
r
m
odule KYL eq
uippe
d 500
s. Circuit at each node i
s
sh
o
w
n in Figu
re
9.
While th
e mo
dule
s
that a
r
e on the P
C
as a
se
rver
a
r
e fun
c
tionin
g
wirel
e
ss Tx/
R
x KYL
500
s that fun
c
tion receives the data fro
m
the mete
r
and the
n
se
nt to a com
pute
r
thro
ugh a
seria
l
cabl
e.
2. Rese
arch
Metho
d
To dete
c
t fo
rest fire
s th
en
perfo
rme
d
measur
ement
s of te
mpe
r
a
t
ure, flame, l
e
vels
of
methane,
ga
soline, CO an
d CO
2 fro
m
the bu
rnin
g p
eat. There a
r
e seve
ral
sce
nario
s to
perf
o
rm
these me
asurements, na
m
e
ly:
1.
Measurement
without burni
ng peat in the
spa
c
e sim
u
la
tor
Fire sen
s
ing
system u
s
es three se
nso
r
s to d
e
tect wh
ether a conditio
n
can be
con
s
id
ere
d
a
s
early sympt
o
m of fire or not. In th
is fir
e
sen
s
ing
system used 3 pi
ece
s
of sen
s
ors
that will dete
c
t fire, sm
oke and tem
p
e
r
ature cha
n
g
e
s. Th
en the
sen
s
in
g sy
stem is
put in
an
encl
o
sed sp
a
c
e sim
u
lator
and tempe
r
at
ure mea
s
u
r
e
m
ents, levels
of methane, gasoli
ne, CO a
nd
CO2 in the
sp
ace
simulato
r.
2.
Measu
r
em
en
t of burned p
eat in the enclose
d sp
ace simulator
At this stage
of peat bu
rned ga
rb
age
inco
rpo
r
ated
into the sim
u
lator. Th
e result
s of
measurement
s from three sens
ors viewe
d
and sto
r
ed
on a com
pute
r
appli
c
ation
prog
ram.
3.
Measurement
of levels of air in the city
At this stage
, observatio
n
and testing
in an open
area in the
town of Duri to take
measurement
s of temperature, le
vels
of methane, hy
dro
c
a
r
bo
ns
a
nd CO
2 that can b
e
used
as
an indic
a
tor of fire.
4.
RF Module T
e
sting
The te
st aim
s
to
dete
r
mi
ne the
pe
rfo
r
man
c
e
an
d
sign
al p
r
op
a
gation
ch
ara
c
teri
stics
(pathlo
s
s, p
o
w
er d
ensity,
power
and
se
nsitivity) of K
Y
L 20
0U
in shado
we
d
a
r
e
a
s (o
utdoo
r), and
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 11, No. 3, September 20
13: 56
3 – 574
568
the measurement result
s
obtained will
be compared
with the
results of the calcul
ation
RS
SI
(Re
c
eive
d Signal Strength
Indication
). Sensin
g sy
st
em use
d
to perfo
rm mea
s
ureme
n
ts in
the
scena
rio sho
w
n in Figu
re
7.
Figure 7. Measu
r
em
ent of burn
ed pe
at in the encl
o
se
d spa
c
e
simul
a
tor
3. Results a
nd Analy
s
is
In Table 2
can be
see
n
result
s of me
asu
r
em
ents of
temperature,
levels of
methane,
hydro
c
a
r
bo
ns, CO and CO
2 in a simula
tor without
th
e burni
ng pe
at. From the results of the
s
e
measurement
s the magnitu
de of te
mperature, metha
ne, hydro
c
a
r
bon
s, CO an
d CO2 a
r
e st
able
durin
g the proce
s
s of colle
cting data a
n
d
in
accordan
ce with n
o
rm
al air co
nditio
n
s.
Table 2. Re
sults of the me
asu
r
em
ents
wi
thout bu
rne
d
peat in the encl
o
sed spa
c
e
simulat
o
r
Data
Tem
p
erat
ure
(
o
C)
Metana
(pp
m
)
Hidrokar
bo
n
(pp
m
)
C
O
(p
pm
)
C
O
2
(p
pm
)
87637
31
1
41
5
97
87638
31
1
41
5
97
87639
31
1
41
5
97
87640
31
1
41
5
97
87641
31
1
41
5
97
87642
31
1
41
5
97
87643
31
1
41
5
97
87644
31
1
41
5
97
87645
31
1
41
5
97
87646
31
1
41
5
97
87647
31
1
41
5
97
87648
31
1
41
5
97
87649
31
1
41
5
97
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Wirel
e
s
s
Sen
s
or
Netw
or
k for Fo
re
st Fire
Detectio
n (M
.Y. Hariya
wa
n)
569
3.1. Measure
ment of the
burning pea
t
in the space
simulator
In Figure 8 it can be seen
the result
s of m
easu
r
eme
n
ts of temperature, and le
vels of
methane fo
r
burni
ng p
eat
in the sim
u
lat
o
r. From
the
results of te
mperature
m
easure
m
ent
s, the
magnitud
e
of
the tempe
r
at
ure
ro
se to
3
4
°
C
comp
a
r
ed to
31
° C with n
o
bu
rn
ing pe
at in th
e
simulato
r. Wh
ile methan
e l
e
vels in
crea
se dramat
icall
y
in an ave
r
a
ge of 7
ppm
comp
ared to
only
1 ppm witho
u
t
burning p
eat
in the simula
tor.
Figure 8. Re
sults of mea
s
u
r
eme
n
ts of
temperature a
n
d
methane g
a
s
levels
in the encl
o
sed sp
ace sim
u
lator
In Figure 9, it can be seen
the results of
the measure
m
ent of hy
dro
c
arbon g
a
s a
nd CO
2
levels fo
r b
u
rning p
eat in
the
simulato
r.
From th
e m
e
asu
r
em
ent of
hydro
c
a
r
bo
n
ga
se
s in
crea
sed
dram
atically t
o
an
ave
r
ag
e
of 95
ppm
co
mpared to
4
1
ppm
with
n
o
burni
ng
peat
i
n
the
sim
u
lat
o
r.
While
the ave
r
age
level
of
CO2
eq
ual to
97
ppm
with
out bu
rnin
g p
eat in the
si
m
u
lator, alth
ou
gh
at one point h
ad increa
se
d dram
atically to 103 ppm.
Figure 9. Re
sults of mea
s
u
r
eme
n
ts
of hydro
c
a
r
bo
n ga
s and
CO2 le
vels
in the encl
o
sed sp
ace sim
u
lator
From the
bu
rning p
eat in the sim
u
lato
r
sho
w
e
d
an in
cre
a
se in tem
peratu
r
e, leve
ls of metha
n
e
,
a
hydro
c
a
r
bo
n gas, an
d CO
2 and me
asurem
ent re
sults ca
n be
use
d
as a
n
indicator of the
pres
enc
e
of a fores
t
fire.
31.5
32
32.5
33
33.5
34
34.5
87650
87700
87750
87800
87850
87900
o
C
sequence
number
0
2
4
6
8
10
87650
87700
87750
87800
87850
87900
Metan
a
sequence
number
89
90
91
92
93
94
95
96
97
87650
87700
87750
87800
87850
87900
ppm
sequence
number
96
97
98
99
100
101
102
103
104
87650
87700
87750
87800
87850
87900
sequence
number
ppm
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 11, No. 3, September 20
13: 56
3 – 574
570
3.2. Measure
ment of lev
e
ls of air in th
e to
w
n
o
f
Du
ri
In Figure 1
0
it can be see
n
the re
sults
of
measurem
ents of temp
eratu
r
e, and l
e
vels of
methane in t
he ope
n are
a
in the town of Duri. Fr
om
the results o
f
temperatu
r
e
measu
r
em
e
n
ts,
the magnitu
d
e
of the temp
eratu
r
e rose to 34 °
C
com
pare
d
to 31
°
C with
no b
u
rning p
eat in t
he
simulato
r. While methan
e
levels rise to
a point 4 ppm while then
back down to 1 ppm with
no
burni
ng pe
at in the simulat
o
r.
Figure 10. Re
sults of me
as
urem
ents of tempe
r
ature a
nd methan
e g
a
s level
s
in the town of
Duri
In Figure 11 i
t
can be se
e
n
the results
of
measu
r
em
ents of hydro
c
arbon g
a
s a
nd CO
2
levels in th
e
open
are
a
i
n
the town o
f
Duri. F
r
om
the mea
s
u
r
ement of hy
dro
c
a
r
bo
n ga
se
s
increa
sed q
u
i
te dramati
c
a
lly to 80 ppm comp
are
d
to 41 ppm with no bu
rn
ing peat in the
simulato
r. Wh
ile the averag
e level of CO
2 equal to 97
ppm witho
u
t burni
ng pe
at in the simulat
o
r.
Figure 11. Re
sults of me
asurem
ents of h
y
drocarbon g
a
s an
d CO
2
levels in the ci
ty of Duri
From the re
sults of measurem
ents of tempe
r
ature, levels of methane, a hydroca
r
bo
n
gas
and
CO2
in an
ope
n a
r
ea i
n
the to
wn of
Du
ri
sh
ows no
si
gns of fore
st
fire
s. It can
be
seen
from the values
of temperatu
r
e, methane,
hydrocarbo
n ga
s, and CO
2 are still below the
measurement
results in
a
simulato
r. Although
th
e levels of met
hane
and ot
her hyd
r
o
c
a
r
bon
gases con
s
id
erably high
er than normal,
but this is
more du
e to the factor of air pollution fro
m
heavy vehicle
s
in the town
of Duri.
31.5
32
32.5
33
33.5
34
34.5
100000
101000
102000
103000
104000
105000
o
C
sequence
number
‐
1
0
1
2
3
4
5
100000
101000
102000
103000
104000
105000
Meth
an
sequence
number
0
10
20
30
40
50
60
70
80
90
100000
101000
102000
103000
104000
105000
Hi
drokarb
o
n
(ppm
)
sequence
number
95.8
96
96.2
96.4
96.6
96.8
97
97.2
100000
101000
102000
103000
104000
105000
C
O
2 (ppm
)
sequence
number
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Wirel
e
s
s
Sen
s
or
Netw
or
k for Fo
re
st Fire
Detectio
n (M
.Y. Hariya
wa
n)
571
3.3. RF Mod
u
le Testing
3.3.1.
Powe
r Re
cei
v
ed Measure
m
ent
Measurement
s were
perfo
rmed in o
r
de
r
to det
ermi
ne
the trend
of receive
d
po
wer of the
distan
ce.
Ta
ble 3 i
s
the
measurement
of aver
age
p
o
we
r of
each
node
re
ceive
r
with
3 diffe
rent
baud
rate.
Table 3. Re
ceived Powe
r
Level
Distance
(meter
)
Receive Pow
e
r
Level (
d
Bm)
Data
1200 bps
Data
9600 bps
Data
19200 bps
30 OK
-87
OK
-35.33
OK
-81.33
60 OK
-90
OK
-35.67
OK
-91
90 OK
-92
OK
-36.33
OK
-97.67
120 OK
-93
OK
-36.67
NOT
OK
-100.33
150 OK
-95
OK
-36.67
-
-
180 OK
-97
OK
-37
-
-
210 OK
-98.33
OK
-37.33
-
-
240
NOT OK
-102
NOT OK
-42.67
-
-
Figure 12. Power
Re
ceived
Measu
r
em
en
t
Ac
c
o
rding to the Table
III and Figure 12, it c
a
n be
s
e
en
that the average power
rec
e
ived
at ea
ch
nod
e
will
be
sm
all
e
r if th
e
dista
n
ce
is fa
rth
e
r away, or it
can b
e
said
th
at the
quality
of
data re
ceived
by the recei
v
er will get worse t
he lon
g
e
r it along
wi
th the distan
ce. Data can
be
viewed
on th
e re
ceived
p
o
we
r at no
de
baud
rate 1
2
00 bp
s a
nd 1
9
,200 b
p
s i
s
very small, e
v
en
inclu
d
ing the
stand
ard
no
ise (<-80
d
Bm), but du
e
to the se
nsiti
v
ity of the module KYL
-2
00U
rea
c
he
d 12
3
d
Bm, then even su
ch
a small po
wer
re
ceive
r
can still receive th
e data well. As for
the ba
ud
rate
9600
bp
s, th
e received
co
nsid
era
b
le
po
wer i
s
still in
good
qu
ality. In this test, t
h
e
averag
e max
i
mum tra
n
sm
issi
on di
stan
ce in
a sha
dowed a
r
ea
of the mod
u
l
e
is 3
10 m
for
baud
rate
120
0 bp
s, 23
0 m
for b
aud
rate
9600
bp
s, an
d 90
m to
19
200
bp
s ba
u
d
rate. It is qu
ite
different from
the maximum transmissi
on dista
n
ce
o
f
the module
s
on the co
ndit
i
on of LOS (L
ine
Of Sight) liste
d on the data
s
he
et, which is 100
0m.
3.3.2.
Powe
r De
nsit
y Mea
s
u
r
em
ent
Based o
n
the
Fig 13, it can be seen that
t
he power de
nsity will decrease with increasi
ng
distan
ce bet
ween Tx and Rx. This
is because the dist
ance is invers
ely propo
rtion
a
l to the power
dens
i
ty.
‐
98
,
6
7
‐
40
‐
98,
33
‐
110
‐
100
‐
90
‐
80
‐
70
‐
60
‐
50
‐
40
‐
30
‐
20
‐
10
0
30
60
90
120
150
180
210
Po
w
e
r
Le
v
e
l
(d
B
m
)
Dis
t
a
n
c
e
(
m
et
er
)
1200
9600
19200
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 11, No. 3, September 20
13: 56
3 – 574
572
Figure 13. Power
Den
s
ity Measurement
3.3.3. Pathloss
Mea
s
urem
ent
Based
on
Fig
u
re
14, it can
be
see
n
that
if
the farthe
r
anatara Tx a
nd Rx, the
greater it
s
pathlo
ss
well.
This i
s
co
nsi
s
tent with
the
re
su
lts
of m
easure
m
ent
s of po
wer lev
e
ls
re
ceived
by
Rx can b
e
se
en in table
s
4
.
1 and 4.2 as well as
4.6 a
nd 4.7 the previous
cha
r
t, the gre
a
ter th
e
distan
ce b
e
twee
n Tx and
Rx will be smaller th
e
n
the average
power received by the Rx
for
gro
w
ing lo
sse
s
that occu
rre
d, following th
is
equation
11,727
ln
56,
999
.
Figure 14. Path Loss Mea
s
urem
ent
3.3.4.
RSSI (Re
c
ei
ved Signal
Strength Indi
cati
on) Me
asure
m
ent
Figure 1
5
exp
l
ains that the
RSSI value g
enerated
will
be
smalle
r if t
he di
stan
ce
b
e
twee
n
the Tx and Rx farther awa
y
. Thus RSSI value is
inve
rsely p
r
op
orti
onal to the value of pathlo
ss.
RSSI value is obtaine
d
in accorda
n
ce
with
the data re
ce
ived by the receiver p
o
we
r
measurement
before, that the gr
eater th
e distan
ce b
e
t
ween Tx a
n
d
Rx, the small
e
r the received
power. T
hat
way, the
d
a
ta obtai
ned
from th
e
result
s of
me
asu
r
em
ents
with d
a
ta
RSSI
prop
ortio
nal
calculation.
Just recei
v
ed
power nearin
g th
e cal
c
ulate
d
value of RSSI
measurement
results b
aud
rate is
960
0b
ps. By
doing
so, ba
udrate
9600
bp
s are suitabl
e for u
s
e
in system
s de
sign
ed.
0
50
100
150
200
250
30
60
9
0
120
150
1
8
0
210
240
27
0
300
330
36
0
w/
m2
Dis
t
a
n
c
e
(m
)
x1
0
-7
y
=
11,
72
7l
n
(
x)
+
56
,
9
99
0
10
20
30
40
50
60
70
80
90
100
30
60
9
0
12
0
1
50
180
210
24
0
2
70
3
0
0
330
36
0
dB
Dist
a
n
c
e
(m
)
Evaluation Warning : The document was created with Spire.PDF for Python.