I
nd
o
ne
s
ia
n J
o
urna
l o
f
E
lect
rica
l En
g
ineering
a
nd
Co
m
pu
t
er
Science
Vo
l.
24
,
No
.
2
,
N
o
v
em
b
e
r
2
0
2
1
,
p
p
.
8
1
5
~
8
2
4
I
SS
N:
2
5
0
2
-
4
7
5
2
,
DOI
: 1
0
.
1
1
5
9
1
/ijeecs.v
24
.i
2
.
p
p
8
1
5
-
8
2
4
815
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ee
cs.ia
esco
r
e.
co
m
Feature
extractio
n wit
h
mel scal
e
s
epa
ra
tion metho
d
on no
ise
a
udio
rec
o
rding
s
Ro
y
Rudo
lf
H
uizen
,
F
lo
re
ntina
T
a
t
rin K
urnia
t
i
F
a
c
u
lt
y
o
f
I
n
fo
rm
a
ti
c
s a
n
d
C
o
m
p
u
ter,
In
st
it
u
t
Tek
n
o
l
o
g
i
d
a
n
Bisn
i
s S
TIKOM
Ba
li
,
Ba
li
,
In
d
o
n
e
sia
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
u
l
7
,
2
0
2
1
R
ev
is
ed
Sep
1
,
2
0
2
1
Acc
ep
ted
Sep
15
,
2
0
2
1
Th
is
p
a
p
e
r
f
o
c
u
se
s
o
n
imp
r
o
v
i
n
g
th
e
a
c
c
u
ra
c
y
o
f
n
o
ise
a
u
d
i
o
r
e
c
o
rd
in
g
s.
Hig
h
-
q
u
a
li
t
y
a
u
d
i
o
re
c
o
rd
in
g
,
e
x
trac
ti
o
n
u
si
n
g
th
e
m
e
l
fre
q
u
e
n
c
y
c
e
p
stra
l
c
o
e
fficie
n
ts
(M
F
CC)
m
e
th
o
d
p
ro
d
u
c
e
s
h
i
g
h
a
c
c
u
ra
c
y
.
W
h
il
e
t
h
e
l
o
w
-
q
u
a
li
ty
is
b
e
c
a
u
se
o
f
n
o
ise
,
t
h
e
a
c
c
u
ra
c
y
is
lo
w.
Im
p
r
o
v
e
d
a
c
c
u
ra
c
y
b
y
i
n
v
e
stig
a
ti
n
g
th
e
e
ffe
c
t
o
f
b
a
n
d
wid
t
h
o
n
th
e
m
e
l
sc
a
le.
Th
e
p
ro
p
o
se
d
imp
r
o
v
e
m
e
n
t
u
se
s th
e
m
e
l
sc
a
le
se
p
a
ra
ti
o
n
m
e
th
o
d
s
i
n
to
two
fre
q
u
e
n
c
y
c
h
a
n
n
e
ls
(M
F
CC
d
u
a
l
-
c
h
a
n
n
e
l).
F
o
r
th
e
c
o
m
p
a
riso
n
m
e
th
o
d
u
sin
g
th
e
m
e
l
sc
a
le
b
a
n
d
wid
th
with
o
u
t
se
p
a
ra
ti
o
n
(M
F
CC
si
n
g
le
-
c
h
a
n
n
e
l
).
F
e
a
tu
re
a
n
a
ly
sis
u
sin
g
k
-
m
e
a
n
c
lu
ste
rin
g
.
Th
e
d
a
ta
u
se
s
a
n
o
ise
v
a
rian
c
e
o
f
u
p
to
-
1
6
d
B
.
Tes
ti
n
g
o
n
th
e
M
F
CC
sin
g
le
-
c
h
a
n
n
e
l
m
e
th
o
d
fo
r
-
1
6
d
B
n
o
ise
h
a
s
a
n
a
c
c
u
ra
c
y
o
f
4
7
.
5
%
,
w
h
il
e
t
h
e
M
F
CC
d
u
a
l
-
c
h
a
n
n
e
l
m
e
th
o
d
h
a
s
a
n
a
c
c
u
ra
c
y
b
e
tt
e
r
o
f
7
6
.
2
5
%
.
T
h
e
n
e
x
t
tes
t
u
se
d
a
d
a
p
ti
v
e
n
o
ise
-
c
a
n
c
e
li
n
g
(
ANC
)
t
o
re
d
u
c
e
n
o
ise
b
e
fo
re
e
x
trac
ti
o
n
.
Th
e
re
su
lt
is
th
a
t
th
e
M
F
CC
sin
g
le
-
c
h
a
n
n
e
l
m
e
th
o
d
h
a
s
a
n
a
c
c
u
ra
c
y
o
f
8
2
.
5
%
a
n
d
t
h
e
M
F
CC
d
u
a
l
-
c
h
a
n
n
e
l
m
e
th
o
d
h
a
s
a
n
a
c
c
u
ra
c
y
b
e
tt
e
r
o
f
8
3
.
7
5
%
.
Hi
g
h
-
q
u
a
li
t
y
a
u
d
io
re
c
o
r
d
in
g
tes
ti
n
g
fo
r
t
h
e
M
F
CC sin
g
le
-
c
h
a
n
n
e
l
m
e
th
o
d
h
a
s a
n
a
c
c
u
ra
c
y
o
f
9
2
.
5
%
a
n
d
th
e
M
F
CC
d
u
a
l
-
c
h
a
n
n
e
l
m
e
th
o
d
h
a
s
a
n
a
c
c
u
ra
c
y
b
e
tt
e
r
o
f
9
7
.
5
%
.
Th
e
tes
t
re
su
lt
s
sh
o
w
th
e
e
ffe
c
t
o
f
m
e
l
sc
a
le
b
a
n
d
wid
t
h
t
o
in
c
re
a
se
a
c
c
u
ra
c
y
.
Th
e
M
F
CC
d
u
a
l
-
c
h
a
n
n
e
l
m
e
th
o
d
h
a
s h
i
g
h
e
r
a
c
c
u
ra
c
y
.
K
ey
w
o
r
d
s
:
Ad
ap
tiv
e
n
o
is
e
ca
n
ce
llatio
n
K
-
m
ea
n
clu
s
ter
in
g
MFC
C
d
u
al
-
ch
an
n
el
MFC
C
s
in
g
le
-
ch
an
n
el
No
is
e
au
d
io
r
ec
o
r
d
in
g
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
R
o
y
R
u
d
o
lf
Hu
izen
Facu
lty
o
f
I
n
f
o
r
m
atics a
n
d
C
o
m
p
u
ter
I
n
s
titu
t T
ek
n
o
lo
g
i d
an
B
is
n
is
STKO
M
B
ali
J
l.
R
ay
a
Pu
p
u
tan
No
.
8
6
,
Den
p
asar
,
B
ali
8
0
2
3
4
,
I
n
d
o
n
esia
E
m
ail: r
o
y
@
s
tik
o
m
-
b
ali.
ac
.
id
1.
I
NT
RO
D
UCT
I
O
N
User
s
o
f
d
ig
ital
d
ev
ices
ar
e
i
n
cr
ea
s
in
g
.
T
h
is
in
cr
ea
s
e
is
in
f
lu
en
ce
d
b
y
r
elativ
ely
lo
w
p
r
ices,
f
r
ee
ap
p
licatio
n
s
,
an
d
f
ast
in
ter
n
et
ac
ce
s
s
.
T
h
is
b
r
in
g
s
ch
an
g
es
an
d
h
a
b
its
.
On
p
r
ev
i
o
u
s
co
m
m
u
n
icatio
n
d
ev
ices
u
s
in
g
th
e
telep
h
o
n
e,
s
witch
to
ap
p
licatio
n
s
(
W
h
atsAp
p
an
d
T
eleg
r
am
)
[
1
]
,
[
2
]
.
T
h
is
ch
an
g
e
attr
ac
ted
th
e
atten
tio
n
o
f
law
en
f
o
r
ce
m
e
n
t
b
y
s
tar
tin
g
to
u
s
e
d
i
g
ital
ev
id
e
n
ce
to
p
r
o
v
e
a
ca
s
e
[
3
]
,
[
4
]
.
D
ig
ital
ev
id
en
ce
is
a
b
r
ea
k
th
r
o
u
g
h
to
f
in
d
o
u
t
s
o
m
eo
n
e'
s
in
v
o
lv
em
en
t
o
r
ac
ti
v
ity
,
an
ex
am
p
le
o
f
d
i
g
ital
e
v
id
en
ce
is
r
ec
o
r
d
e
d
co
n
v
er
s
atio
n
s
.
C
o
n
v
er
s
atio
n
c
an
b
e
d
ef
i
n
ed
as
in
ter
ac
tiv
e
c
o
m
m
u
n
icatio
n
b
etwe
en
in
d
iv
i
d
u
als
.
R
ec
o
r
d
in
g
a
co
n
v
er
s
atio
n
will c
lar
if
y
th
e
c
h
r
o
n
o
lo
g
y
o
f
a
n
ev
en
t
[
5
]
.
T
h
e
au
d
io
r
ec
o
r
d
s
ar
e
u
s
ef
u
l f
o
r
k
n
o
win
g
an
ev
e
n
t.
T
h
e
p
r
o
ce
s
s
i
s
b
y
co
n
v
er
tin
g
th
e
au
d
i
o
f
r
eq
u
e
n
cy
u
s
in
g
a
m
icr
o
p
h
o
n
e
in
to
a
n
e
lectr
ical
s
ig
n
al
an
d
s
av
in
g
it
in
th
e
f
o
r
m
o
f
a
f
ile
[
6
]
.
Au
d
io
tap
p
in
g
b
y
th
e
au
th
o
r
ities
is
an
ac
tiv
ity
t
o
r
e
co
r
d
with
o
u
t
b
ein
g
n
o
ticed
.
W
ir
etap
p
in
g
r
ec
o
r
d
in
g
s
ar
e
n
o
t
n
ec
ess
ar
ily
g
o
o
d
q
u
ality
,
b
ec
a
u
s
e
th
ey
ca
n
'
t
ch
o
o
s
e
th
e
en
v
i
r
o
n
m
e
n
tal
s
itu
atio
n
wh
en
th
e
tap
p
in
g
tak
es
p
l
ac
e
[
6
]
,
[
7
]
.
No
is
y
en
v
ir
o
n
m
en
t,
wir
etap
p
i
n
g
r
ec
o
r
d
in
g
is
n
o
is
e,
th
e
au
d
io
r
ec
o
r
d
in
g
q
u
ality
is
lo
w
[
7
]
.
T
h
is
af
f
ec
ts
th
e
ac
cu
r
ac
y
o
f
th
e
in
f
o
r
m
atio
n
c
o
n
tain
ed
in
th
e
r
ec
o
r
d
in
g
[
5
]
,
[
8
]
.
I
n
f
o
r
m
atio
n
o
n
a
a
u
d
io
r
ec
o
r
d
in
clu
d
es
in
d
iv
i
d
u
al
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
5
2
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci,
Vo
l.
24
,
No
.
2
,
No
v
em
b
er
2
0
2
1
:
8
1
5
-
824
816
p
r
o
f
iles
,
lo
ca
tio
n
s
,
an
d
c
h
r
o
n
o
lo
g
y
o
f
an
ev
en
t
[
9
]
.
Ob
tai
n
in
g
in
f
o
r
m
atio
n
b
y
ex
tr
ac
tin
g
wo
r
d
s
am
p
les,
to
o
b
tain
f
ea
tu
r
es
f
o
r
id
en
tific
a
tio
n
[
1
0
]
,
[
1
1
]
.
T
h
e
p
r
o
ce
s
s
is
b
y
m
atch
in
g
th
e
f
ea
tu
r
es
o
f
ev
id
en
ce
an
d
co
m
p
ar
is
o
n
.
Similar
ity
o
f
f
ea
tu
r
es
m
ea
n
s
th
at
th
e
a
u
d
io
r
e
co
r
d
s
co
m
e
f
r
o
m
t
h
e
s
am
e
in
d
iv
id
u
al
[
1
2
]
.
T
h
e
s
im
ilar
ity
o
f
f
ea
tu
r
es
is
ca
lcu
lated
f
r
o
m
th
e
h
ig
h
v
alu
e
o
f
ac
cu
r
ac
y
.
T
h
e
e
x
tr
ac
tio
n
m
et
h
o
d
is
a
f
ac
to
r
th
at
af
f
ec
ts
th
e
ac
cu
r
ac
y
v
alu
e.
T
h
e
m
el
f
r
eq
u
en
cy
ce
p
s
tr
al
co
ef
f
icien
ts
(
MFC
C
)
m
eth
o
d
is
a
r
e
liab
le
m
eth
o
d
with
h
ig
h
ac
cu
r
ac
y
f
o
r
h
ig
h
-
q
u
ality
au
d
io
r
ec
o
r
d
in
g
s
[
1
3
]
.
T
h
e
ac
cu
r
ac
y
r
ate
is
m
o
r
e
th
an
9
0
%
[
1
4
]
,
[
1
5
]
.
T
h
e
h
ig
h
ac
cu
r
ac
y
o
f
th
e
MFC
C
m
eth
o
d
is
d
u
e
to
th
e
m
el
s
ca
le
wh
ich
h
as
ch
ar
ac
ter
is
tic
s
s
im
ilar
to
h
u
m
an
h
ea
r
in
g
[
1
1
]
.
T
h
e
wo
r
d
s
am
p
l
e
if
ex
tr
ac
ted
b
y
th
e
MFC
C
m
eth
o
d
,
t
h
e
v
alu
e
o
f
th
e
f
r
eq
u
e
n
cy
co
m
p
o
n
en
t
will
m
atch
th
e
ch
ar
ac
ter
is
tics
o
f
th
e
m
el
s
ca
le.
T
h
e
r
esu
ltin
g
f
ea
tu
r
es
r
ep
r
esen
t
h
u
m
an
h
ea
r
in
g
.
Me
l
s
ca
le
s
en
s
itiv
ity
is
o
n
e
o
f
th
e
f
ac
to
r
s
th
at
in
cr
ea
s
e
th
e
ac
cu
r
ac
y
v
a
lu
e.
T
h
is
m
eth
o
d
is
s
u
p
er
io
r
f
o
r
th
e
id
en
tific
atio
n
o
f
h
ig
h
-
q
u
ality
au
d
io
r
ec
o
r
d
in
g
.
I
d
en
tific
atio
n
o
f
a
u
d
io
r
ec
o
r
d
in
g
s
h
as
tim
e
in
ter
v
als
o
f
u
p
to
ten
s
o
f
y
ea
r
s
with
co
m
p
ar
i
s
o
n
au
d
io
r
ec
o
r
d
in
g
s
(
ag
i
n
g
-
f
ac
t
o
r
)
,
ac
cu
r
ac
y
d
ec
r
ea
s
es
d
u
e
to
ch
a
n
g
es
in
f
r
eq
u
en
c
y
co
m
p
o
n
en
ts
[
1
6
]
.
T
h
e
s
am
e
ap
p
lies
to
au
d
io
r
ec
o
r
d
in
g
s
wi
th
n
o
is
e.
C
o
m
p
ar
in
g
n
o
is
e
wo
r
d
s
am
p
les
ca
u
s
es
a
d
ec
r
ea
s
e
in
ac
cu
r
ac
y
d
u
e
to
ch
an
g
es
i
n
th
e
f
r
eq
u
en
cy
co
m
p
o
n
en
t.
E
ac
h
wo
r
d
s
am
p
le
co
n
s
is
ts
o
f
s
ev
er
al
f
r
e
q
u
en
c
y
c
o
m
p
o
n
e
n
ts
,
n
am
ely
th
e
f
u
n
d
am
e
n
tal
f
r
eq
u
e
n
cy
an
d
th
e
r
eso
n
an
t
f
r
eq
u
e
n
cy
.
T
h
e
wir
etap
p
in
g
au
d
io
r
ec
o
r
d
in
g
co
n
tain
s
n
o
is
e.
T
h
e
f
r
eq
u
e
n
cy
c
o
m
p
o
n
en
ts
ar
e
t
h
e
au
d
io
f
r
eq
u
en
cy
an
d
th
e
n
o
is
e
f
r
eq
u
en
cy
.
So
th
at
t
h
e
f
ea
tu
r
es
o
b
tain
ed
c
o
n
s
is
t
o
f
a
co
m
b
in
atio
n
o
f
th
e
wo
r
d
s
am
p
le
f
r
eq
u
e
n
cy
with
n
o
is
e.
I
f
test
ed
with
a
co
m
p
ar
is
o
n
s
a
m
p
le,
th
e
ac
c
u
r
ac
y
is
lo
w
[
1
2
]
.
T
h
ese
r
esu
lts
lea
d
to
in
ac
cu
r
ate
an
aly
s
is
[
1
7
]
.
I
n
o
r
d
er
to
in
c
r
ea
s
e
th
e
ac
cu
r
ac
y
v
alu
e
o
f
l
o
w
-
q
u
ality
wir
etap
p
in
g
r
ec
o
r
d
in
g
s
d
u
e
to
n
o
is
e,
b
y
d
ev
elo
p
in
g
a
n
ex
tr
ac
tio
n
m
eth
o
d
.
2.
RE
S
E
ARCH
M
E
T
H
O
D
I
n
th
is
s
tu
d
y
,
th
e
p
r
o
p
o
s
ed
i
m
p
r
o
v
e
m
en
t
is
th
e
m
el
s
ca
le
s
ep
ar
atio
n
m
eth
o
d
,
b
ef
o
r
e
e
x
tr
ac
tin
g
th
e
wo
r
d
s
am
p
le
b
y
d
iv
id
in
g
it
in
t
o
two
f
r
eq
u
e
n
cy
p
a
r
ts
(
MFC
C
d
u
al
-
ch
an
n
el)
.
T
h
e
s
ep
a
r
atio
n
is
b
ased
o
n
lin
ea
r
h
u
m
an
h
ea
r
in
g
ch
ar
ac
ter
is
tics
f
o
r
f
r
e
q
u
en
cies
b
elo
w
1
k
Hz
a
n
d
lo
g
a
r
ith
m
ic
f
o
r
f
r
e
q
u
en
cies
ab
o
v
e
1
k
Hz.
T
h
e
s
tag
es
o
f
th
e
MF
C
C
d
u
al
-
ch
an
n
el
m
eth
o
d
as
s
h
o
wn
in
Fig
u
r
e
1
.
T
h
e
r
esear
ch
d
ata
u
s
es
a
s
am
p
le
o
f
wo
r
d
s
with
n
o
is
e,
o
b
tain
ed
f
r
o
m
c
o
n
v
er
s
atio
n
r
ec
o
r
d
in
g
s
b
y
a
d
d
in
g
r
a
n
d
o
m
n
o
is
e
u
s
in
g
a
c
o
m
p
u
ter
.
A
r
a
n
d
o
m
n
o
is
e
v
ar
ian
t is u
s
ed
f
r
o
m
lo
w
to
h
ig
h
.
Fig
u
r
e
1
.
I
d
en
tific
atio
n
s
tag
e
with
MFC
C
d
u
al
-
ch
an
n
el
2
.
1
.
Ada
ptiv
e
no
is
e
-
ca
nceling
T
h
e
n
o
is
e
r
ed
u
ctio
n
m
eth
o
d
i
n
th
is
s
tu
d
y
u
s
es
ad
ap
tiv
e
n
o
is
e
-
ca
n
ce
lin
g
(
ANC
)
with
th
e
least
m
ea
n
s
q
u
ar
e
(
LMS
)
alg
o
r
ith
m
[
1
8
]
,
[
1
9
]
.
T
h
is
m
eth
o
d
h
as
a
s
im
p
le
an
d
r
eliab
le
s
tr
u
ctu
r
e
[
2
0
]
,
[
2
1
]
.
T
h
e
s
tr
u
ctu
r
e
o
f
th
e
L
MS
alg
o
r
ith
m
is
s
h
o
wn
in
Fig
u
r
e
2
(
a)
.
I
d
en
tific
at
io
n
o
f
n
o
is
e
au
d
io
r
ec
o
r
d
in
g
s
,
th
e
f
ir
s
t
s
tep
is
to
r
ed
u
ce
n
o
is
e
b
ef
o
r
e
ex
tr
ac
tio
n
.
T
h
is
p
r
o
ce
s
s
r
eq
u
ir
es
r
ef
er
en
ce
n
o
is
e
(
X
k
)
.
I
n
p
u
t
(
X
k
)
a
s
r
ef
er
en
ce
n
o
is
e
is
p
r
o
ce
s
s
ed
b
y
a
n
ad
a
p
tiv
e
lin
ea
r
co
m
b
in
e
r
with
a
s
in
g
le
in
p
u
t,
as sh
o
wn
in
Fig
u
r
e
2
(
b
)
[
1
9
]
,
[
2
2
]
.
T
h
e
r
ef
e
r
en
ce
n
o
is
e
as
in
(
1
)
b
ec
o
m
es
th
e
in
p
u
t
o
f
th
e
a
d
ap
tiv
e
lin
ea
r
c
o
m
b
in
er
.
T
h
e
r
ef
e
r
en
ce
n
o
is
e
X
lk
-
1
p
ass
es
th
e
d
elay
tim
e
(
z
-
1
)
,
th
e
v
alu
e
o
f
X
lk
-
1
is
a
f
f
ec
t
ed
b
y
th
e
ad
ap
tiv
ely
ch
a
n
g
in
g
weig
h
t
(
w
k
)
.
T
h
e
o
u
tp
u
t
o
f
th
e
a
d
ap
tiv
e
lin
ea
r
c
o
m
b
in
er
is
s
h
o
wn
in
(
2
)
.
X
lk
-
1
=
[
x
k
x
k
-
1
x
k
-
2
…
x
k
-
L
]
T
(
1
)
=
∑
−
1
=
0
(
2
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2502
-
4
7
5
2
F
ea
tu
r
e
ex
tr
a
ctio
n
w
ith
mel
s
ca
le
s
ep
a
r
a
tio
n
meth
o
d
o
n
n
o
is
e
a
u
d
io
r
ec
o
r
d
in
g
s
(
R
o
y
R
u
d
o
lf Hu
iz
en
)
817
T
h
e
ad
ap
ti
v
e
lin
ea
r
co
m
b
i
n
er
o
u
tp
u
t
(
y
k
)
will
b
e
co
r
r
ec
ted
t
h
r
o
u
g
h
iter
atio
n
s
th
at
co
n
tin
u
e
u
n
til
th
e
m
ea
n
s
q
u
ar
e
d
er
r
o
r
(
MSE
)
is
m
in
im
al,
as sh
o
wn
in
(
3
)
.
ξ=
E
[
2
]
(
3
)
Min
im
u
m
MSE
m
ea
n
s
th
at
t
h
e
n
o
is
e
in
th
e
wo
r
d
s
am
p
le
is
r
ed
u
ce
d
[
2
3
]
.
T
h
e
n
o
is
e
r
ed
u
ctio
n
is
d
eter
m
in
ed
b
y
th
e
ac
cu
r
ac
y
o
f
th
e
weig
h
t
v
alu
e
o
n
th
e
a
d
a
p
tiv
e
lin
ea
r
co
m
b
in
e
r
.
E
ac
h
s
tag
e
o
f
iter
atio
n
,
th
e
am
o
u
n
t
o
f
er
r
o
r
is
co
r
r
ec
ted
to
th
e
d
esire
d
lim
it.
T
h
e
r
ed
u
cti
o
n
p
r
o
ce
s
s
er
r
o
r
in
ANC is
s
h
o
wn
in
(
4
)
.
ek
=
dk
-
Xk
W
T
k
(
4
)
B
ased
o
n
(
4
)
,
t
h
e
d
eter
m
in
i
n
g
f
ac
to
r
f
o
r
th
e
iter
atio
n
s
p
ee
d
an
d
th
e
m
in
im
u
m
MSE
is
d
eter
m
in
ed
f
r
o
m
th
e
ac
cu
r
ac
y
o
f
t
h
e
we
ig
h
t
v
alu
e.
I
n
o
r
d
er
to
o
b
tain
th
e
m
ax
im
u
m
r
e
d
u
ctio
n
v
alu
e
an
d
m
i
n
im
u
m
iter
atio
n
,
th
e
Steep
est De
s
ce
n
t m
eth
o
d
is
u
s
ed
,
b
y
e
n
ter
in
g
t
h
e
s
tep
s
ize
(
µ)
,
as sh
o
wn
in
(
5
)
.
+
1
=
−
(
5
)
I
n
(
5
)
th
e
c
o
r
r
ec
tio
n
o
f
th
e
w
eig
h
t
v
alu
e
u
s
in
g
a
ze
r
o
g
r
ad
ien
t
is
s
h
o
wn
in
(
6
)
.
C
h
an
g
es
in
th
e
weig
h
t
v
alu
e
(
s
u
b
tr
ac
tin
g
o
r
a
d
d
in
g
)
a
r
e
af
f
ec
te
d
b
y
th
e
r
ef
e
r
en
ce
n
o
is
e
v
alu
e
a
n
d
th
e
m
ag
n
itu
d
e
o
f
t
h
e
ANC
er
r
o
r
,
s
h
o
wn
in
(
6
)
.
B
ased
o
n
(
5
)
an
d
(
6
)
,
th
e
weig
h
t f
o
r
n
o
is
e
r
ed
u
ctio
n
o
b
tain
ed
b
y
(
7
)
is
k
n
o
wn
as th
e
L
MS
alg
o
r
ith
m
[
2
0
]
,
[
2
4
]
.
=
−
2
(
6
)
+
1
=
+
2
(
7
)
(
a)
(
b
)
Fig
u
r
e
2
.
Ad
a
p
tiv
e
n
o
is
e
-
ca
n
c
elin
g
with
L
MS
alg
o
r
ith
m
;
(
a)
L
MS
alg
o
r
ith
m
an
d
(
b
)
a
d
a
p
tiv
e
lin
ea
r
co
m
b
in
er
with
a
s
in
g
le
in
p
u
t
2
.
2
.
M
F
CC
du
a
l
-
cha
nn
el
E
x
tr
ac
tio
n
in
th
e
MFC
C
d
u
al
-
ch
an
n
el
m
eth
o
d
b
eg
i
n
s
b
y
s
ep
ar
atin
g
th
e
wo
r
d
s
am
p
le
in
t
o
two
p
ar
ts
.
On
ch
an
n
el
1
,
f
ilter
in
g
u
s
es
lo
w
p
ass
f
ilter
(
L
PF
)
f
o
r
f
r
eq
u
e
n
cies
less
th
an
1
k
Hz.
W
h
ile
ch
an
n
el
2
u
s
es
b
an
d
p
ass
f
ilter
(
B
PF
)
with
a
f
r
e
q
u
en
cy
b
etwe
en
1
k
Hz
to
4
k
Hz.
I
m
p
u
ls
e
r
esp
o
n
s
es
f
o
r
lo
w
p
ass
f
ilter
(
L
PF
)
,
h
ig
h
p
ass
f
ilter
,
an
d
b
a
n
d
p
ass
f
ilter
(
B
PF
)
r
ef
er
to
(
8
)
,
(
9
)
,
an
d
(
1
0
)
.
ℎ
[
]
=
Ω
0
(
Ω
0
)
(
8
)
h
[
n
]
=
s
in
c
(
π
n
)
−
Ω
0
π
s
in
c
(
Ω
0
n
)
(
9
)
h
bp
=
h
L
PH
−
h
L
PL
(
1
0
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
5
2
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci,
Vo
l.
24
,
No
.
2
,
No
v
em
b
er
2
0
2
1
:
8
1
5
-
824
818
T
h
e
d
eter
m
in
atio
n
o
f
th
e
b
an
d
wid
th
o
f
b
o
th
ch
an
n
els
r
ef
er
s
to
th
e
lin
ea
r
an
d
lo
g
ar
ith
m
ic
ch
ar
ac
ter
is
tics
o
f
h
u
m
a
n
h
ea
r
i
n
g
at
a
ce
r
tain
f
r
eq
u
en
cy
[
1
6
]
.
T
h
e
n
o
is
e
r
ec
o
r
d
i
n
g
(
s
am
p
le
wo
r
d
)
is
s
h
o
wn
in
Fig
u
r
e
3
(
a
)
.
R
e
c
o
r
d
s
a
r
e
s
e
p
a
r
a
t
e
d
b
y
f
i
n
i
t
e
i
m
p
u
l
s
e
r
es
p
o
n
s
e
(
F
I
R
)
[
2
5
]
,
[
2
6
]
.
T
h
e
r
e
s
u
l
t
s
h
o
w
n
i
n
F
i
g
u
r
e
3
(
b
)
fo
r
c
h
a
n
n
e
l
1
s
h
o
w
s
t
h
a
t
t
h
e
n
o
i
s
e
t
e
n
d
s
t
o
d
e
c
r
e
as
e
c
o
m
p
a
r
e
d
t
o
c
h
a
n
n
e
l
2
i
n
F
i
g
u
r
e
3
(
c
)
.
W
h
i
l
e
t
h
e
r
ec
o
r
d
i
n
g
o
f
n
o
i
s
e
r
e
d
u
c
e
d
b
y
A
N
C
i
s
s
h
o
w
n
i
n
F
i
g
u
r
e
3
(
d
)
.
T
h
e
r
e
s
u
l
t
s
o
f
t
h
e
s
e
p
a
r
a
ti
o
n
a
r
e
s
h
o
wn
i
n
F
i
g
u
r
e
3
(
e
)
,
f
o
r
ch
an
n
el
1
an
d
Fig
u
r
e
3
(
f
)
f
o
r
c
h
an
n
el
2
.
(
a)
(
b
)
(
c)
(
d
)
(
e)
(f)
Fig
u
r
e
3
.
W
o
r
d
s
s
am
p
le
;
(
a
)
wo
r
d
s
am
p
le
with
n
o
is
e
;
(
b
)
wo
r
d
s
am
p
le
with
a
n
o
is
e,
o
u
tp
u
t o
f
ch
a
n
n
el
1
;
(
c)
wo
r
d
s
am
p
le
with
a
n
o
is
e,
o
u
tp
u
t
o
f
ch
a
n
n
el
2
;
(
d
)
w
o
r
d
s
am
p
le
with
n
o
is
e,
r
ed
u
ce
d
u
s
in
g
ANC
;
(
e)
noi
se
-
r
e
d
u
ce
d
with
ANC,
o
u
tp
u
t c
h
an
n
el
1
;
(f)
noi
se
-
r
ed
u
ce
d
with
ANC,
o
u
tp
u
t c
h
an
n
el
2
T
h
e
wo
r
d
s
am
p
les
th
at
h
av
e
b
ee
n
s
ep
ar
atio
n
ar
e
th
e
n
ex
tr
ac
ted
u
s
in
g
MFC
C
with
th
e
f
o
llo
win
g
s
tep
s
;
f
r
am
e
b
lo
ck
in
g
,
win
d
o
win
g
,
f
ast
Fo
u
r
ier
tr
a
n
s
f
o
r
m
(
FFT
)
,
m
el
s
ca
le
f
ilter
b
an
k
,
d
is
cr
ete
co
s
in
e
tr
an
s
f
o
r
m
(
DC
T
)
an
d
m
el
f
r
e
q
u
en
cy
ce
p
s
tr
al
co
ef
f
icien
ts
[
1
2
]
.
T
h
e
f
r
eq
u
e
n
cy
co
m
p
o
n
e
n
t
co
n
tain
ed
in
th
e
s
am
p
le
wo
r
d
s
in
ea
ch
ch
an
n
e
l,
th
e
f
r
eq
u
en
c
y
v
alu
e
is
o
b
ta
in
ed
b
y
d
iv
i
d
in
g
th
e
s
am
p
le
wo
r
d
s
in
to
s
ev
er
al
f
r
am
es.
T
h
e
f
r
am
e
le
n
g
th
is
s
et
s
o
th
at
ea
ch
f
r
am
e
h
as
a
f
r
eq
u
en
cy
v
alu
e.
Settin
g
th
e
f
r
am
e
len
g
th
u
s
in
g
(
1
1
)
.
I
f
th
e
f
r
am
e
len
g
th
is
N
an
d
s
h
if
ted
b
y
M
th
en
ea
c
h
f
r
am
e
h
as
M
o
v
e
r
lap
.
T
h
e
f
r
a
m
e
b
lo
ck
in
g
p
r
o
ce
s
s
is
s
h
o
wn
in
Fig
u
r
e
4
.
(
1
:n
)
=
s
’
(
N+
M(
l
-
1
)
)
(
1
1
)
Fig
u
r
e
4
.
Fra
m
es b
l
o
ck
in
g
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2502
-
4
7
5
2
F
ea
tu
r
e
ex
tr
a
ctio
n
w
ith
mel
s
ca
le
s
ep
a
r
a
tio
n
meth
o
d
o
n
n
o
is
e
a
u
d
io
r
ec
o
r
d
in
g
s
(
R
o
y
R
u
d
o
lf Hu
iz
en
)
819
E
ac
h
f
r
am
e
is
win
d
o
wed
,
th
i
s
p
r
o
ce
s
s
aim
s
to
r
ed
u
ce
s
p
e
ctr
u
m
leak
ag
e
d
u
e
to
t
h
e
lo
w
s
am
p
lin
g
p
r
o
ce
s
s
.
T
h
e
ty
p
e
o
f
win
d
o
wi
n
g
u
s
ed
is
Ham
m
in
g
,
s
h
o
wn
i
n
(
1
2
)
an
d
(
1
3
)
;
(
n
)
=
0
.
5
4
–
0
.
4
6
c
o
s
(
2
n
/(
N
-
1
)
)
(
1
2
)
x
(
n
)
=x
l*
w(
n
)
(
1
3
)
e
ach
f
r
am
e
af
ter
win
d
o
win
g
is
ca
lcu
lated
th
e
f
r
eq
u
en
cy
v
alu
e,
u
s
in
g
d
is
cr
ete
Fo
u
r
ie
r
tr
an
s
f
o
r
m
(
DFT
)
(
co
n
v
e
r
tin
g
f
r
o
m
tim
e
d
o
m
ain
to
f
r
eq
u
en
c
y
)
.
A
to
tal
o
f
N
DFT
d
ata
is
ca
lcu
lated
u
s
in
g
FF
T
,
to
d
eter
m
in
e
th
e
f
r
eq
u
e
n
cy
v
al
u
e,
(
1
4
)
is
u
s
ed
.
(
)
=
∑
(
)
.
−
[
2
]
=
1
(
1
4
)
f
m
el_
ch
1
,
2
=
2
5
9
5
*
lo
g
1
0
[
1
+
f
l
in
700
]
(
1
5
)
On
ch
an
n
el
1
th
e
m
el
s
ca
le
i
s
f
o
r
s
ca
lin
g
f
r
e
q
u
en
cies
b
elo
w
1
KHz
a
n
d
c
h
an
n
el
2
is
f
o
r
s
ca
lin
g
f
r
eq
u
e
n
cies
b
etwe
en
1
to
4
KHz
.
T
h
e
f
r
eq
u
en
cy
o
f
t
h
e
p
r
o
ce
s
s
is
th
e
m
el
f
r
eq
u
e
n
cy
(
f
mel_ch1,
2
)
[
2
7
]
.
T
h
e
ch
ar
ac
ter
is
tics
o
f
th
e
m
el
s
ca
l
e
o
n
ch
an
n
els
1
an
d
2
f
o
llo
w
t
h
e
p
r
in
cip
le
th
at
n
o
t
all
f
o
llo
w
a
lin
ea
r
p
atter
n
,
s
o
th
at
ea
ch
c
h
an
n
el
f
o
llo
ws
a
lin
ea
r
an
d
e
x
p
o
n
en
tial
p
atter
n
.
T
h
e
f
r
e
q
u
en
c
y
is
th
e
n
o
n
t
h
e
m
el
s
ca
le
with
(
1
5
)
.
T
h
e
wid
th
o
f
th
e
m
el
s
ca
le
p
la
n
e
o
f
c
h
an
n
el
1
an
d
ch
an
n
el
2
is
s
h
o
wn
in
Fig
u
r
e
5
.
(
a)
(
b
)
Fig
u
r
e
5
.
Me
l scale
with
f
ilter
b
an
k
o
n
MFC
C
d
u
al
-
ch
an
n
el
;
(
a)
c
h
an
n
el
1
,
a
n
d
(
b
)
c
h
an
n
el
2
T
h
e
n
ex
t
s
tep
,
t
h
e
m
el
f
r
e
q
u
en
cy
v
alu
e
is
th
en
ca
lcu
lated
f
o
r
th
e
lo
g
m
el
f
r
eq
u
en
cy
u
s
in
g
(
1
6
)
.
Fu
r
th
er
m
o
r
e
,
with
(
1
7
)
t
h
e
m
e
l
ce
p
s
tr
u
m
co
ef
f
icien
t
v
alu
e
is
o
b
tain
e
d
wh
ic
h
is
th
e
f
ea
t
u
r
e
ce
p
tr
al
co
e
f
f
icien
t
o
f
th
e
wo
r
d
s
am
p
le
f
o
r
ea
ch
c
h
an
n
el.
_
ℎ
12
,
(
)
=
(
∑
|
_
ℎ
1
,
2
(
)
|
2
_
ℎ
1
,
2
=
1
)
(
1
6
)
_
ℎ
1
,
2
=
∑
_
_
ℎ
1
,
2
=
1
[
(
−
1
2
)
]
(
1
7
)
Fro
m
th
e
ex
tr
ac
tio
n
,
th
e
f
ea
tu
r
es o
b
tain
ed
ar
e
s
h
o
wn
in
c
h
an
n
els 1
an
d
2
s
h
o
wn
in
Fig
u
r
e
6
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
5
2
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci,
Vo
l.
24
,
No
.
2
,
No
v
em
b
er
2
0
2
1
:
8
1
5
-
824
820
(
a)
(
b
)
Fig
u
r
e
6
.
Featu
r
es th
e
r
esu
lt o
f
ex
tr
ac
tin
g
th
e
n
o
is
e
wo
r
d
s
am
p
le
an
d
r
e
d
u
ce
b
y
ANC
;
(
a)
c
h
an
n
el
1
f
ea
t
u
r
es
an
d
(
b
)
c
h
a
n
n
el
2
f
ea
tu
r
es
2
.
3
.
K
-
m
e
a
n c
lus
t
er
ing
T
h
e
f
ea
tu
r
e
a
n
aly
s
is
o
n
ea
c
h
ch
an
n
el
u
s
es
k
-
m
ea
n
clu
s
ter
in
g
,
th
e
d
ata
f
o
r
ea
ch
f
ea
tu
r
e
in
th
e
wo
r
d
s
am
p
le
is
ca
lcu
lated
f
o
r
th
e
cl
u
s
ter
ce
n
ter
[
2
3
]
,
[
2
8
]
.
I
d
en
tif
icatio
n
d
eter
m
in
es
wh
eth
er
a
s
am
p
le
o
f
wo
r
d
s
is
id
en
tical
o
r
n
o
t
b
y
co
m
p
ar
in
g
th
e
d
is
tan
ce
to
th
e
ce
n
ter
o
f
th
e
clu
s
ter
.
I
d
en
tical
if
th
e
ce
n
tr
o
id
o
f
th
e
w
o
r
d
s
am
p
le
h
as
a
h
ig
h
p
r
o
x
im
ity
v
alu
e
to
th
e
c
o
m
p
ar
is
o
n
wo
r
d
s
am
p
le
an
d
n
o
t
id
en
tical
i
f
t
h
e
two
s
am
p
les
ar
e
r
elativ
ely
f
ar
ap
ar
t b
etwe
en
th
e
ce
n
tr
o
id
s
.
T
h
e
s
tep
s
f
o
r
k
-
m
ea
n
clu
s
ter
in
g
ar
e
as f
o
llo
ws;
−
Step
1
: D
eter
m
in
e
th
e
s
tar
tin
g
p
o
in
t K
r
an
d
o
m
ly
f
r
o
m
th
e
f
ea
tu
r
e
d
ata
s
et.
T
h
e
n
ex
t iter
atio
n
u
s
es (
1
8
)
.
=
1
∑
=
1
(
1
8
)
−
Step
2
: E
ac
h
f
ea
tu
r
e
d
ata
with
in
itial c
en
tr
o
id
is
ca
lcu
lated
u
s
in
g
(
1
9
)
.
(
,
)
=
√
∑
(
−
)
2
=
1
(
1
9
)
−
Step
3
: U
p
d
ate
th
e
ce
n
tr
o
id
b
a
s
ed
o
n
th
e
ca
lcu
latio
n
o
f
s
tag
e
2
.
−
Step
4
: T
h
e
s
ec
o
n
d
s
tep
is
r
ep
ea
ted
u
n
til th
er
e
is
n
o
ch
an
g
e
i
n
th
e
c
en
tr
o
id
v
alu
e
o
r
it is
s
tab
le.
T
h
e
r
e
s
u
l
t
s
o
f
i
d
e
n
t
i
f
i
c
a
t
i
o
n
w
i
t
h
t
h
e
e
x
t
r
a
c
t
i
o
n
m
e
t
h
o
d
a
r
e
c
a
l
c
u
l
a
t
e
d
f
o
r
t
h
e
a
c
c
u
r
a
c
y
v
a
l
u
e
u
s
i
n
g
(
20
)
.
Acc
u
r
ac
y
=
Tp
+
TN
Tp
+
TN
+
FP
+
FN
(
2
0
)
T
o
d
eter
m
in
e
th
e
p
er
f
o
r
m
an
c
e
o
f
th
e
MFC
C
d
u
al
-
ch
an
n
el
an
d
MFC
C
s
in
g
le
-
ch
an
n
el
m
eth
o
d
s
b
y
ca
lcu
latin
g
th
e
v
alu
es
o
f
tr
u
e
p
o
s
itiv
e
(
TP
)
,
tr
u
e
n
eg
ativ
e
(
TN
)
,
f
alse
p
o
s
itiv
e
(
FP
)
,
f
alse
n
eg
ativ
e
(
FN
)
[
7
]
.
T
P
is
a
s
am
p
le
o
f
wo
r
d
s
s
tatin
g
tr
u
e
,
an
d
th
e
test
r
esu
lts
ar
e
id
en
tical.
T
N
is
a
s
am
p
le
o
f
wo
r
d
s
s
tatin
g
tr
u
e
,
an
d
th
e
test
r
esu
lts
ar
e
n
o
n
-
id
en
tical.
Fo
r
FP
,
th
e
s
am
p
le
wo
r
d
is
f
alse,
a
n
d
th
e
te
s
t
s
tates
id
en
tical.
Me
an
wh
ile,
FN
is
a
s
am
p
le
o
f
f
alse
wo
r
d
s
,
an
d
th
e
test
r
esu
lts
s
tate
th
at
th
ey
ar
e
n
o
n
-
i
d
en
t
ical.
T
o
d
eter
m
in
e
th
e
p
er
f
o
r
m
an
ce
o
f
th
e
MFC
C
d
u
al
-
ch
an
n
el
an
d
MFC
C
s
i
n
g
le
-
ch
an
n
el
m
eth
o
d
s
b
y
ca
lc
u
latin
g
th
e
ac
cu
r
ac
y
v
alu
e
b
y
co
m
p
a
r
in
g
T
P a
n
d
T
N
with
T
P,
T
N,
FP
,
FN a
s
s
h
o
wn
in
(
2
0
)
.
3.
RE
SU
L
T
S
A
ND
D
IS
CU
SS
I
O
N
E
x
p
er
im
en
t
o
n
MFC
C
s
in
g
le
-
ch
an
n
el
an
d
MFC
C
d
u
al
-
ch
an
n
el
m
eth
o
d
s
u
s
in
g
au
d
io
r
ec
o
r
d
in
g
d
ata
with
o
u
t n
o
is
e
an
d
with
n
o
is
e.
Fo
r
au
d
io
r
ec
o
r
d
in
g
s
with
n
o
i
s
e
u
s
in
g
a
n
o
is
e
v
ar
ian
t f
r
o
m
l
o
w
to
h
ig
h
.
T
h
e
test
d
ata
also
u
s
es
au
d
io
r
ec
o
r
d
in
g
s
with
r
ed
u
ce
d
n
o
is
e
wit
h
ANC.
T
h
e
f
ea
tu
r
e
an
aly
s
is
o
f
ea
ch
m
eth
o
d
u
s
es
k
-
m
ea
n
clu
s
ter
in
g
.
T
h
e
ex
p
er
i
m
en
tal
r
esu
lts
f
o
r
th
e
MFC
C
s
in
g
le
-
ch
an
n
el
m
eth
o
d
ar
e
s
h
o
wn
in
Fig
u
r
e
7
.
E
x
p
er
im
en
ts
u
s
in
g
th
e
MFC
C
s
in
g
le
-
ch
an
n
el
m
eth
o
d
to
s
a
m
p
le
wo
r
d
s
with
o
u
t
n
o
is
e,
t
h
e
r
esu
lts
ar
e
s
h
o
wn
in
F
ig
u
r
e
7
(
a)
.
T
h
e
clu
s
ter
ce
n
ter
b
etwe
en
th
e
test
an
d
co
m
p
ar
is
o
n
s
am
p
les h
as
a
h
i
g
h
clo
s
en
ess
v
alu
e.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2502
-
4
7
5
2
F
ea
tu
r
e
ex
tr
a
ctio
n
w
ith
mel
s
ca
le
s
ep
a
r
a
tio
n
meth
o
d
o
n
n
o
is
e
a
u
d
io
r
ec
o
r
d
in
g
s
(
R
o
y
R
u
d
o
lf Hu
iz
en
)
821
Ho
wev
er
,
i
n
th
e
s
am
p
le
wo
r
d
ex
p
er
im
en
t
with
n
o
is
e,
th
e
clu
s
ter
ce
n
ter
o
f
th
e
test
s
am
p
le
a
n
d
th
e
co
m
p
ar
is
o
n
o
f
th
e
p
r
o
x
im
ity
v
alu
es a
r
e
lo
w,
as sh
o
wn
in
Fig
u
r
e
7
(
b
)
.
T
h
e
th
ir
d
test
o
f
v
ar
ian
ce
u
s
ed
ANC
to
r
ed
u
ce
th
e
n
o
is
e
o
n
th
e
wo
r
d
s
am
p
le
b
ef
o
r
e
ex
tr
a
ctio
n
.
T
h
e
r
esu
lt
is
th
at
th
e
ce
n
tr
o
id
v
alu
es
in
th
e
test
wo
r
d
s
am
p
le
ar
e
clo
s
er
,
as
s
h
o
wn
in
Fig
u
r
e
7
(
c)
.
T
h
e
ex
p
er
i
m
en
t
f
o
r
t
h
e
MFC
C
d
u
al
-
ch
a
n
n
el
m
eth
o
d
u
s
ed
a
wo
r
d
s
am
p
le
with
o
u
t n
o
is
e,
a
wo
r
d
s
am
p
le
with
n
o
is
e,
an
d
a
wo
r
d
s
am
p
le
with
n
o
is
e
r
ed
u
ce
d
b
y
ANC.
T
h
e
test
r
esu
lts
ar
e
s
h
o
wn
in
Fig
u
r
e
8
.
T
h
e
e
x
p
e
r
i
m
e
n
t
f
o
r
t
h
e
M
F
C
C
d
u
a
l
-
c
h
a
n
n
e
l
m
e
t
h
o
d
u
s
e
s
a
w
o
r
d
s
a
m
p
l
e
w
i
t
h
o
u
t
n
o
i
s
e
v
a
l
u
e
o
f
t
h
e
c
l
u
s
t
e
r
c
e
n
t
r
o
i
d
o
n
c
h
a
n
n
e
l
s
1
a
n
d
2
h
a
s
a
h
i
g
h
c
l
o
s
e
n
e
s
s
t
o
th
e
c
o
m
p
a
r
i
s
o
n
w
o
r
d
s
a
m
p
l
e
as
s
h
o
w
n
i
n
F
i
g
u
r
e
8
.
I
n
th
e
ex
p
er
im
en
t o
f
th
e
MFC
C
d
u
al
-
ch
an
n
el
m
et
h
o
d
f
o
r
t
h
e
wo
r
d
s
am
p
le
with
o
u
t n
o
is
e,
i
n
ch
an
n
els
1
a
n
d
2
,
th
e
p
r
o
x
im
ity
o
f
th
e
clu
s
ter
ce
n
ter
to
th
e
co
m
p
ar
is
o
n
ce
n
ter
is
h
ig
h
,
as
s
h
o
wn
in
Fig
u
r
e
8
(
a)
.
As
f
o
r
th
e
wo
r
d
s
am
p
le
with
n
o
is
e,
th
e
r
esu
lts
o
n
ch
an
n
el
1
ar
e
clo
s
er
t
o
th
e
co
m
p
ar
is
o
n
th
an
ch
an
n
el
2
,
as
s
h
o
wn
in
Fig
u
r
e
8
(
b
)
.
I
n
t
h
e
test
o
f
th
e
wo
r
d
s
am
p
le
with
r
ed
u
ce
d
n
o
is
e
u
s
in
g
ANC,
th
e
r
esu
lt
is
th
at
th
e
clu
s
ter
ce
n
ter
b
etwe
en
th
e
test
wo
r
d
s
am
p
le
an
d
th
e
c
o
m
p
ar
is
o
n
wo
r
d
s
am
p
le
is
h
ig
h
,
as sh
o
wn
i
n
Fig
u
r
e
8
(
c)
.
T
h
e
e
x
p
e
r
i
m
e
n
t
u
s
e
d
d
a
t
a
w
i
t
h
n
o
i
s
e
v
a
r
i
a
n
c
e
f
r
o
m
0
d
B
t
o
-
1
6
d
B
.
T
h
e
m
o
d
e
l
s
t
e
s
t
e
d
a
r
e
M
F
C
C
s
i
n
g
l
e
-
c
h
a
n
n
e
l
a
n
d
M
F
C
C
d
u
a
l
-
c
h
a
n
n
e
l
,
t
h
e
a
c
c
u
r
a
c
y
v
a
l
u
e
s
,
a
s
s
h
o
w
n
i
n
F
i
g
u
r
e
9
.
S
N
R
v
a
r
i
a
n
c
e
i
n
w
o
r
d
s
a
m
p
l
e
s
t
o
d
e
t
e
r
m
i
n
e
t
h
e
p
e
r
f
o
r
m
a
n
c
e
o
f
M
F
C
C
s
i
n
g
l
e
-
c
h
a
n
n
e
l
a
n
d
M
F
C
C
d
u
a
l
-
c
h
a
n
n
e
l
m
e
t
h
o
d
s
.
W
o
r
d
s
a
m
p
l
e
s
w
i
t
h
S
N
R
-
1
0
d
B
a
n
d
-
1
6
d
B
w
i
t
h
o
u
t
b
e
i
n
g
r
e
d
u
c
e
d
b
y
A
N
C
,
t
h
e
r
e
s
u
l
t
s
f
o
r
t
h
e
M
F
C
C
s
i
n
g
l
e
-
c
h
a
n
n
e
l
m
e
t
h
o
d
h
a
v
e
a
c
c
u
r
a
c
y
v
a
l
u
e
s
o
f
5
7
.
5
%
a
n
d
4
7
.
5
%
.
M
e
a
n
w
h
i
l
e
,
w
i
t
h
t
h
e
M
F
C
C
d
u
a
l
-
c
h
a
n
n
e
l
m
e
t
h
o
d
,
t
h
e
a
c
c
u
r
a
c
y
v
a
l
u
e
s
a
r
e
8
2
%
a
n
d
7
6
.
2
%
.
B
a
s
e
d
o
n
t
h
e
s
e
e
x
p
e
r
i
m
e
n
t
s
,
s
h
o
w
i
n
g
t
h
e
M
F
C
C
d
u
a
l
-
c
h
a
n
n
e
l
m
e
t
h
o
d
i
s
m
o
r
e
r
e
s
i
s
t
a
n
t
t
o
n
o
i
s
e
t
h
a
n
t
h
e
M
F
C
C
s
i
n
g
l
e
-
c
h
a
n
n
e
l
m
e
t
h
o
d
,
t
h
e
e
x
p
e
r
i
m
e
n
t
a
l
r
e
s
u
l
t
s
,
a
s
s
h
o
w
n
i
n
F
i
g
u
r
e
9
(
a
)
.
T
h
e
h
ig
h
n
o
is
e
in
th
e
wo
r
d
s
am
p
le
af
f
ec
ts
th
e
ac
cu
r
ac
y
th
e
h
ig
h
er
th
e
n
o
is
e
v
alu
e,
th
e
lo
wer
th
e
ac
cu
r
ac
y
v
alu
e.
T
o
in
cr
ea
s
e
th
e
ac
cu
r
ac
y
v
alu
e
b
y
s
ep
a
r
a
tin
g
th
e
b
a
n
d
wid
th
o
n
m
el
s
ca
le
(
MFC
C
d
u
al
-
ch
an
n
el)
a
n
d
r
ed
u
c
i
n
g
n
o
is
e
(
ANC)
in
th
e
s
am
p
le
wo
r
d
s
b
ef
o
r
e
e
x
tr
ac
tio
n
.
T
h
e
ex
p
er
im
e
n
ts
with
SNR
-
1
6
d
B
o
n
wo
r
d
s
am
p
le
s
r
ed
u
ce
b
y
ANC,
th
e
r
esu
lts
with
th
e
MFC
C
s
in
g
le
-
ch
an
n
el
m
eth
o
d
with
a
n
ac
cu
r
ac
y
o
f
8
2
.
5
%,
wh
ile
th
e
MFC
C
d
u
al
-
ch
an
n
el
m
eth
o
d
with
an
ac
cu
r
ac
y
v
alu
e
o
f
8
3
.
7
5
%.
T
h
e
am
o
u
n
t
o
f
n
o
is
e
r
ed
u
ctio
n
with
ANC
s
h
o
ws
a
p
r
o
p
o
r
tio
n
al
in
cr
ea
s
e
i
n
th
e
v
alu
e
o
f
ac
cu
r
ac
y
.
T
ests
u
s
in
g
h
ig
h
-
q
u
ality
r
ec
o
r
d
in
g
s
(
with
o
u
t
n
o
is
e)
,
th
e
r
esu
lts
f
o
r
th
e
MFC
C
s
in
g
le
-
ch
an
n
el
m
eth
o
d
h
a
v
e
an
ac
cu
r
ac
y
v
alu
e
o
f
9
2
.
5
%,
th
e
MFC
C
d
u
al
-
ch
an
n
el
m
eth
o
d
h
as a
n
ac
cu
r
ac
y
v
al
u
e
o
f
9
7
.
5
%,
as sh
o
wn
in
Fig
u
r
e
9
(
b
)
.
(
a)
(
b
)
(
c)
Fig
u
r
e
7
.
MFC
C
s
in
g
le
-
ch
an
n
el
;
(
a)
ex
tr
a
cted
wo
r
d
s
am
p
le
with
o
u
t n
o
is
e
,
(
b
)
t
h
e
e
x
tr
ac
tio
n
r
esu
lt f
o
r
th
e
n
o
is
e
wo
r
d
s
am
p
le
is
-
6
d
B
,
(
c
)
e
x
tr
a
ctio
n
r
esu
lt f
o
r
-
6
d
B
n
o
is
e
wo
r
d
s
am
p
le
r
ed
u
ce
d
b
y
A
NC
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
5
2
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci,
Vo
l.
24
,
No
.
2
,
No
v
em
b
er
2
0
2
1
:
8
1
5
-
824
822
(
a)
(
b
)
(
c)
Fig
u
r
e
8
.
MFC
C
d
u
al
-
ch
an
n
el
m
eth
o
d
;
(
a
)
ex
tr
ac
te
d
wo
r
d
s
am
p
le
with
o
u
t n
o
is
e
;
(
b
)
t
h
e
e
x
tr
ac
tio
n
r
esu
lt f
o
r
th
e
n
o
is
e
wo
r
d
s
am
p
le
is
-
6
d
B
;
(
c)
e
x
tr
ac
tio
n
r
esu
lt f
o
r
-
6
d
B
n
o
is
e
wo
r
d
s
am
p
le
r
ed
u
ce
d
b
y
ANC
(
a)
(
b
)
Fig
u
r
e
9
.
MFC
C
s
in
g
le
-
ch
an
n
el
an
d
MFC
C
d
u
al
-
ch
an
n
el
m
eth
o
d
s
;
(
a)
w
ith
o
u
t A
NC
an
d
(
b
)
w
ith
ANC
4.
CO
NCLU
SI
O
N
B
ased
o
n
th
e
ex
p
er
im
en
t,
it
i
s
co
n
clu
d
ed
th
at
th
e
r
ec
o
r
d
ed
n
o
is
e
af
f
ec
ts
th
e
ac
cu
r
ac
y
.
Hig
h
n
o
is
e
ca
u
s
es
lo
w
ac
cu
r
ac
y
.
I
n
t
h
e
s
am
p
le
wo
r
d
s
with
SNR
-
1
6
d
B
u
s
in
g
th
e
MFC
C
s
in
g
le
-
ch
an
n
el
m
eth
o
d
with
4
7
.
5
%
ac
c
u
r
ac
y
an
d
th
e
MFC
C
d
o
u
b
le
-
c
h
an
n
e
l
m
eth
o
d
wit
h
7
6
.
2
5
%
ac
cu
r
ac
y
.
T
h
e
u
s
e
o
f
ANC
ca
n
i
n
cr
ea
s
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2502
-
4
7
5
2
F
ea
tu
r
e
ex
tr
a
ctio
n
w
ith
mel
s
ca
le
s
ep
a
r
a
tio
n
meth
o
d
o
n
n
o
is
e
a
u
d
io
r
ec
o
r
d
in
g
s
(
R
o
y
R
u
d
o
lf Hu
iz
en
)
823
th
e
ac
cu
r
ac
y
o
f
th
e
MFC
C
s
i
n
g
le
-
ch
an
n
el
b
y
8
2
.
5
%
a
n
d
MFC
C
d
u
al
-
ch
an
n
el
m
eth
o
d
b
y
8
3
.
7
5
%.
I
n
au
d
i
o
r
ec
o
r
d
in
g
with
o
u
t
n
o
is
e
f
o
r
th
e
MFC
C
s
in
g
le
-
ch
an
n
el
m
eth
o
d
,
t
h
e
ac
cu
r
ac
y
v
alu
e
is
9
2
.
5
%.
T
h
e
MFC
C
d
u
al
-
ch
an
n
el
ac
cu
r
a
c
y
is
9
7
.
5
%.
B
ased
o
n
th
e
test
r
esu
lts
,
th
e
MFC
C
d
u
al
-
ch
an
n
el
m
eth
o
d
h
as
h
ig
h
er
ac
cu
r
ac
y
th
an
th
e
MFC
C
s
in
g
le
-
ch
an
n
el
f
o
r
r
ec
o
r
d
in
g
with
o
r
with
o
u
t
n
o
is
e.
T
h
e
u
s
e
o
f
th
e
m
el
s
ca
le
s
ep
ar
atio
n
m
eth
o
d
in
MFC
C
d
u
al
-
ch
an
n
e
l c
an
in
cr
ea
s
e
th
e
ac
cu
r
ac
y
v
alu
e.
RE
F
E
R
E
NC
E
S
[1
]
S
.
C.
S
a
t
h
e
a
n
d
N.
M
.
D
o
n
g
r
e
,
“
Da
ta
a
c
q
u
isit
i
o
n
tec
h
n
iq
u
e
s
in
m
o
b
il
e
f
o
re
n
sic
s,”
2
0
1
8
2
n
d
I
n
ter
n
a
t
io
n
a
l
Co
n
fer
e
n
c
e
o
n
In
v
e
n
t
ive
S
y
ste
ms
a
n
d
Co
n
tro
l
(ICIS
C)
,
2
0
1
8
,
p
p
.
2
8
0
-
2
8
6
,
d
o
i:
1
0
.
1
1
0
9
/ICIS
C.
2
0
1
8
.
8
3
9
9
0
7
9
.
[2
]
H.
F
.
Tay
e
b
a
n
d
C.
Va
ro
l,
“
An
d
r
o
id
M
o
b
il
e
De
v
ice
F
o
re
n
sic
s:
A
Re
v
iew
,
”
2
0
1
9
7
th
I
n
ter
n
a
ti
o
n
a
l
S
y
mp
o
si
u
m
o
n
Dig
it
a
l
Fo
re
n
sic
s
a
n
d
S
e
c
u
rity (I
S
DFS
)
,
2
0
1
9
,
p
p
.
1
-
7
,
d
o
i:
1
0
.
1
1
0
9
/IS
DFS
.
2
0
1
9
.
8
7
5
7
4
9
3
.
[3
]
A.
Va
ro
l
a
n
d
Y.
Ü.
S
ö
n
m
e
z
,
"
Re
v
iew
o
f
e
v
id
e
n
c
e
a
n
a
ly
sis
a
n
d
re
p
o
rti
n
g
p
h
a
se
s
in
d
ig
i
tal
fo
re
n
sic
s
p
ro
c
e
ss
,
"
2
0
1
7
In
ter
n
a
t
io
n
a
l
Co
n
fer
e
n
c
e
o
n
Co
mp
u
ter
S
c
ien
c
e
a
n
d
En
g
i
n
e
e
rin
g
(UBM
K)
,
2
0
1
7
,
p
p
.
9
2
3
-
9
2
8
,
d
o
i:
1
0
.
1
1
0
9
/UBM
K.
2
0
1
7
.
8
0
9
3
5
6
3
.
[4
]
Z.
Ali,
M
.
Im
ra
n
,
a
n
d
M
.
Alsu
la
ima
n
,
“
An
Au
t
o
m
a
t
ic
Dig
it
a
l
Au
d
io
A
u
th
e
n
ti
c
a
ti
o
n
/F
o
re
n
sic
s
S
y
ste
m
,
”
in
IEE
E
Acc
e
ss
,
v
o
l.
5
,
p
p
.
2
9
9
4
-
3
0
0
7
,
2
0
1
7
,
d
o
i:
1
0
.
1
1
0
9
/ACCE
S
S
.
2
0
1
7
.
2
6
7
2
6
8
1
.
[5
]
G
.
De
lg
a
d
o
-
G
u
ti
é
rre
z
,
F
.
Ro
d
ríg
u
e
z
-
S
a
n
to
s,
O.
Jim
é
n
e
z
-
Ra
m
írez
,
a
n
d
R
.
Vá
z
q
u
e
z
-
M
e
d
i
n
a
,
“
Ac
o
u
s
t
i
c
e
n
v
iro
n
m
e
n
t
id
e
n
ti
fica
ti
o
n
b
y
Ku
ll
b
a
c
k
–
Leib
l
e
r
d
iv
e
r
g
e
n
c
e
,
”
Fo
re
n
sic
S
c
ien
c
e
In
ter
n
a
ti
o
n
a
l
,
v
o
l
.
2
8
1
,
p
p
.
1
3
4
-
1
4
0
,
2
0
1
7
,
d
o
i:
1
0
.
1
0
1
6
/j
.
f
o
rsc
ii
n
t
.
2
0
1
7
.
1
0
.
0
3
1
.
[6
]
V.
A.
Ha
d
o
lt
i
k
a
r,
V.
R.
Ra
t
n
a
p
a
rk
h
e
,
a
n
d
R
.
Ku
m
a
r
,
“
Op
ti
m
iza
ti
o
n
o
f
M
F
CC
p
a
ra
m
e
ters
fo
r
m
o
b
il
e
p
h
o
n
e
re
c
o
g
n
it
i
o
n
fr
o
m
a
u
d
io
re
c
o
rd
i
n
g
s,”
2
0
1
9
3
r
d
I
n
ter
n
a
ti
o
n
a
l
c
o
n
fer
e
n
c
e
o
n
El
e
c
tro
n
ics
,
Co
m
mu
n
ica
t
io
n
a
n
d
Aer
o
sp
a
c
e
T
e
c
h
n
o
l
o
g
y
(ICECA
)
,
2
0
1
9
,
p
p
.
7
7
7
-
7
8
0
,
d
o
i:
1
0
.
1
1
0
9
/I
CECA.
2
0
1
9
.
8
8
2
2
1
7
7
.
[7
]
M
.
S
.
At
h
u
l
y
a
,
Vi
n
a
y
sh
a
n
k
a
r
a
n
d
P
.
S
.
S
a
t
h
id
e
v
i,
“
M
it
ig
a
ti
n
g
e
ff
e
c
ts
o
f
n
o
ise
i
n
F
o
re
n
sic
S
p
e
a
k
e
r
Re
c
o
g
n
it
io
n
,
”
2
0
1
7
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
W
ire
les
s
Co
mm
u
n
ica
ti
o
n
s,
S
ig
n
a
l
Pro
c
e
ss
in
g
a
n
d
Ne
two
rk
in
g
(W
iS
PNE
T
)
,
2
0
1
7
,
p
p
.
1
6
0
2
-
1
6
0
6
,
d
o
i:
1
0
.
1
1
0
9
/W
iS
P
NET.
2
0
1
7
.
8
3
0
0
0
3
1
.
[8
]
R.
Ah
m
a
d
a
n
d
S
.
S
u
y
a
n
t
o
,
“
Th
e
Im
p
a
c
t
o
f
Lo
w
-
P
a
ss
F
il
ter
in
S
p
e
a
k
e
r
Id
e
n
ti
fica
ti
o
n
,
”
2
0
1
9
In
ter
n
a
ti
o
n
a
l
S
e
mi
n
a
r
o
n
Res
e
a
rc
h
o
f
In
f
o
rm
a
ti
o
n
T
e
c
h
n
o
l
o
g
y
a
n
d
In
telli
g
e
n
t
S
y
ste
ms
(IS
RIT
I)
,
2
0
1
9
,
p
p
.
1
3
3
-
1
3
6
,
d
o
i:
1
0
.
1
1
0
9
/IS
RITI
4
8
6
4
6
.
2
0
1
9
.
9
0
3
4
5
9
2
.
[9
]
H.
Ya
n
g
,
P
.
Zh
o
u
,
T.
F
u
k
u
d
a
,
a
n
d
H.
A.
Zh
a
o
,
“
De
-
No
isin
g
Us
in
g
Du
a
l
T
h
re
sh
o
ld
F
u
n
c
ti
o
n
fo
r
S
p
e
a
k
e
r
Re
c
o
g
n
it
i
o
n
a
t
Lo
w
S
NR,”
2
0
1
8
In
ter
n
a
t
io
n
a
l
Co
n
fer
e
n
c
e
o
n
M
a
c
h
in
e
L
e
a
rn
i
n
g
a
n
d
Cy
b
e
r
n
e
ti
c
s
(I
CM
L
C)
,
2
0
1
8
,
p
p
.
1
2
1
-
1
2
5
,
d
o
i:
1
0
.
1
1
0
9
/ICM
L
C.
2
0
1
8
.
8
5
2
7
0
3
3
.
[1
0
]
S
.
Ra
n
jan
,
P
.
G
a
rh
wa
l,
A.
B
h
a
n
,
M
.
Ar
o
ra
a
n
d
A.
M
e
h
ra
,
"
F
ra
m
e
wo
rk
f
o
r
Im
a
g
e
F
o
rg
e
ry
De
tec
ti
o
n
a
n
d
Clas
sifica
ti
o
n
Us
in
g
M
a
c
h
in
e
Lea
rn
in
g
,
"
2
0
1
8
2
n
d
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
T
re
n
d
s
i
n
El
e
c
tro
n
ics
a
n
d
In
fo
rm
a
t
ics
(ICOEI)
,
2
0
1
8
,
p
p
.
1
-
9
,
d
o
i:
1
0
.
1
1
0
9
/ICOEI.
2
0
1
8
.
8
5
5
3
9
24
.
[1
1
]
A.
Niwa
tk
a
r
a
n
d
Y.
K.
Ka
n
se
,
“
F
e
a
tu
re
Ex
trac
ti
o
n
u
sin
g
Wav
e
le
t
Tran
sfo
rm
a
n
d
E
u
c
li
d
e
a
n
Dista
n
c
e
fo
r
sp
e
a
k
e
r
re
c
o
g
n
it
i
o
n
sy
ste
m
,
”
2
0
2
0
In
ter
n
a
ti
o
n
a
l
C
o
n
fer
e
n
c
e
o
n
In
d
u
str
y
4
.
0
T
e
c
h
n
o
l
o
g
y
(I4
T
e
c
h
)
,
2
0
2
0
,
p
p
.
1
4
5
-
1
4
7
,
d
o
i:
1
0
.
1
1
0
9
/I4
Tec
h
4
8
3
4
5
.
2
0
2
0
.
9
1
0
2
6
8
3
.
[1
2
]
F
.
A
m
e
l
i
a
a
n
d
D
.
G
u
n
a
w
a
n
,
“
D
W
T
-
M
F
C
C
M
e
t
h
o
d
f
o
r
S
p
e
a
k
e
r
R
e
c
o
g
n
i
t
i
o
n
S
y
s
t
e
m
w
i
t
h
N
o
i
s
e
,
”
2
0
1
9
7
t
h
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
S
m
a
r
t
C
o
m
p
u
t
i
n
g
&
C
o
m
m
u
n
i
c
a
t
i
o
n
s
(
I
C
S
C
C
)
,
2
0
1
9
,
p
p
.
1
-
5
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
S
C
C
.
2
0
1
9
.
8
8
4
3
6
6
0
.
[1
3
]
F
.
Y
.
L
e
u
a
n
d
G
.
L
.
L
i
n
,
“
A
n
M
F
C
C
-
b
a
s
e
d
s
p
e
a
k
e
r
i
d
e
n
t
i
f
i
c
a
t
i
o
n
s
y
s
t
e
m
,
”
2
0
1
7
I
E
E
E
3
1
s
t
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
A
d
v
a
n
c
e
d
I
n
f
o
r
m
a
t
i
o
n
N
e
t
w
o
r
k
i
n
g
a
n
d
A
p
p
l
i
c
a
t
i
o
n
s
(
A
I
N
A
)
,
2
0
1
7
,
p
p
.
1
0
5
5
-
1
0
6
2
,
d
o
i
:
1
0
.
1
1
0
9
/
A
I
N
A
.
2
0
1
7
.
1
3
0
.
[1
4
]
O.
C
.
Ai,
M
.
Ha
rih
a
ra
n
,
S
.
Ya
a
c
o
b
,
a
n
d
L.
S
.
Ch
e
e
,
“
Clas
sifica
ti
o
n
o
f
sp
e
e
c
h
d
y
sfl
u
e
n
c
ies
wit
h
M
F
CC
a
n
d
LP
CC
fe
a
tu
re
s,”
Exp
e
rt
S
y
ste
ms
wit
h
A
p
p
li
c
a
ti
o
n
s
,
v
o
l.
3
9
,
n
o
.
2
,
p
p
.
2
1
5
7
-
2
1
6
5
,
F
e
b
ru
a
ry
2
0
1
2
,
d
o
i:
1
0
.
1
0
1
6
/j
.
e
sw
a
.
2
0
1
1
.
0
7
.
0
6
5
.
[1
5
]
M
.
T.
S
.
Al
-
Ka
lt
a
k
c
h
i,
H
.
A.
Al
-
Ra
h
e
e
m
Tah
a
,
M
.
A.
S
h
e
h
a
b
,
a
n
d
M
.
A
.
M
.
A
b
d
u
ll
a
h
,
“
Co
m
p
a
riso
n
o
f
fe
a
tu
re
e
x
trac
ti
o
n
a
n
d
n
o
rm
a
li
z
a
ti
o
n
m
e
t
h
o
d
s
fo
r
s
p
e
a
k
e
r
re
c
o
g
n
it
i
o
n
u
sin
g
g
r
id
-
a
u
d
io
v
isu
a
l
d
a
tab
a
se
,
”
In
d
o
n
e
sia
n
J
o
u
rn
a
l
o
f
El
e
c
trica
l
E
n
g
i
n
e
e
rin
g
a
n
d
Co
mp
u
ter
S
c
ien
c
e
(IJ
EE
C
S
)
,
v
o
l.
1
8
,
n
o
.
2
,
p
p
.
7
8
2
-
7
8
9
,
2
0
2
0
,
d
o
i:
1
0
.
1
1
5
9
1
/
ij
e
e
c
s.v
1
8
.
i2
.
p
p
7
8
2
-
7
8
9
.
[1
6
]
R.
R.
Hu
ize
n
,
J.
E.
Isti
y
a
n
t
o
,
a
n
d
A.
E.
P
u
tra,
“
Id
e
n
ti
f
ica
ti
o
n
o
f
Vo
ice
Uttera
n
c
e
with
Ag
in
g
F
a
c
to
r
Us
in
g
t
h
e
M
e
th
o
d
o
f
M
F
CC M
u
lt
ich
a
n
n
e
l
,
”
In
ter
n
a
ti
o
n
a
l
J
o
u
r
n
a
l
o
f
A
d
v
a
n
c
e
d
S
t
u
d
ies
i
n
Co
m
p
u
ter
S
c
ien
c
e
a
n
d
E
n
g
in
e
e
rin
g
(IJ
A
S
CS
E)
,
v
o
l.
6
,
n
o
.
0
1
,
p
p
.
1
4
-
1
9
,
2
0
1
7
.
[On
l
in
e
]
A
v
a
il
a
b
le
:
h
t
tp
:/
/www
.
ij
a
sc
se
.
o
rg
/v
o
l
u
m
e
-
6
-
th
e
m
e
-
b
a
se
d
-
issu
e
-
1
/M
F
CC
_
M
u
lt
ich
a
n
n
e
l.
p
d
f.
[1
7
]
Z.
Ala
v
i
a
n
d
B.
Az
imi
,
“
Ap
p
li
c
a
ti
o
n
o
f
En
v
ir
o
n
m
e
n
t
No
ise
Clas
sif
ica
ti
o
n
to
wa
rd
s
S
o
u
n
d
Re
c
o
g
n
it
i
o
n
fo
r
Co
c
h
lea
r
Im
p
la
n
t
Us
e
rs,”
2
0
1
9
6
th
I
n
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
El
e
c
trica
l
a
n
d
El
e
c
tro
n
ics
En
g
in
e
e
rin
g
(
ICEE
E)
,
2
0
1
9
,
p
p
.
1
4
4
-
1
4
8
,
d
o
i:
1
0
.
1
1
0
9
/ICE
EE
2
0
1
9
.
2
0
1
9
.
0
0
0
3
5
.
[1
8
]
K.
M
o
n
d
a
l
(Da
s)
e
t
a
l
.
,
“
All
-
p
a
ss
fil
tere
d
x
lea
st
m
e
a
n
sq
u
a
re
a
lg
o
rit
h
m
fo
r
n
a
rro
w
b
a
n
d
a
c
ti
v
e
n
o
ise
c
o
n
tr
o
l,
”
Ap
p
li
e
d
Aco
u
stics
,
v
o
l.
1
4
2
,
p
p
.
1
-
1
0
,
2
0
1
8
,
d
o
i:
1
0
.
1
0
1
6
/j
.
a
p
a
c
o
u
s
t.
2
0
1
8
.
0
7
.
0
2
6
.
[1
9
]
E.
C.
M
e
n
g
u
c
,
“
Ad
a
p
t
iv
e
f
o
u
rie
r
li
n
e
a
r
c
o
m
b
i
n
e
r
b
a
se
d
o
n
m
o
d
ifi
e
d
lea
st
m
e
a
n
k
u
r
to
sis
a
l
g
o
rit
h
m
,
”
2
0
1
8
6
t
h
In
ter
n
a
t
io
n
a
l
Co
n
fer
e
n
c
e
o
n
Co
n
tro
l
En
g
in
e
e
rin
g
&
In
fo
r
ma
ti
o
n
T
e
c
h
n
o
lo
g
y
(CEIT
)
,
2
0
1
8
,
p
p
.
1
-
5
,
d
o
i:
1
0
.
1
1
0
9
/CE
IT
.
2
0
1
8
.
8
7
5
1
8
9
1
.
[2
0
]
Y.
S
.
Wan
g
,
T.
P
.
F
e
n
g
,
X.
L.
W
a
n
g
,
H.
G
u
o
,
a
n
d
H.
Z.
Qi
,
“
An
i
m
p
ro
v
e
d
L
M
S
a
lg
o
r
it
h
m
fo
r
a
c
ti
v
e
so
u
n
d
-
q
u
a
li
ty
c
o
n
tro
l
o
f
v
e
h
icle
in
terio
r
n
o
ise
b
a
se
d
o
n
a
u
d
it
o
ry
m
a
sk
in
g
e
ffe
c
t,
”
M
e
c
h
a
n
ica
l
S
y
ste
ms
a
n
d
S
i
g
n
a
l
Pro
c
e
ss
in
g
,
v
o
l.
1
0
8
,
p
p
.
5
8
-
7
2
,
2
0
1
8
,
d
o
i:
1
0
.
1
0
1
6
/
j.
y
m
ss
p
.
2
0
1
8
.
0
2
.
0
1
8
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
5
2
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci,
Vo
l.
24
,
No
.
2
,
No
v
em
b
er
2
0
2
1
:
8
1
5
-
824
824
[2
1
]
X.
Zh
a
n
g
,
X.
Z
o
u
,
M
.
S
u
n
,
T.
F
.
Z
h
e
n
g
,
C.
Jia
,
a
n
d
Y
.
Wa
n
g
,
“
No
ise
R
o
b
u
st
S
p
e
a
k
e
r
Re
c
o
g
n
it
io
n
Ba
se
d
o
n
Ad
a
p
ti
v
e
F
ra
m
e
Weig
h
ti
n
g
in
G
M
M
fo
r
i
-
Ve
c
to
r
E
x
trac
ti
o
n
,
”
in
IEE
E
Acc
e
ss
,
v
o
l.
7
,
p
p
.
2
7
8
7
4
-
2
7
8
8
2
,
2
0
1
9
,
d
o
i:
1
0
.
1
1
0
9
/ACCES
S
.
2
0
1
9
.
2
9
0
1
8
1
2
.
[2
2
]
P
.
K.
Da
sh
,
D.
P
.
S
wa
in
,
A.
C.
L
iew
,
a
n
d
S
.
Ra
h
m
a
n
,
“
An
a
d
a
p
ti
v
e
li
n
e
a
r
c
o
m
b
i
n
e
r
fo
r
o
n
-
l
in
e
tra
c
k
in
g
o
f
p
o
we
r
sy
ste
m
h
a
rm
o
n
ics
,
”
in
IEE
E
T
ra
n
sa
c
ti
o
n
s
o
n
Po
we
r
S
y
ste
ms
,
v
o
l
.
1
1
,
n
o
.
4
,
p
p
.
1
7
3
0
-
1
7
3
5
,
No
v
.
1
9
9
6
,
d
o
i:
1
0
.
1
1
0
9
/5
9
.
5
4
4
6
3
5
.
[2
3
]
V.
Div
y
a
,
R.
De
e
p
ik
a
,
C
.
Ya
m
in
i,
a
n
d
P
.
S
o
b
i
y
a
a
,
“
An
Eff
icie
n
t
K
-
M
e
a
n
s
Clu
ste
rin
g
In
it
i
a
li
z
a
ti
o
n
Us
in
g
Op
ti
m
iza
ti
o
n
Alg
o
rit
h
m
,
”
2
0
1
9
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
Ad
v
a
n
c
e
s
i
n
Co
m
p
u
t
in
g
a
n
d
Co
mm
u
n
ica
ti
o
n
En
g
i
n
e
e
rin
g
(
ICACCE
)
,
2
0
1
9
,
p
p
.
1
-
7
,
d
o
i:
1
0
.
1
1
0
9
/ICACCE4
6
6
0
6
.
2
0
1
9
.
9
0
7
9
9
9
8
.
[2
4
]
X.
Zh
a
n
g
,
Y.
C
u
i,
Y.
Wa
n
g
,
M
.
S
u
n
,
a
n
d
H.
Hu
,
“
An
imp
ro
v
e
d
A
E
d
e
tec
ti
o
n
m
e
th
o
d
o
f
ra
il
d
e
fe
c
t
b
a
se
d
o
n
m
u
lt
i
-
lev
e
l
AN
C
with
VSS
-
LM
S
,
”
M
e
c
h
a
n
ica
l
S
y
ste
ms
a
n
d
S
i
g
n
a
l
Pro
c
e
ss
in
g
,
v
o
l
.
9
9
,
p
p
.
4
2
0
-
4
3
3
,
2
0
1
8
,
d
o
i:
1
0
.
1
0
1
6
/j
.
y
m
ss
p
.
2
0
1
7
.
0
6
.
0
2
9
.
[2
5
]
A.
E.
W.
G
e
o
rg
e
,
S
.
S
o
,
R.
G
h
o
s
h
,
a
n
d
K.
K.
P
a
li
wa
l,
“
Ro
b
u
stn
e
s
s
m
e
tri
c
-
b
a
se
d
tu
n
in
g
o
f
th
e
a
u
g
m
e
n
ted
Ka
lma
n
fil
ter
fo
r
t
h
e
e
n
h
a
n
c
e
m
e
n
t
o
f
sp
e
e
c
h
c
o
rru
p
ted
wit
h
c
o
l
o
u
re
d
n
o
ise
,
”
S
p
e
e
c
h
Co
mm
u
n
.
,
v
o
l.
1
0
5
,
p
p
.
6
2
-
7
6
,
2
0
1
8
,
d
o
i:
1
0
.
1
0
1
6
/j
.
s
p
e
c
o
m
.
2
0
1
8
.
1
0
.
0
0
2
.
[2
6
]
M
.
S
.
Ali,
“
Ca
sc
a
d
e
d
ri
p
p
le
c
a
rry
a
d
d
e
r
b
a
se
d
S
RC
S
A
fo
r
e
fficie
n
t
F
IR
fi
lt
e
r,
”
In
d
o
n
e
sia
n
J
o
u
rn
a
l
o
f
El
e
c
trica
l
En
g
i
n
e
e
rin
g
a
n
d
C
o
mp
u
ter
S
c
ien
c
e
,
v
o
l.
9
,
n
o
.
2
,
p
p
.
2
5
3
-
2
5
6
,
2
0
1
8
,
d
o
i:
1
0
.
1
1
5
9
1
/
ij
e
e
c
s.v
9
.
i
2
.
p
p
2
5
3
-
2
5
6
.
[2
7
]
M
.
A.
Isla
m
,
“
M
o
d
ifi
e
d
m
e
l
-
f
re
q
u
e
n
c
y
c
e
p
stra
l
c
o
e
fficie
n
ts
(M
M
F
CC)
in
r
o
b
u
st
tex
t
-
d
e
p
e
n
d
e
n
t
s
p
e
a
k
e
r
id
e
n
ti
fica
ti
o
n
,
”
2
0
1
7
4
t
h
In
ter
n
a
ti
o
n
a
l
C
o
n
fer
e
n
c
e
o
n
A
d
v
a
n
c
e
s
in
El
e
c
trica
l
En
g
i
n
e
e
rin
g
(I
CAE
E)
,
2
0
1
7
,
p
p
.
5
0
5
-
5
0
9
,
d
o
i:
1
0
.
1
1
0
9
/ICAEE
.
2
0
1
7
.
8
2
5
5
4
0
8
.
[2
8
]
A.
C.
O.
Re
d
d
y
a
n
d
K.
M
a
d
h
a
v
i,
“
Hie
ra
rc
h
y
b
a
se
d
firefly
o
p
t
i
m
ize
d
k
-
m
e
a
n
s
c
lu
ste
rin
g
fo
r
c
o
m
p
lex
q
u
e
sti
o
n
a
n
sw
e
rin
g
,
”
I
n
d
o
n
e
sia
n
J
o
u
rn
a
l
o
f
El
e
c
trica
l
En
g
in
e
e
rin
g
a
n
d
Co
mp
u
ter
S
c
ien
c
e
(IJ
EE
C
S
)
,
v
o
l
.
1
7
,
n
o
.
1
,
p
p
.
2
6
4
-
2
7
2
,
2
0
1
9
,
d
o
i:
1
0
.
1
1
5
9
1
/i
jee
c
s.v
1
7
.
i
1
.
p
p
2
6
4
-
2
7
2
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
Ro
y
Ru
d
o
lf
H
u
ize
n
re
c
e
iv
e
d
h
i
s
Ba
c
h
e
lo
r
o
f
En
g
in
e
e
ri
n
g
in
El
e
c
tri
c
a
l
En
g
i
n
e
e
rin
g
(
1
9
9
9
)
fro
m
Un
i
v
e
rsitas
S
e
m
a
ra
n
g
(US
M
)
S
e
m
a
ra
n
g
,
Ce
n
tral
Ja
v
a
.
M
a
ste
r
o
f
El
e
c
tri
c
a
l
E
n
g
in
e
e
rin
g
(2
0
0
6
)
a
n
d
D
o
c
to
r
o
f
C
o
m
p
u
ter
S
c
ien
c
e
(2
0
1
8
)
fro
m
U
n
iv
e
rs
it
a
s
G
a
d
jah
M
a
d
a
(UG
M
)
Yo
g
y
a
k
a
rta,
In
d
o
n
e
sia
.
Lec
tu
re
r
a
n
d
re
se
a
rc
h
e
r
a
t
t
h
e
De
p
a
rtme
n
t
o
f
El
e
c
tri
c
a
l
E
n
g
i
n
e
e
rin
g
,
Un
iv
e
rsitas
S
e
m
a
ra
n
g
(USM
)
fr
o
m
2
0
0
0
t
o
2
0
0
8
,
a
n
d
a
t
t
h
e
F
a
c
u
lt
y
o
f
In
f
o
rm
a
ti
c
s
a
n
d
Co
m
p
u
ters
,
I
n
stit
u
t
Te
k
n
o
lo
g
i
d
a
n
Bi
sn
is
S
TKOM
Ba
li
,
fro
m
2
0
0
8
t
o
t
h
e
p
re
se
n
t.
Re
se
a
rc
h
in
tere
st i
n
si
g
n
a
l
p
ro
c
e
ss
in
g
a
n
d
d
ig
it
a
l
f
o
re
n
sic
s.
Fl
o
r
e
n
tin
a
Ta
tr
in
K
u
r
n
i
a
ti
re
c
e
iv
e
d
th
e
Ba
c
h
e
lo
r
o
f
E
n
g
in
e
e
rin
g
d
e
g
re
e
in
C
iv
il
En
g
i
n
e
e
rin
g
,
fro
m
Un
iv
e
rsitas
S
e
m
a
ra
n
g
(USM
),
Ce
n
tral
Ja
v
a
,
In
d
o
n
e
sia
(2
0
0
0
)
a
n
d
M
a
ste
r
o
f
En
g
i
n
e
e
rin
g
in
In
f
o
rm
a
ti
c
s
fro
m
Un
iv
e
rsitas
Atm
a
Ja
y
a
Yo
g
y
a
k
a
rta
(UA
JY
),
Yo
g
y
a
k
a
rta,
In
d
o
n
e
sia
(
2
0
1
5
).
S
i
n
c
e
2
0
0
8
s
h
e
h
a
s
b
e
e
n
a
lec
tu
re
r
a
n
d
re
s
e
a
rc
h
e
r
a
t
th
e
F
a
c
u
lt
y
o
f
in
fo
rm
a
ti
c
s
a
n
d
c
o
m
p
u
ter,
In
st
it
u
t
Te
k
n
o
lo
g
i
d
a
n
Bisn
is
S
TKO
M
Ba
li
,
In
d
o
n
e
sia
.
S
h
e
is
in
tere
ste
d
in
a
d
a
p
ti
v
e
n
o
ise
c
a
n
c
e
ll
a
ti
o
n
,
p
a
tt
e
r
n
re
c
o
g
n
it
i
o
n
a
n
d
d
i
g
it
a
l
fo
re
n
sic
s.
Evaluation Warning : The document was created with Spire.PDF for Python.