TELKOM
NIKA
, Vol.14, No
.1, March 2
0
1
6
, pp. 110~1
1
8
ISSN: 1693-6
930,
accredited
A
by DIKTI, De
cree No: 58/DIK
T
I/Kep/2013
DOI
:
10.12928/TELKOMNIKA.v14i1.2454
110
Re
cei
v
ed Au
gust 17, 20
15
; Revi
sed
No
vem
ber 2
5
, 2015; Accepte
d
De
cem
ber
11, 2015
Underwater Channel Characterization to Design
Wireless Sensor Network by Bellhop
Nim
a
Bahr
a
m
i*, Nor Hisham
Haji Kh
am
is, Am
eru
ddin Baha
ro
m
,
Az
li Yah
y
a
Dep
a
rtment of Electrical E
ngi
neer
ing, Un
iver
siti T
e
knologi
Mala
ysi
a
(UT
M
), Skudai, Mala
ysi
a
*Corres
p
o
ndi
n
g
author, e-ma
i
l
: bnima
2@l
i
ve
.utm.my
A
b
st
r
a
ct
Acoustic u
nde
rw
ater link du
e to low
atten
uatio
n is e
m
p
l
oyed for u
n
d
e
r
w
a
ter sensor
netw
o
rk
(UW
S
N). Due
to w
a
ter chan
gin
g
phys
i
cal
prop
erties
a
n
d
different env
ir
on
me
ntal
co
nd
itions, the so
u
nd
prop
agati
on fo
r every und
er
w
a
ter chann
el
are different
.
So, the most
imp
o
rtant par
ameters to de
sign
relia
bl
e UW
SN
are: th
e s
h
a
d
o
w
z
o
ne
d
e
ter
m
i
nati
on,
opti
m
u
m
n
odes
p
l
ace
m
e
n
t, hi
gh
sign
al to
n
o
is
e
ratio
and
more
pow
er efficie
n
cy of
nod
es. T
h
is p
a
per si
mula
tes
u
nderw
a
ter ch
a
nne
l for 7 K
H
z
carrier fre
que
n
cy,
w
i
th measur
ed
physical w
a
ter prop
erties and re
mo
te sensi
ng data.
Based o
n
defi
ned sce
nari
o
, th
e
researc
h
det
ermi
nes
opti
m
u
m
n
o
d
e
s pl
ac
ement a
nd l
i
n
k
bud
get a
nal
ysis bas
ed o
n
botto
m to sur
f
ac
e
acoustic
li
nk. T
he
pa
per
util
i
z
es Be
llh
op
ac
o
u
stic too
l
b
o
x a
s
a s
i
mul
a
tor,
GeoMapA
pp
p
r
ogra
m
to c
o
ll
e
c
ts
bathy
metry dat
a and Avis
o+
datab
ase to d
e
termin
e
w
i
nd
speed o
n
sea
surface. As a result the pa
per
based on sim
u
lation of sound pr
opagation in channel and transm
i
ss
ion loss deter
m
i
nat
ion in depth and
rang
e, finds th
e opti
m
u
m
n
o
d
e
s positi
ons a
n
d
lin
k b
u
d
get calcul
atio
n to pr
ove the resu
lts.
Ke
y
w
ords
:
Un
derw
a
ter chan
nel ch
aracteri
zation, Un
derw
a
ter sound pr
op
agati
on, Un
der
w
a
ter link bud
g
e
t,
Und
e
rw
ater no
de pl
ace
m
e
n
t, Bell
hop
Copy
right
©
2016 Un
ive
r
sita
s Ah
mad
Dah
l
an
. All rig
h
t
s r
ese
rved
.
1. Introduc
tion
Re
cently, ma
ny re
sea
r
che
s
focus on
un
derwa
te
r wi
re
les
s
s
e
n
s
or
n
e
twor
k
(U
WS
N)
due
to wide em
pl
oyment in Military, Oil indu
stry,
Fishin
g, Biology and
climate ch
ang
es dete
c
tion [
1
].
Due to p
o
o
r
perfo
rman
ce
of high fre
q
u
ency
cau
s
e
of high atten
uation, the a
c
ou
stic
ban
d
is
utilized for lo
ng-h
aul un
de
rwate
r
co
mm
unication.
Thi
s
band h
a
s
many limitations
such as:
low
spe
ed si
gnal
travelling (15
00 m/s) and l
i
mit band
widt
h (KHz) [2]. Therefore, th
e impleme
n
t of
UWS
N
fa
ce
s to many
difficulties.
On
the ot
he
r h
a
nd, the
und
e
r
wate
r
nod
es by re
ason
of
compl
e
xity and wate
r resi
stant pa
ckagi
ng in hig
h
pressure situ
ation are expen
sive. It equal
s t
o
low n
ode
den
sity and lo
ng
er rang
e bet
wee
n
no
de
s i
n
UWSN. T
h
e nod
e pla
c
e
m
ent in lo
ng-hau
l
comm
uni
cati
on is ve
ry co
mplex, due t
o
tran
smi
ssi
o
n
loss, si
gnal
to noise rati
o and m
u
ltip
ath
cal
c
ulatio
n. T
h
is
pap
er fo
cu
se
s o
n
u
n
derwa
te
r cha
nnel simul
a
tion
to determ
i
ne
lin
k budg
et
betwe
en no
d
e
s.
1.1. Under
w
ater Soun
d Propaga
tion
Acou
stic p
r
op
agation in wa
ter is ba
sed o
n
S
nell-Descartes la
w. Thi
s
law dete
r
mi
nes ray
refra
c
tion
be
haviors a
c
co
rding to
many
factor
s
su
ch
as: i
n
ci
dent
and
refle
c
tion
angl
es and
th
e
spe
ed profile for two medi
a
.
The Figure 1 and E
quati
on (1
) illustrates Snell
-
Descarte
s la
w [2].
12
12
co
s
β
co
s
β
cc
(1)
A
ccu
ra
cy
co
n
d
it
ion:
21
1
(
or
co
s
ββ
12
/
cc
)
(2)
Refra
c
tion i
s
cau
s
e of diffe
rent speed
b
e
twee
n two
medium
s. Th
e tran
smitted
ray has
three
conditio
n
s: Fi
rst,
ray
penetrates to
other m
edi
a. Secon
d
,
refle
c
tion ray
travels between
two
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Und
e
rwate
r
Cha
nnel
Cha
r
acte
ri
zation t
o
De
sign
Wireless Sen
s
o
r
Network by…
(
Nim
a
Bahra
m
i
)
111
media i
n
terfa
c
e
(critical a
n
g
le), th
e Equ
a
tion
(3)
represe
n
ts th
e fo
rmula.
Third,
the ray come
s
back to its incid
ent medi
a again (tot
al refl
ectio
n
),
when this
phen
omen
on
occurs whi
c
h
transmitted
ray angl
e i
s
h
i
gher tha
n
cri
t
ical a
ngle.
T
he Fi
gure
1
sho
w
s the
to
tal
reflection.
A
wave path b
e
t
ween two poi
nts will be n
a
m
ed an a
c
ou
stic ray [2].
Figure 1. Refl
ection a
nd ref
r
actio
n
of plane wave
Figure 2. Refraction of a wave with a discontin
uou
s a
nd
contin
uou
s
a
rccos
c
β
(
12
/
cc
)
(3)
Whe
n
the me
dia layer is m
o
re than two then
the la
w b
e
com
e
s li
ke a
s
Equation
(4
):
1
1
ii
ii
co
s
β
cos
β
cc
(4)
The ref
r
a
c
tion relation al
o
ng co
ordi
nate
z can b
e
gen
erali
z
ed into:
C
o
ns
t
a
nt
co
s
β
Z
CZ
(5)
The Fig
u
re
2
sho
w
s the refractio
n
waves
with discontinuo
us a
n
d
co
ntinuou
s sou
nd spee
d
distrib
u
tion wi
th depth [2].
1.2. Signal Atte
nua
tion a
nd Loss in Under
w
a
t
er
Signal attenu
ation is d
u
e
to the viscous
ab
sorption ph
enom
e
non, it refe
rs to fluid
resi
stan
ce
a
gain
s
t flow.
The b
o
ri
c a
c
id and
magn
esium
sulfate
salt a
r
e the
two imp
o
rta
n
t
sea
w
ate
r
ma
terials, whi
c
h
are cau
s
in
g
this
ph
eno
menon. T
he
sign
al attenu
ation is
heav
ily
depe
ndent
o
n
employe
d
f
r
equ
en
cy. If the carrier fre
quen
cy is
hig
her th
an h
u
n
d
red
he
rtz, th
e
attenuation
can be de
scrib
ed
by Thorp equatio
n [2, 3].
α
=
2
2
0.
1
1
f
f
+
2
2
40
4100
f
f
+
2.7
5
Χ
42
10
f
+
0.003
(6)
The p
r
op
agat
ed si
gnal i
s
f
a
ce
d to
surfa
c
e a
g
itat
ion, t
h
is o
c
cu
rre
nce is
rea
s
o
n
o
f
signa
l
disp
ersion a
n
d
energy loss. The surfa
c
e
loss i
s
co
mp
utable by ma
ny models
su
ch a
s
: Eckhart,
Schul
kin
-
Ma
rch is
[2, 3].
The
signal
after hitting to
seafloor defle
cted and l
o
ses its inten
s
ity as
well. Th
e
bottom
loss is d
epen
dent on
seafl
oor
coverage
and t
opo
gra
phy, and ba
sed on utilize
d
frequ
en
cy i
s
cal
c
ulable by LFBL, HFBL
and HFEVA
model [3].
1.3. Under
w
ater
Noise
Gene
rally, un
derwate
r noi
se divided into
four se
ction
s
: turbulen
ce,
ship
ping, waves an
d
thermal n
o
ise
[4]. The extent of the freq
uen
cy in
ea
ch gro
up is
different fo
r exa
m
ple: turbul
e
n
ce
noise is exte
nded fo
r f < 1
0
Hz
and thi
s
rang
e for
shi
pping n
o
ise i
s
the fre
quen
cie
s
between
10
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 : 110 – 1
1
8
112
to 100 Hz. Th
e most im
port
ant noi
se that
can
cove
r a
c
ousti
c ba
nd is surfa
c
e
wav
e
noi
se, whi
c
h
is sp
rea
d
on
100 Hz until 5
0
KHz. The e
quation (7) d
e
scrib
ed wave noise [3, 5].
10log
w
Nf
=
50 +
7.5
1/
2
w
+2
0 log f - 40log (f + 0.
4)
(7)
Whe
r
e W i
s
wind
spe
ed b
a
se
d on mete
r per
se
con
d
(m/s).
1.4. Under
w
ater ch
annel
The sig
n
ifica
n
t of underwater ch
ann
el
charac
te
risti
c
is time-va
r
ying behavio
r. Every
phen
omen
on
whi
c
h i
s
a
b
l
e
to chang
e
medium
su
ch
as
su
rface a
g
itation,
se
afloor
und
ulatio
n
and phy
sical
prope
rtie
s of water, are the rea
s
o
n
s of ch
ann
el time-varyi
ng pro
p
e
r
ty [6].
Und
e
rwate
r
chann
el i
s
co
nsi
s
t of ma
n
y
duct
s
t
hat sep
a
rate
d
by
different so
u
nd spe
ed.
T
h
is
differen
c
e i
s
the re
sult of chang
es in p
h
y
sical
wate
r p
r
ope
rtie
s (te
m
perat
ure,
salinity and de
pth)
[2, 3]. To si
mulate a
c
curate und
erwat
e
r
chan
nel m
odel, the
rem
o
te se
nsi
ng i
n
formatio
n from
netwo
rk n
ode
s po
sition
s an
d CTD
(cond
uctivity
, temperatu
r
e, dept
h) data a
r
e e
s
sential.
1.5. Sound speed
Sound vel
o
cit
y
in water is
function
of p
r
essure, temp
eratu
r
e
and
salinity. It mea
n
s th
e
sou
nd
spe
e
d
betwe
en
surface a
nd
sea
bottom ha
s
variation
(de
pend
on de
p
t
h), and it wi
ll
modify with
cl
imate chan
ge
s a
nd
se
ason
s a
s
we
ll. T
h
ere
are ma
ny model
s
su
ch
as Ma
cken
zi
e,
UNES
C
O,
Chen and
Millero for SSP cal
c
ulation, ba
sed on the water
physi
cal
properties
range.
This
pap
er
selecte
d
Ma
cken
zie fo
rmu
l
a re
garding
to ca
se
stud
y CTD
mea
s
urem
ent. Th
i
s
equatio
n i
s
a
c
curate
for th
e temp
eratu
r
e 0
to 3
5
°C,
sali
nity 0 to
40 p
a
rt
s p
e
r
thousand
an
d
depth 0 to 10
00 m. The eq
uation (8
) sho
w
s the M
a
cke
n
zie'
s
eq
uatio
n
[2, 3]
.
SS = 1449.2
+
4.6T
−
0.055
2
T
+0.0002
9
3
T
+ (1.
3
4
−
0.01T
) (S
−
35)
+ 0.01
6 D
(8)
Whe
r
e T is te
mperature in
degree
s Cel
s
i
u
s, S is
salini
t
y in parts per thousa
nd an
d D is equ
al to
depth in mete
rs.
1.6. Under
w
ater no
de
No
wad
a
ys, the resea
r
che
s
focu
s on sen
s
or n
ode
s wit
h
smalle
r si
ze
, ultra powe
r
saving
,
more
sto
r
ag
e to b
u
fferin
g
, high
process
cap
a
ci
ty
and
more
band
width
efficien
cy. In fact
,
unde
rwater n
ode
s are m
o
dems
whi
c
h
sen
s
o
r
s a
r
e
con
n
e
c
ted to
them. The node ta
sk i
s
informatio
n collectio
n from
se
a an
d tran
smitted to
oth
e
r n
ode
s, g
e
nerally. T
h
is i
n
formatio
n m
a
y
inclu
de data
from Oxygen
, sedime
nt, CTD
se
ns
ors or processe
d data such
as p
o
sitioni
n
g
,
detectio
n
and
tracking, exp
l
oring p
r
o
c
ed
ure
s
[7].
The nod
e maintena
nce
and install
a
tion pro
c
e
s
s in se
afloo
r
due to no
des n
on-
availability are expe
nsive
and
co
stly. So, the b
a
tte
ry
re
pla
c
eme
n
t in
sho
r
t time
is i
m
po
ssibl
e
.
This
pro
b
le
m will b
e
solving with
reliable
de
sig
n
of commu
nicatio
n
lin
k and b
e
st
n
ode
placement. In this pap
er,
all simulatio
n
s a
r
e ba
se
d on acou
stic rel
e
a
s
e
s
modem by 86
7
-
A
model from T
e
ledyne Bent
hos
Comp
an
y. The Table 1 sho
w
s feature
s
of this m
odem [8].
Table 1. Mod
e
m feature
s
o
f
867-A mode
l
Maximum depth
305 meter
Receiver frequen
cy
7-15 KHz
Transmitter f
r
equ
enc
y
7-15 KHz
Tx po
wer
192 dB
Batter
y
life (
T
x)
250,000 Pings
Battery
life (Rx)
2
Y
e
a
r
1.7. BELLHO
P
Bellhop
is a
n
a
c
ou
stic to
olbox fo
r
ray
tr
a
c
ing
p
r
ed
iction,
whi
c
h
is code
d b
a
s
ed
on
FORT
RA
N a
nd MATLAB
simulato
rs f
o
r all pc pl
atforms. Bellho
p
is able to
simulate
wav
e
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Und
e
rwate
r
Cha
nnel
Cha
r
acte
ri
zation t
o
De
sign
Wireless Sen
s
o
r
Network by…
(
Nim
a
Bahra
m
i
)
113
prop
agatio
n i
n
water
col
u
mns
with
co
nsid
eratio
n o
f
sou
nd
sp
e
ed p
r
ofile, to
p an
d b
o
ttom
boun
dari
e
s,
reflection
s a
n
d
Tx, Rx qua
lification
s
. Fo
r stu
d
y on th
eoreti
c
al b
a
ckgroun
d refer to
referenc
e [9].
The Fig
u
re
3
depict
s the
input and
ou
tput f
iles for
the toolbox. All informatio
n from
cha
nnel
qual
ity, Tx and
Rx char
acte
ristics, altimet
r
y and
bath
y
metry data
are
coded
in
environ
menta
l
file as an inp
u
t (file.env) [9-11].
Bellhop i
s
a
b
le to a
nalyze and
calcul
ate sig
nal tra
n
smi
ssi
on l
o
ss in diffe
ren
t
rang
es
and
depth
s
,
over the
chann
el. The
acou
stic to
olbox o
u
tput
s a
r
e:
so
un
d spee
d p
r
o
f
ile,
transmissio
n
loss, Impul
se
re
spo
n
se, Eigen ra
ys and time
serie
s
in
re
ceiver b
a
se
d
on
simulate
d ch
annel. Thi
s
p
aper fo
cu
se
s
on overa
ll tra
n
smi
ssi
on lo
ss achievem
e
n
t by underwater
cha
nnel
simul
a
tion throu
gh
Bellhop toolb
o
x [12].
Figure 3. Bellhop inp
u
t and
output stru
ct
ure
2. Rese
arch
Metho
d
This pa
pe
r simulates u
n
d
e
rwater
cha
n
nel
for acou
stic si
gnal p
r
opag
ation ba
sed o
n
Bellhop a
c
o
u
s
tic tool
box. The si
mulate
d ch
ann
el co
vers all
atten
uation
s
, su
rfa
c
e a
nd b
o
ttom
backscatterin
g
ba
se
d o
n
real
wave l
e
vel an
d g
eog
raphi
c
sea
bed
inform
ation,
on
Nove
mb
er
2013 in
De
sa
ru bea
ch o
n
easte
rn sho
r
e of Joho
r in
Malaysia. Be
cau
s
e of all d
a
ta is extra
c
ted
from re
al dat
aba
se
s from
node
s po
sitio
n
s, the si
mul
a
ted ch
ann
el
is very accu
rate and relia
ble.
In
this se
ctio
n,
the
pa
per explain
s
all si
mulation
pro
c
edures, th
e F
i
gure
4
sho
w
s meth
odol
og
y
cha
r
t. The si
mulation refers to Table 2 a
s
sumption h
a
s
bee
n don
e.
Figure 4. Methodol
ogy ch
a
r
t
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 : 110 – 1
1
8
114
Table 2. Un
d
e
rwater
cha
n
nel pro
p
e
r
ties
Range
28
Km
Freque
nc
y 7000
Hz
Seafloor
cover
a
ge
Rock
Sea surface
w
i
nd
speed
4 (m/s)
Source fan-b
eam
0.5° deg
Communication type
Simplex
After determi
nation
of the
ca
se
stu
d
y
area,
th
e G
e
oMapAp
p
p
r
ogra
m
i
s
utili
zed
to
colle
ct se
aflo
or ge
ographi
c data [13]. The re
se
arch
is employe
d
Aviso+ onli
n
e databa
se, t
o
obtain
su
rfa
c
e wi
nd
sp
eed
[14]. The
sal
i
nity,
tempera
t
ure
and
de
pth of
po
sition
are
mea
s
u
r
e
d
on November
2013
as
an experim
ental
measurement
. The SSP is
determined
by Mac
k
e
nzie
model, a
nd
al
l these info
rm
ation a
r
e
co
d
ed to b
e
llho
p
toolbox a
s
a
n
enviro
n
ment
al file, then th
e
TL evaluatio
n
pro
c
e
ss
run
s
to dete
r
min
e
the
be
st no
de pla
c
em
ent
. Finally, the link bu
dget i
s
c
a
lc
ulated.
2.1. En
v
i
ronmental Da
ta
The e
n
viron
m
ental
re
cords
are a
c
ce
ssi
ble f
r
om
o
n
line d
a
taba
ses
su
ch
as:
Nation
a
l
Oce
ani
c and
Atmosphe
ri
c Administ
ra
tion (NOAA) [15] or Wood
s Hole
Oce
ano
gra
p
h
i
c
Institution (WHOI) [16], Be
dford In
stitute of
Ocea
nog
raphy (BIO) [1
7]. But, this paper, utili
zed
self me
asure
m
ent of
ca
se
study o
n
No
vember
201
3
in Desaru b
each on
ea
st
ern
sh
ore
of
Joh
o
r i
n
M
a
l
a
ysia, o
n
the
South
Chin
a
Sea. Th
e lin
k
start
s
o
n
(104.87
5° E,
1.625°
N) u
n
t
il
(104.625° E,
1.625° N).
The dis
t
anc
e
between
nodes
is
near to
0.25
degree longitude, around
28 Km. Fi
gure 6
sh
ows
no
des lo
cation,
Nod
e1
duty
i
s
b
r
oa
dcast
signal to
Node
2, by me
an
s
of
s
i
mplex communication link
.
Figure 6. Nod
e
s lo
cation
2.2 Altim
e
try
and Ba
thy
m
etry
Data
To get
altim
e
try data, th
e pa
per emp
l
oy
ed Aviso
+
online
data
b
a
se. T
h
is dat
aba
se
colle
cts
surfa
c
e agitatio
n
data from Aviso
+
sa
tellite
that is based
on rada
r technolo
g
y [14].
Oce
an bathy
metry and top
ogra
phy obtai
ned by GeoM
apApp p
r
og
ra
m [13].
3. Results a
nd Discu
ssi
on
In this
se
ctio
n, the
chan
ne
l simul
a
tion
with tran
smi
ssi
on lo
ss dete
r
mination i
n
di
fferent
rang
es
and d
epths a
r
e
co
nsid
ere
d
. All simulate
d re
sults are ba
se
d on re
al dat
a that colle
ct
ed
from mention
ed so
urce
s. So, the results
are very a
c
cu
rate and
relia
ble.
3.1. Sound Speed Profile
The Ta
ble 3 i
s
sho
w
n phy
sical p
r
op
ertie
s
of case stu
d
y points, b
a
s
ed
on expe
ri
mental
measurement
on Novem
b
er. The Figure 6 repr
esents SSP graph, whi
c
h i
s
obtained
wit
h
Macken
zie fo
rmula. Ba
sed
on Figu
re 5 the maximum
sou
nd spee
d is on 30 m
e
ter de
pth wi
t
h
1540.2
67 (m/
s
) an
d minim
u
m is 154
0.1
27 on surfa
c
e
.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Und
e
rwate
r
Cha
nnel
Cha
r
acte
ri
zation t
o
De
sign
Wireless Sen
s
o
r
Network by…
(
Nim
a
Bahra
m
i
)
115
Table 3. Ca
se study physi
cal prope
rtie
s
Dept
h
(m)
Tem
p
erat
ure
(De
g
)
Salinit
y
[psu
]
SS
(m/s
)
0 28.7811
32.4825
1540.127
5
28.9116
32.559
1540.554
10 28.8389
32.6056
1540.646
15 28.8389
32.6515
1540.668
20 28.8675
32.7002
1540.871
25
28.7892
32.8968
1541.02
30 28.7362
33.1014
1541.267
35 28.6461
33.0611
1541.135
40 28.3205
32.9181
1540.313
45 28.4225
33.0024
1540.365
50 28.3103
33.0063
1540.531
Figure 7. Sound sp
eed p
r
of
ile
3.2. Sound Propaga
tion
Acco
rdi
ng to our sce
n
a
r
io the Tx is pl
aced on se
aflo
or on 50 met
e
r depth. Based o
n
Figure 8,
signal travellin
g
start from 50
meter
and
go directly until
15
Km and due to
seabed
gradi
ent, the
sign
al bent u
p
wa
rd to
surf
ace in
22
Km
approximatel
y. After this point the sig
n
a
l
travels p
a
th from b
o
ttom to su
rface, re
peat
edly. Ba
sed
on a
c
ou
stic p
r
op
agati
on graph
an
d
netwo
rk a
ppli
c
ation, the d
e
ci
sion to ch
oose
the best node positi
on for netwo
rk de
sig
ner i
s
possibl
e. Th
e
aim
s
of
pa
p
e
r
are
n
ode
placement
an
d lin
k
budg
et cal
c
ul
ation.
Then
the
gra
ph
sho
w
s 22
u
n
til 22.5 K
m
ra
nge i
s
suita
b
le fo
r buoy p
o
siti
on. With
tra
n
smi
ssi
on l
o
ss
determi
nation
the net
work
desi
gne
r is a
b
le to
find
ou
t the be
st ra
nge a
nd
re
ce
iver de
pth to
maximize
sig
nal to noise ration and mo
re power effici
ency.
Figure 8. Signal pro
pag
ation in de
sign
e
d
cha
nnel
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 : 110 – 1
1
8
116
3.3. Transmission Loss i
n
Rang
e
The Figu
re 9
represents d
i
stri
butio
n of aco
u
sti
c
sig
n
a
l transmissi
on loss, whil
e sign
al
is travelling in
chan
nel. To
focu
s on tra
n
s
missio
n loss on su
rface, the pap
er si
m
u
lates T
L
wit
h
the re
ceive
r
node
pla
c
em
ent on
2 an
d
5 meter belo
w
than
surfa
c
e, sup
p
o
s
edl
y. According
to
Figure 10, the minimum T
L
for 2 and 5
mete
r de
pths
are 22.1
5
an
d 22.25 Km range.
Figure 9. Signal TL in ch
a
nnel
Figure 10. TL
versu
s
depth
3.4. Transmission Loss i
n
Dep
t
h
After range d
e
termin
ation, the pape
r sh
ould fi
nds o
u
t the best dep
th position. So, th
e
Figure 11 sh
o
w
s the TL for
Two foun
d ra
nge
s. Fi
nally, excellent de
pth position f
o
r 22.15 Km is
2.92 mete
r
wi
th 50
dB TL
a
nd thi
s
value
for 22.2
5
Km
rang
e i
s
2.1
m with
61.5
d
B
TL. So, the
link bu
dget calcul
ation for
desi
gne
r is p
o
ssible.
Figure 11. TL
versu
s
depth
3.5. Link Bu
dget
With calculat
ed overall tra
n
smi
ssi
on lo
ss wh
ich
is
i
n
cluded sig
nal attenuation, surface
and ba
ckscat
tering lo
ss, th
e sign
al to no
ise ratio by E
quation (9) i
s
asse
ssable [2
-5].
SNR=
SL-TL-NL+
D
I
≥
DT
(9)
Whe
r
e SL i
s
transmitted si
gnal level, NL is noi
se l
e
vel, DT is
dete
c
tion threshol
d an
d
DI refe
rs to d
i
rectivity inde
x that is calculated
from i
n
tensity of a
dire
ctional
so
urce divide
t
o
intensity of
a
n
Om
ni di
re
ctional
so
urce
of equ
al p
o
wer. Th
e Fi
gure 12
sho
w
s SNR value
s
fo
r
0
to 28 Km based on (T
able
1) and
wave
noise cal
c
ulat
ion (Equ
ation
7). Whe
r
e th
e transmitted
power is 1
92
dB, Noise lev
e
l is 47.14 dB
and
DI is 4.7
d
B and dete
c
tion threshold
is 60 dB.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Und
e
rwate
r
Cha
nnel
Cha
r
acte
ri
zation t
o
De
sign
Wireless Sen
s
o
r
Network by…
(
Nim
a
Bahra
m
i
)
117
Figure 12. Lin
k
bud
get gra
p
h
Acco
rdi
ng to
Figure 12
the
re
ceive
r
no
d
e
on
su
rfa
c
e,
with the
be
st mentione
d
positio
n ha
s
30
dB received p
o
we
r more than dete
c
tion thre
shol
d.
The Fi
gu
re
7
sh
ows, th
e
sound
spee
d i
s
in
crea
sed
u
n
til 30
meters, and
betwee
n
30
to
40 m
is de
creasi
ng. T
he
Figure 8
de
m
onstrates,
if
the velo
city in
cre
a
ses e
nou
gh, an
a
c
ou
stic
path in
clined
relative to the
straig
ht ca
n
unde
rgo total
reflectio
n
. O
n
the othe
r h
and, a de
cre
a
se
in sp
eed
will
increa
se th
e
gra
z
ing
an
gle
of the
signal
. Based
on
F
i
gure
10
arou
nd 22
Km th
e
sign
al h
a
s two pe
aks
nea
r to surfa
c
e, t
he
stra
n
g
e
r
p
eak ha
s T
L
v
a
lue
s
b
e
twe
e
n
50
until
61.5
dB. The grap
h 11 determi
n
e
s the be
st node po
sition i
n
depth. As a result the be
st receiver no
de
positio
n a
s
a
buoy
stand
s on 2
2
.15 K
m
ra
nge
and
2.92 m
e
ter
depth
with th
e minimu
m o
v
erall
TL with
50
d
B
value. So, based o
n
lin
k budg
et
calcu
l
ation
the re
ceived
po
we
r has 30 dB
po
wer
more tha
n
de
tection thresh
old in optimu
m
point, but
after that, receiver face
d to
many multipath
sign
al whi
c
h
are up
pe
r tha
n
DT.
4. Conclusio
n
This p
ape
r focu
se
s o
n
u
nderwate
r
ch
annel a
nalysi
s
for a
c
o
u
sti
c
si
gnal travelling by
Bellhop
acou
stic to
olbox.
The
out
co
mes re
sult
s
of
this re
se
a
r
ch
a
r
e: T
h
e
sig
nal
cro
s
sing
betwe
en d
u
ct
s is m
o
st d
e
p
end o
n
sound
spe
ed
cha
n
g
i
ng rate, a
c
o
u
s
tic
sign
al after tra
n
smittin
g
prefe
r
to travels in lo
w sp
eed du
cts
un
less t
he environmental
co
n
d
ition is
chan
ged. Based o
n
the TL sim
u
l
a
tion, the opt
imum re
ceive
r
po
siti
on on
surfa
c
e i
s
det
ermin
ed. Accordin
g to 867
-A
modem featu
r
es; the lin
k budg
et analysis d
e
mon
s
trates the nod
e place
m
ent
desi
gn ha
s 3
0
dB
safety margin
. To have hig
her di
re
ctivity index va
lue to better
sign
al to noise rat
i
o, the desi
g
n
e
r
sho
u
ld choo
se dire
ctional t
r
an
smitter wit
h
narro
w fan beam cha
r
a
c
teristi
c
.
Ackn
o
w
l
e
dg
ements
The a
u
thors
woul
d like to t
han
k the
Uni
v
ersiti
Te
kn
ol
ogi Mal
a
ysia
(UT
M
) a
nd M
i
nistry of
Educatio
n (MoE) Mal
a
ysia fo
r finan
cial
sup
por
t
throug
h Research
Unive
r
sity Gra
n
t (GUP
05H41).
R
e
fe
re
nc
es
[1]
Heid
ema
n
n
J,
Ye W, Wills J,
S
y
ed
A, L
i
Y.
Res
earch
ch
a
llen
ges
a
nd
ap
plicati
ons
for
u
nderw
a
ter
sensor netw
o
rking
. In: Wireles
s
Commun
i
cati
ons a
nd N
e
t
w
o
r
king C
onfer
en
ce, WCNC 20
0
6
. 200
6; 1:
228
–2
35.
[2]
Lurton
X. An in
troductio
n
to u
nder
w
a
ter aco
u
stics. P
rincip
l
e
s and a
p
p
licat
ions
. Spri
ng
er.
[3]
Hod
ges RP. U
nder
w
a
ter aco
u
stics: Anal
ysi
s
, design a
nd
performa
nce of
sonar. John
W
ile
y
& So
ns.
201
1.
[4]
Bahram
i N, Khamis NH
H, Bahar
om A.
Evalu
a
tion of u
nderw
a
ter aco
u
stical int
e
rmittent amb
i
e
n
t
nois
e
. In: Signal Proc
essing
& Its Applic
ations (CSPA)
, 2015 IEEE 11th I
n
ternational Colloquium
on
IEEE. 2015: 11
-14.
[5]
Domin
go MC.
Overvie
w
of
chann
el mo
d
e
ls for und
er
w
a
t
e
r
w
i
re
less
communic
a
tio
n
net
w
o
rks.
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 : 110 – 1
1
8
118
Physica
l Co
mmu
n
ic
ation
. 2
0
08; 1(3): 16
3-1
82.
[6] Garcia
JE.
A
d
hoc
positi
o
nin
g
for se
ns
ors in
und
erw
a
ter aco
u
stic
netw
o
rks
. In: OCEANS’04.
MTTS/IEEE
TECHNO-OCEA
N’04. 2
0
0
4
; 4:
233
8-23
40.
[7]
Alba
lad
e
j
o
C, Soto F
,
T
o
rres R, S´anch
e
z P,
L´opez
JA. A lo
w
-
c
o
s
t
sensor bu
o
y
s
y
stem for
monitori
ng sh
al
lo
w
mari
ne e
n
v
i
ronme
n
ts.
Sensors
. 2012;1
2(
7): 9613-
96
34.
[8]
Benthos T
.
T
e
l
e
d
y
n
e
be
nthos
unders
ea s
y
stems and e
q
u
i
p
m
ent. 2015.
[9] Porter
MB.
T
he BELLHOP
manu
al a
nd us
er
s guid
e
: PRELI
M
INARY DRA
F
T
. Heat, Light, and Sou
nd
Rese
arch, Inc. La Jol
l
a, CA, USA. 2011.
[10]
Rodri
g
u
e
z OC. General d
e
scri
p
tion of
the BE
LLHOP ra
y tra
c
ing pr
ogram.
200
8.
[11]
Z
e
iger V, B
a
d
r
i-
Hoe
her S,
Hoe
her PA.
M
ode
l-Base
d Po
sitioni
ng
. Proc
Oceano
lo
g
y
Internati
o
n
a
l
(OI12). Lond
on
, UK. 2012.
[12]
Bahram
i N, Kh
amis N
HH, Ba
harom A. Stu
d
y
of Un
der
w
a
t
e
r Ch
an
nel
Estimatio
n
Bas
ed
on D
i
fferen
t
Nod
e
Placem
e
n
t in Shal
lo
w
W
a
ter.
Sensor
s Journa
l.
201
5; (99): 1.
[13]
Ha
xb
y W
,
Melk
oni
an A, C
o
p
l
a
n
J, Ch
an S,
R
y
a
n
W
.
GeoMa
p
App
free
w
a
re
soft
w
a
re, v. 2.
3. Lamo
n
t-
Doh
e
rt
y
Earth
Observator
y. Palisa
des. 2
010.
[14]
Aviso
satel
lite altimetr
y
data. http://
w
w
w
.
av
is
o.altimetr
y
.fr.
[15]
Natio
nal Ocea
nic an
d atmosp
heric
ad
mi
ni
stra
ti
o
n
.
h
ttp://
w
w
w
.
noaa.gov.
[16]
W
oods ho
le oc
ean
ogra
p
h
i
c in
stitution. http://www
.
w
h
o
i
.e
du.
[17] Bedford
i
n
stit
ut
e of ocea
nogr
a
p
h
y
. http://
w
w
w
.
b
i
o.gc.ca.
Evaluation Warning : The document was created with Spire.PDF for Python.