Inter
national
J
our
nal
of
Electrical
and
Computer
Engineering
(IJECE)
V
ol.
4,
No.
3,
June
2014,
pp.
314
–
321
ISSN:
2088-8708
314
A
Unified
Call-to-Pray
er
Framew
ork
in
Muslim
W
orld
Naeem
Al-Oudat
and
Abdel
Ilah
Alshbatat
Dept.
of
Electrical
and
Computer
Engineering,
T
afila
T
echnical
Uni
v
ersity
,
Jordan
Article
Inf
o
Article
history:
Recei
v
ed
Feb
8,
2014
Re
vised
Apr
5,
2014
Accepted
Apr
22,
2014
K
eyw
ord:
Unified
call-to-prayer
Distrib
uted
mosques
Sound
pressure
le
v
el
Logical
Fermat
point
Local
amplifier
-g
ain
setting
algorithm
ABSTRA
CT
In
man
y
Muslim
countries
there
are
man
y
sounds
that
are
fired
at
nearly
the
same
time
via
loudspeak
ers.
This
sound
is
a
call-to-pr
ayer
(Azan).
Azan
is
fired
from
the
so-called
mosques
in
man
y
countries
where,
these
s
mosques
are
still
using
its
o
wn
tim
ing
to
trigger
such
call
and
its
o
wn
amplifier
g
ain
re
g
ardless
of
other
mosques
in
the
re
gion.
This
results
in
an
out
of
sync
call-to-prayer
firing
and
a
v
ery
noisy
and
distracting
mix
of
sounds
in
man
y
places
at
the
same
re
gion.
In
this
paper
,
a
unified
call-to-prayer
frame
w
ork
is
proposed
that
sheds
light
on
these
issues
and
gi
v
es
solution
directions
for
the
abo
v
e
mentioned
issues
in
the
currently
used
systems.
Copyright
c
2014
Institute
of
Advanced
Engineering
and
Science
.
All
rights
r
eserved.
Corresponding
A
uthor:
Naeem
Al-Oudat
Dep.
of
Electrical
and
Computer
Engineering,
T
afila
T
echnical
Uni
v
ersity
P
.O.Box
179,
T
afila,
66110,
Jordan.
+962777526844
naeemodat@ttu.edu.jo
1.
INTR
ODUCTION
In
man
y
Muslim
countries,
there
is
a
v
ery
well-kno
wn
sound
that
is
fired
fi
v
e
times
a
day
.
This
sound
is
the
call-to-prayer
or
Azan,
which
is
a
timely
fired
call
for
the
fi
v
e
prayers
namely
F
ajir
,
Duher
,
Asr
,
Magrib
and
Isha
[1].
This
sound
is
fired
from
each
Muslim
mosque
(Masjid).
What
determines
ho
w
much
sound
one
can
hear
(Azan
soundscape)
in
an
y
place
is
ho
w
f
ar
one
is
from
Masjids.
Masjids
positions
are
determined
mainly
based
on
the
con
v
enience
of
access
for
the
habitants
in
an
y
re
gion.
A
soundscape
is
defined
as
the
o
v
erall
sonic
en
vironment
in
a
re
gi
on
[2].
In
an
y
soundscape
in
Muslim
countries,
Azan
is
fired
from
each
mosque
using
loudspeak
ers
that
are
directed
in
man
y
directions
on
top
of
the
minarets
of
the
mosques.
The
number
of
mosques
in
a
square
kilometer
is
roughly
not
less
than
four
.
Each
one
of
the
mosques
fire
the
same
sound
at
the
same
time,
for
that
reason
an
y
clock
that
is
out
of
sync
in
an
y
of
the
mosques
causes
a
noisy
and
not
so
pleasant
sound
at
prayer
times.
Further
,
each
mosque
usually
sets
its
amplifiers
to
the
highest
g
ain
independently
of
other
mosques
which
results
in
a
v
ery
loud
Azan
in
some
places.
These
tw
o
issues
(out
of
sync.
clocks
and
highest
amplifiers
g
ains)
can
be
solv
ed
or
at
least
mitig
ated
if
a
proper
setting
of
a
unified
call-to-prayers
is
applied,
where
g
ains
of
mosque’
s
sound-amplifiers
are
set
appropriately
and
clocks
in
the
mosques
are
synchronized.
In
this
paper
,
a
unified
call-to-prayers
frame
w
ork
is
proposed
that
pro
vides
a
solution
to
the
abo
v
e
tw
o
issues.
The
set
of
mosques
in
an
y
re
gion
is
a
dist
rib
uted
real-time
system
that
cooperati
v
ely
achie
v
es
a
common
goal;
namely
announcing
the
prayer
times
to
the
residents
in
that
re
gion.
Ho
we
v
er
a
unified
call-to-prayer
i
s
not
a
ne
w
idea,
it
has
been
applied
in
man
y
cities
in
Muslim
w
orld;
Amman
in
Jordan
is
using
an
FM
transmitter
[3],
Emirates
is
using
satellite
broadcasting
[4,
3],
and
Cairo
in
Egypt
where
the
y
are
using
a
common
transmitter
[5,
3].
The
unified-call-to-prayer
presented
in
this
paper
is
a
no
v
el
solution
that
considers
setting
amplifiers
g
ains
not
only
firing
a
unified
Azan.
The
rest
of
this
paper
is
or
g
anized
as
follo
ws.
A
Unified
Call-to-Prayer
Frame
w
ork
is
presented
and
a
problem
is
formulated
in
Section
2..
In
Section
3.
neighbor
disco
v
ery
paradigm
is
discussed
as
an
important
step
in
proposed
frame
w
ork.
A
heuristi
c
algorithm
to
set
proper
g
ains
for
mosques
in
an
y
re
gion
is
presented
and
discussed
in
Section
4..
Section
5.
pro
vides
a
proof
of
concept
via
e
xperimental
study
.
Finally
,
Section
6.
concludes
the
paper
and
gi
v
e
a
future
research
directions.
Evaluation Warning : The document was created with Spire.PDF for Python.
IJECE
ISSN:
2088-8708
315
2.
UNIFIED
CALL-T
O-PRA
YER
FRAMEW
ORK
In
this
section,
the
frame
w
ork
of
unified
call-to-prayer
is
gi
v
en
and
a
proposed
solution
is
discussed.
a
b
c
c
A
m
1
m
2
m
3
m
4
c
1
c
2
c
3
c
4
(a)
S
ystem
Model
(b)
Modified
Syst
em
Figure
1.
A
system
of
4
mosques
with
their
(a)
random
co
v
ering
areas
and
(b)
a
better
settings
of
their
co
v
ering
areas
2.1.
Ov
er
view
of
the
Framew
ork
As
sho
wn
in
Fig.
1(a),
the
set
of
mosques
is
a
group
of
distrib
uted
sites
each
has
its
o
wn
range
of
sound
broadcasting
to
its
neighbors.
The
dotted
circles
represent
this
range
which
is
the
border
of
desired
Sound
Pressure
Le
v
els.
There
are
se
v
eral
o
v
erlapped
re
gions
where
the
sound
pressure
le
v
el
is
the
sum
of
more
than
one
source.
These
re
gions
are
sho
wn
in
the
figure,
A,
B
and
C.
In
these
re
gions
the
sound
is
pretty
much
a
mix
of
dif
ferent
sound
sources
which
gi
v
e
unpleasant
feeling
at
these
re
gions
which
is
proportional
to
the
number
of
sources
heard.
Fig.
1(a)
sho
ws
only
four
mosques
which
is
only
a
simplified
sample
of
the
actual
distrib
ution
of
mosques.
These
unpleasant
feelings
in
these
re
gions
result
from
an
y
or
all
of
the
follo
wings:
(1)
time
dif
ferences
in
Azan
firing,
(2)
se
v
eral
persons
who
fire
the
Azan.
These
tw
o
issues
can
be
solv
ed
by
calling
to
prayers
at
the
same
time
and
by
the
same
person.
Further
the
range
circles
can
be
recalculated
to
reduce
the
o
v
erlapping
areas.
More
acceptable
ranges
are
depicted
in
Fig.
1(b).
2.2.
Framew
ork
Details
and
Pr
oblem
Statement
T
o
achi
e
v
e
the
required
goal
of
reducing
the
area
of
the
o
v
erlapped
areas,
this
problem
can
be
look
ed
at
as
an
optimization
problem.
The
goal
is
to
minimize
the
area
of
o
v
erlapped
re
gions
while
meeting
the
conditions
of
satisfying
the
residents
in
these
re
gions.
One
of
the
measures
used
to
decide
the
satisf
actions
of
the
residents
in
these
re
gions
is
the
sound
pressure
le
v
el
(SPL).
SPL
at
an
y
distance
is
proportional
to
the
in
v
erse
of
the
dist
ance
from
the
source.
Hence
if
SPL
p
1
at
distance
r
1
is
kno
wn
then
SPL
p
2
at
r
2
[6]
is
gi
v
en
for
open
en
vironment
in
Eq.
1.
p
2
=
p
1
20
l
og
10
r
2
r
1
(1)
Se
v
eral
remarks
can
be
dra
wn
when
e
xamining
the
problem
at
hand:
R1.
There
are
n
nodes
in
a
desired
area
A
,
where
each
node
(minaret
of
the
mosque)
is
equipped
with
a
transcei
v
er
,
and
an
omnidirectional
loudspeak
er
(se
v
eral
similar
loudspeak
ers
that
co
v
er
360
around
the
minaret).
R2.
The
link
between
an
y
tw
o
nodes
can
be
af
fected
with
se
v
eral
parameters;
T
emperature,
W
ind
V
elocity
,
dif
fraction
properties
and
Humidity
.
The
borders
of
the
SPL
co
v
erage
areas
of
the
tw
o
nodes
shrink
or
e
xpand
based
on
these
parameters.
Sound
propag
ation
is
af
fected
mainly
by
the
ground
surf
ace
temp,
where
direction
of
the
A
Unified
Call-to-Pr
ayer
F
r
ame
work
in
Muslim
W
orld
(Naeem
Al-Oudat)
Evaluation Warning : The document was created with Spire.PDF for Python.
316
ISSN:
2088-8708
sound
propag
ation
bends
more
to
w
ard
the
colder
en
vironment.
As
of
wind
v
elocit
y
,
the
sound
tends
to
bend
to
w
ard
the
lo
west
speed
block
of
air
when
the
sound
propag
ates
in
the
wind
direction
and
bends
to
w
ard
the
highest
speed
block
of
air
when
the
sound
p
r
op
a
g
at
ion
is
not
in
the
same
direction
of
the
wind.
Sound
t
ends
to
reflect
from
hard
obstacles
when
obstacle
size
is
lar
ge
compared
to
sound
w
a
v
elength
and
tends
to
bend
around
when
obstacle
size
is
small
as
compared
to
sound
w
a
v
elength.
As
of
humidity
parameter
,
dry
air
attenuates
sound
more
than
humid
air
does.
Further
the
required
SPL
at
an
y
place
is
less
when
the
background
noise
in
the
re
gion
is
lo
w
.
R3.
When
looking
at
the
area
as
a
whole
there
will
be
re
gions
that
are
co
v
ered
by
more
than
one
source.
R4.
The
lack
of
time
synchronization
between
mosques
results
in
time
lapses
between
Azans
heard
in
o
v
erlapped
re
gions.
Azan
is
composed
from
dedicated
w
ords;
consequently
,
the
audiences
in
the
o
v
erlapped
re
gions
hear
dif
ferent
w
ords
from
dif
ferent
sources.
R1
imposes
the
constraint
of
co
v
ering
the
desired
area
A
.
R3
gi
v
es
the
objecti
v
e
of
our
solution
for
the
problem,
which
is
area
minimization
of
the
o
v
erlapped
re
gions.
R3
and
R4
impose
the
need
of
a
unified
call
to
prayer
and
time
synchronization
between
mosques.
While
R2
sho
ws
the
problem
dynamics,
where
the
radii
of
the
co
v
erage
circles
are
influenced
by
the
mentioned
parameters
e
xcept
wind
v
elocity
,
where
there
is
no
sense
in
considering
the
shape
of
the
co
v
erage
area
as
a
circle
an
y
more.
F
or
the
case
in
Fig.
1(a),
the
problem
can
be
e
xpressed
as
a
linear
program
(LP):
Let
OA
denotes
the
sum
of
the
o
v
erlapped
areas.
Then
OA
=
a
+
b
+
c
,
And
hence
LP
can
be
written
as:
Algorithm
1
LP
formulation
of
UCtP
problem
minimize
OA
such
that:
OA
m
1
m
2
m
3
m
4
+
c
1
+
c
2
+
c
3
+
c
4
=
A
In
light
of
the
abo
v
e
LP
,
the
problem
at
hand
requires
an
enumeration
of
all
possible
o
v
erlapping
between
the
co
v
erage
circles
then
the
optimal
solution
is
the
one
that
gi
v
es
a
smaller
o
v
erlapped
area.
This
means
that
the
comple
xity
gro
ws
e
xponentially
as
the
number
of
nodes
increases.
Since
an
optimal
solution
needs
a
lot
of
space
and
time
to
minimize
the
o
v
erlapped
areas,
which
are
indeed
not
feasible
due
to
the
constraints
and
dynamics
of
this
problem,
a
f
aster
solution
is
needed.
Therefore
a
heuristic
algorithm
is
propos
ed,
in
this
paper
,
to
pro
vide
a
f
ast
suboptimal
solution
to
the
problem
at
hand.
The
algorithm
is
based
on
a
local
decision
that
is
tak
en
in
each
node
based
on
the
state
of
the
neighboring
nodes.
The
algorithm
starts
by
e
xchanging
state
describing
messages
between
the
nodes,
and
then
each
node
will
decide
what
g
ain
le
v
el
will
be
used
for
its
amplifier
.
Locally
Gain
setting
Based
on
Neighboring
States
Algorithm
(LGBNSA)
is
the
name
of
our
algorithm.
In
the
follo
wing
section
we
pro
vide
the
details
of
this
algorithm.
3.
NEIGHBOR
DISCO
VER
Y
F
or
an
y
node
to
disco
v
er
its
neighbors,
disco
v
ery
beacon
with
a
highest
g
ain
will
be
broadcasted
periodically
fi
v
e
times
a
day
ahead
of
Azan
times
by
delta
t.
T
o
satisfy
the
pre
vious
goal,
we
assum
ed
that
each
node
has
its
o
wn
GPS
recei
v
er
for
time
synchronization
between
nodes,
and
its
priority
to
start
beacon
broadcasting
is
based
on
its
ID,
as
sho
wn
in
Fig.
2.
Each
beacon
carries
the
ID
of
the
broadcasting
node.
F
ollo
wing
this
procedure
each
node
is
capable
of
recognizing
all
its
neighbors.
F
ailure
to
recei
v
e
beacons
from
other
nodes
is
considered
as
an
indication
that
the
node
has
no
neighbors.
Once
this
happen,
this
node
should
switch
to
its
highest
g
ain
during
scheduled
Azan
firing.
Neighbor
disco
v
ery
algorithm
can
be
classified
to
tw
o
types
based
on
ho
w
frequent
it
is
called;
short
and
long
term
disco
v
ery
.
Long
term
neighbor
disco
v
ery
can
be
periodically
called
for
a
period
of
24
hours,
while
short
term
disco
v
ery
is
called
fi
v
e
times
a
day
.
An
y
node
that
is
heard
by
node
i
when
it
transmits
its
beacon
is
added
to
the
neighbor
table.
T
o
calculate
the
distance
of
that
node
from
node
i
a
sequence
of
transmissions
is
initiated
starting
by
the
highest
po
wer
then
starting
decreasing
that
po
wer
until
it
is
not
heard.
At
that
instant
the
transmission
g
ain
is
translated
to
a
logical
distance
between
the
tw
o
nodes.
This
distance
is
used
in
calculating
the
logical
coordinated
of
the
mo
ving
node.
F
or
each
of
the
border
nodes
(that
does
not
ha
v
e
an
y
node
in
its
co
v
erage
circle
from
the
border
direction)
tw
o
dummy
nodes
(radius
of
the
co
v
erage
circle
of
a
dummy
node
is
zero)
are
added
to
neighbors
table.
IJECE
V
ol.
4,
No.
3,
June
2014:
314
–
321
Evaluation Warning : The document was created with Spire.PDF for Python.
IJECE
ISSN:
2088-8708
317
F
3
F
2
F
1
I
D
2
0
I
D
1
0
I
D
1
Figure
2.
T
iming
diagram
for
setup
and
Azan
calling
periods
4.
LOCALL
Y
GAIN
SETTING
B
ASED
ON
NEIGHBORING
ST
A
TES
ALGORITHM
(LGBNSA)
T
o
handle
the
problem
using
the
proposed
algorithm,
tw
o
cases
arise:
1.
Number
of
neighbors
for
each
node
is
fix
ed.
Based
on
the
result
obtained
from
long
term
disco
v
ery
type.
2.
Number
of
neighbors
is
not
fix
ed
which
is
v
arying
according
to
short
term
disco
v
ery
type.
Each
node
has
what
is
called
a
vicinity
matrix
that
includes
all
nodes
of
distance
2R
or
less
a
w
ay
.
Based
on
triangu-
lation
which
means
all
possible
triangle
s
are
formed
from
the
node
and
its
neighbors.
After
forming
these
triangles
a
Fermat
point
is
calculated
that
gi
v
es
radii
of
the
co
v
erage
circles
for
n
odes
on
v
erte
x
es
of
the
tria
n
gl
e.
These
radii
guarantee
the
co
v
erage
of
the
triangle
formed
from
these
nodes.
K
eep
doing
this
will
e
v
entually
co
v
er
the
whole
re
gion
while
k
eeping
the
o
v
erlapped
areas
to
the
minimum
possible.
Note
that
R
is
not
fix
ed
and
it
does
not
represent
a
ph
ysical
dist
ance
on
earth,
thus
the
other
tw
o
v
erte
x
es
other
than
the
one
at
focus
are
dynamically
mo
ving
a
w
ay
or
close
to
rel
ati
v
e
to
the
one
at
focus.
This
phenomenon
is
a
logical
consequence
as
a
result
of
the
abo
v
e
mentioned
parameters.
Based
on
this
the
node
at
focus
sends
beacons
to
ot
her
nodes
in
its
adjacenc
y
matrix
to
determine
ho
w
f
ar
a
w
ay
the
y
are.
Then
a
ne
w
logical
coordinates
for
these
nodes
are
formulated
and
implemented
in
calculation
of
Fermat
point.
Fig.
3
clarifies
this
idea.
Figure
3.
Logical
Coordinates
of
nodes
j
and
k
as
seen
from
node
i
The
ne
w
point
of
the
virtually
mo
ving
node
can
be
simply
calculated
from
the
line
equation,
assuming
the
mo
v
ement
is
only
happen
along
the
ph
ysical
line
between
the
nodes.
While
it
is
possible
for
the
node
to
sho
w
an
illusion
of
c
o
or
d
i
nates
that
are
not
on
the
same
line,
we
did
not
consider
it
in
this
paper
and
we
reserv
e
that
for
a
future
w
ork.
Let
r
denotes
the
distance
between
the
tw
o
nodes
n
i
at
point
(
x
i
;
y
i
)
and
n
k
at
point
(
x
k
;
y
k
)
.
Let
d
denotes
the
logical
distance
between
the
ne
w
location
of
node
k
at
point
(
x
k
;
y
k
)
and
node
i
.
Then
(
x
k
;
y
k
)
can
be
calculated
as
in
Eq.
2.
y
k
=
y
i
+
d
y
k
y
i
r
;
x
k
=
y
k
y
i
+
y
k
y
i
x
k
x
i
x
1
x
k
x
i
y
k
y
i
(2)
Similarly
,
Eq.
2
is
used
t
o
find
the
logi
cal
coordinates
of
node
j
,
(
x
j
;
y
j
)
.
T
o
calculate
the
radius
for
each
of
the
v
erte
x
es
(
(
x
1
;
y
1
)
,
(
x
2
;
y
2
)
and
(
x
3
;
y
3
)
)
under
consideration,
we
need
to
find
the
Fermat
point
which
is
done
according
to
the
w
ork
in
[7].
Thei
r
equations
(3)
for
finding
Fermat
point
(
x
F
;
y
F
)
is
repeated
here
for
completeness
of
the
w
ork.
x
F
=
K
1
K
2
K
3
2
S
p
3
d
x
1
K
1
+
x
2
K
2
+
x
3
K
3
;
y
F
=
K
1
K
2
K
3
2
S
p
3
d
y
1
K
1
+
y
2
K
2
+
y
3
K
3
(3)
where,
S
=
j
x
1
y
2
+
x
2
y
3
+
x
3
y
1
x
1
y
3
x
3
y
2
x
2
y
1
j
A
Unified
Call-to-Pr
ayer
F
r
ame
work
in
Muslim
W
orld
(Naeem
Al-Oudat)
Evaluation Warning : The document was created with Spire.PDF for Python.
318
ISSN:
2088-8708
r
j
l
=
q
(
x
j
x
l
)
2
+
(
y
j
y
l
)
2
for
f
j
;
l
g
2
f
1
;
2
;
3
g
K
1
=
p
3
2
(
r
2
12
+
r
2
13
r
2
23
)
+
S
K
2
=
p
3
2
(
r
2
23
+
r
2
12
r
2
13
)
+
S
K
3
=
p
3
2
(
r
2
13
+
r
2
23
r
2
12
)
+
S
d
=
1
p
3
(
K
1
+
K
2
+
K
3
)
Then
for
node
i
the
distance
from
Fermat
point,
r
iF
,
is
calculated
as:
r
iF
=
p
(
x
i
x
F
)
2
+
(
y
i
y
F
)
2
(4)
Assuming
that
r
1
=
1
,
p
2
=
p
F
,
and
using
the
result
of
abo
v
e
equation
(4),
p
i
=
p
1
can
be
found
from
Eq.
1.
Algorithm
2
Locally
Gain
setting
Based
on
Neighboring
States
Algorithm
(LGBNSA)
Input:
node
i
and
its
neighbors
Output:
node
i
with
adjusted
radius
R
i
max
f
0,
max.
distance
from
dummy
nodes
if
an
y
g
f
or
all
nodes
in
neighbor
table
of
i
do
Construct
a
triangle
Compute
the
distance
from
i
to
Fermat
point
of
the
triangle
if
R
i
is
less
than
the
computed
distance
then
R
i
computed
distance
end
if
end
f
or
In
this
paper
,
where
the
goal
is
to
shed
light
on
the
e
xisting
problem,
we
made
some
assumptions:
Neighbors
are
fix
ed
and
kno
wn
to
an
y
of
the
nodes.
The
re
gion
that
contains
the
mosques
is
a
flat
area,
i.e.,
there
is
no
mountains
or
v
alle
ys
in
the
re
gion.
Mosques
sound
le
v
els
are
the
same
on
all
directions
and
form
a
circle
centered
at
the
specified
mosque.
5.
EXPERIMENT
AL
STUDIES
In
this
section,
we
present
mosques
transmission
radii
for
actual
mosques
on
earth.
Then
we
apply
the
proposed
algorithm
on
that
distrib
ution
of
the
mosques
and
compare
the
results.
T
o
our
kno
wledge,
no
researchers
ha
v
e
been
discussed
this
problem
before.
Fig.
4(a)
sho
ws
the
distrib
ution
of
mosques
in
Ainalbaida
re
gion
that
lies
in
the
southern
part
of
Jordan.
The
residential
area
of
this
village
is
1800
m
eters
wide
by
3100
meters
long.
Eighteen
mosques
are
distrib
uted
almost
randomly
in
this
area,
i.e.,
not
according
to
the
sound
le
v
els
distrib
ution
around
the
mosque
rather
than
the
con
v
enience
of
access
for
the
residents.
T
able
1
reports
the
con
v
entional
coordinates
of
the
v
erte
x
es
of
the
mosques.
Note
that
the
distances
between
mosques
in
this
coordinate
system
is
a
representati
v
e
system
of
the
actual
coordinate
system.
The
mosques
are
numbered
in
order
as
Fig.
4(b)
sho
ws;
the
origin
of
coordinates
is
placed
in
the
bottom-left
corner
of
the
figure.
Please,
bear
in
mind
that
the
positions
cannot
be
sw
apped
to
an
y
ne
w
location
and
the
coordinates
are
estimated
with
respect
to
the
(x,y)
reference
frame.
Fig.
4(b)
is
generated
by
performing
a
2D
Delaunay
triangulation
[9]
on
a
set
of
eighteen
mosques.
The
resulting
diagram
consists
of
twenty
nine
triangles
distrib
uted
as
sho
wn
in
Fig.
4(b).
Delaunay
triangulation
has
the
property
that
the
circumcircle
of
an
y
triangle
in
the
triangulation
contains
no
mosque
in
its
interior
.
The
centers
of
the
circumcircles
are
then
connected
to
each
other
in
which
the
y
all
produce
a
V
oronoi
diagram
[10]
which
is
a
dual
of
the
Delaunay
triangulation.
T
o
apply
LGBNSA,
the
neighbors
of
the
mosques
are
as
sho
wn
in
T
able
2.
Note
that,
in
this
table
an
y
neighbor
that
is
f
ar
a
w
ay
by
more
than
2R
(R
considered
here
as
200)
is
dropped
from
the
list
of
neighbors
and
hence
an
y
triangle
formed
with
this
neighbor
is
also
dropped
from
triangles
used
in
calculating
Fermat
points.
F
or
e
xample
IJECE
V
ol.
4,
No.
3,
June
2014:
314
–
321
Evaluation Warning : The document was created with Spire.PDF for Python.
IJECE
ISSN:
2088-8708
319
(a)
Di
strib
ution
of
mosques
in
Ainalbaida
re
gion
[8]
1
2
3
4
5
6
7
8
9
1
0
1
1
1
2
1
3
1
4
1
5
1
6
1
7
1
8
(b)
x,y
Coordinates
and
Delaunay
triangulation
of
Ainalbeada
mosques
[8]
Figure
4.
Mosques
in
Ainalbaida
re
gion
(a)
Distrib
ution
and
(b)
triangulation
the
triangle
formed
by
mosques
1,
3,
and
6
in
Fig.
4(b)
is
not
considered
in
an
y
further
calculation
because
distances
between
1
and
6;
and
3
and
6
are
greater
than
200.
Note
that
for
an
y
mosque
where
its
neighbors
list
is
empty
,
its
radius
is
the
highest
(200
in
our
e
xample).
Fig.
5(a)
sho
ws
the
circles
that
are
dra
wn
around
each
mosque
and
represe
n
t
its
co
v
ering
re
gion.
Fig.
5(b)
sho
ws
the
current
co
v
ering
circles
of
the
mosques,
where
each
mosque
uses
its
max
am
p
l
ifier
g
ain
(equi
v
alent
to
200
in
our
case).
W
e
sho
wed
only
some
of
them
in
order
to
k
eep
the
figure
as
clear
as
possible
although
each
of
the
mosques
has
the
same
circle
centered
on
it.
Using
the
information
in
T
able
2,
the
sum
of
radii
squares
is
around
372333.
While
for
the
radii
in
Fig.
5(b)
the
sum
of
their
squares
is
720000.
The
number
for
the
calculate
d
radii
is
small
compared
to
the
current
system.
This
number
reflects
the
sum
of
sound
le
v
els
in
the
re
gion
co
v
ered
by
the
gi
v
en
mosques,
i.e.,
proportional
to
the
areas
of
the
co
v
ering
circles.
That
means
the
o
v
erlapped
areas
are
lar
ger
than
in
current
system
from
that
calculated
by
the
proposed
algorithm.
This
number
is
also
proportional
to
the
po
wer
consumption
by
the
mosques
amplifiers.
6.
CONCLUSION
AND
FUTURE
W
ORK
Call-to-prayer
in
Muslim
w
orld
is
an
important
timely
fired
sound.
This
sound
is
triggered
fi
v
e
times
a
day
in
a
noisy
and
distracting
w
ay
if
not
correctly
synchronized
and
unified.
In
this
paper
a
no
v
el
frame
w
ork
of
unified
call-to-prayer
w
as
proposed.
The
problem
w
as
formulated,
analyzed,
and
a
general
solution
w
as
proposed
that
is
dependent
on
setting
a
suitable
g
ain
for
the
dif
ferent
mosques
amplifiers.
This
w
ork
can
be
implemented
to
serv
e
man
y
other
sound
generating
f
acilities
lik
e
public
school,
churches,
disaster
announcement.
As
a
future
w
ork,
the
transmitting
of
unified
call-to-prayer
will
be
studied
to
achie
v
e
the
requirements
of
con
v
enience
and
co
v
ering
habitants’
areas.
A
suitable
terrain
and
local
en
vironment
models
can
be
applied
to
get
more
accurate
A
Unified
Call-to-Pr
ayer
F
r
ame
work
in
Muslim
W
orld
(Naeem
Al-Oudat)
Evaluation Warning : The document was created with Spire.PDF for Python.
320
ISSN:
2088-8708
T
able
1.
XY
coordinates
of
Ainalbeada
mosques
Mosque
X
Y
1
55
71
2
59
81
3
132
103
4
135
159
5
277
180
6
409
102
7
385
250
8
314
344
9
207
347
10
375
414
11
107
505
12
161
497
13
231
517
14
324
528
15
210
586
16
337
660
17
161
695
18
92
713
T
able
2.
Neighbors
and
calculated
radii
of
Ainalbeada
mosques
M
i
Set
of
neighbors
(
j
;
k
)
Radius
1
(2,3)
11.5
2
(1,3),
(3,4)
72.7
3
(1,2),
(2,4),
(4,5)
77
4
(2,3),
(3,5),
(5,9)
95
5
(3,4),
(4,9),
(9,8),
(8,7),
(7,6)
135.5
6
(5,7)
100
7
(5,6),
(5,8),
(8,10)
107
8
(5,7),
(7,10),
(10,13),
(13,9),
(9,5)
91.5
9
(4,5),
(5,8),
(8,13),
(13,12),
(12,11)
155
10
(7,8),
(8,13),
(13,14)
117.5
11
(9,12),
(12,15),
(15,17),
(17,18)
179.3
12
(9,11),
(11,15),
(15,13),
(13,9)
59
13
(9,8),
(8,10),
(10,14),
(14,15),
(15,12),
(12,9)
152
14
(10,13),
(13,15),
(15,16)
84.5
15
(11,12),
(12,13),
(13,14),
(14,16),
(16,17),
(17,11)
98
16
(14,15),
(15,17)
118
17
(16,15),
(15,11),
(11,18)
108
18
(11,17)
49
distances
between
mosques
in
the
neighborhood.
Further
,
an
y
loudspeak
er
on
the
minaret
of
an
y
mosque
can
be
tak
en
into
account
which
produces
a
non-circular
co
v
ering
re
gion.
REFERENCES
[1]
T
.
S.
Lee,
“T
echnology
and
the
production
of
islamic
space:
the
call
to
prayer
in
sing
apore,
”
Ethnomusicolo
gy
,
v
ol.
43,
no.
1,
pp.
86–100,
1999.
IJECE
V
ol.
4,
No.
3,
June
2014:
314
–
321
Evaluation Warning : The document was created with Spire.PDF for Python.
IJECE
ISSN:
2088-8708
321
1
2
3
4
5
6
7
8
9
1
0
1
1
1
2
1
3
1
4
1
5
1
6
1
7
1
8
(a)
Calculated
radii
of
the
mosques
based
on
the
pro-
posed
algorithm
[8]
1
2
3
4
5
6
7
8
9
1
0
1
1
1
2
1
3
1
4
1
5
1
6
1
7
1
8
(b)
Co
v
erage
circles
of
the
mosques
in
Ainal-
baida
[8]
Figure
5.
Mosques
and
their
co
v
ering
re
gins.
(a)
Calculated
using
the
proposedlgorithm
and
(b)
actual
co
v
ering
circles
[2]
J.
D.
Porteous
and
J.
F
.
Mastin,
“Soundscape.
”
J
ournal
of
Ar
c
hitectur
al
and
Planning
Resear
c
h
,
1985.
[3]
“http://www
.uaeinteract.com/docs/ministry
announces
unified
call
system/13899.htm,
”
last
accessed:
Decem-
ber
,
7,
2013.
[4]
“http://www
.a
wqaf.ae/service.aspx?lang=en§ionid=19&refid=1348,
”
last
accessed:
December
,
7,
2013.
[5]
“http://www
.e
gyptindependent.com/ne
ws/one-v
oice-man
y-mosques-call-prayer
-unified-muezzins-defy-
go
v
ernment-orders,
”
last
accessed:
December
,
7,
2013.
[6]
D.
Bies
and
C.
Hansen,
Engineering
noise
contr
ol:
theory
and
pr
actice
.
T
aylor
&
Francis,
2009.
[7]
A.
Y
.
Uteshe
v
,
“
Analytical
solution
for
the
generalized
fermat-torricelli
problem,
”
arXiv
pr
eprint
arXiv:1208.3324
,
2012.
[8]
“T
afila,
jo.
(03
feb
.
2014).
google
maps.
google.
retrie
v
ed
from
https://maps.google.com/maps/ms?msid=
203417662707219167682.0004e6f
a615a01ba6136a&msa
=0&ll=30.784244,35.598621&spn=0.049699,0.090895.
”
[9]
D.-T
.
Lee
and
B.
J.
Schachter
,
“T
w
o
algorithms
for
constructing
a
delaunay
triangulation,
”
International
J
ournal
of
Computer
&
Information
Sciences
,
v
ol.
9,
no.
3,
pp.
219–242,
1980.
[10]
F
.
Aurenhammer
,
“V
oronoi
diagramsa
surv
e
y
of
a
fundamental
geometric
data
structure,
”
A
CM
Computing
Sur
-
ve
ys
(CSUR)
,
v
ol.
23,
no.
3,
pp.
345–405,
1991.
A
Unified
Call-to-Pr
ayer
F
r
ame
work
in
Muslim
W
orld
(Naeem
Al-Oudat)
Evaluation Warning : The document was created with Spire.PDF for Python.