T
E
L
KO
M
NI
K
A
,
V
ol
.
14,
N
o.
3,
S
ept
em
ber
20
16,
pp.
97
4~
9
80
I
S
S
N
:
1
693
-
6
930
,
ac
c
r
edi
t
ed
A
b
y
D
IK
T
I,
D
e
c
r
e
e
N
o
:
58/
D
I
K
T
I
/
K
ep/
2013
D
O
I
:
10.
12928/
T
E
LK
O
M
N
I
K
A
.
v
1
4
i
3
.
3021
97
4
R
ec
ei
v
ed
N
ov
e
mber
1
2
,
20
1
5
;
R
ev
i
s
ed
May
9
,
20
1
6
;
A
c
c
ept
ed
M
ay
28
,
201
6
Co
m
p
r
essi
v
e
S
en
sin
g
A
l
g
o
r
it
h
m
f
o
r
Dat
a
Co
m
p
r
essio
n
on W
e
a
the
r
M
o
ni
tor
i
ng S
y
s
te
m
R
i
ka S
u
st
i
k
a
*
1
,
B
a
m
b
a
n
g
S
u
g
i
a
r
to
2
R
es
ear
c
h C
e
nt
er
f
or
I
nf
or
m
at
i
c
s
,
I
ndo
nes
i
an
I
ns
t
i
t
u
t
e of
S
c
i
e
nc
e
s
(
LI
P
I
)
Jl
.
C
i
s
i
t
u
21/
154D
,
B
and
ung 4
0135,
I
ndo
ne
s
i
a
*
C
or
r
es
po
ndi
ng a
ut
hor
,
e
-
ma
i
l
:
r
i
k
a@
i
nf
or
m
at
i
k
a.
l
i
p
i
.
g
o.
i
d
1
,
b
am
ba
ng@
i
n
f
or
m
at
i
k
a.
l
i
pi
.
go.
i
d
2
A
b
st
r
act
C
om
pr
es
s
i
v
e
s
en
s
i
ng
(
C
S
)
i
s
new
dat
a
a
c
qui
s
i
t
i
on
a
l
gor
i
t
h
m
t
hat
c
an
be
u
s
ed
f
or
c
om
pr
es
s
i
on.
C
S
t
heor
y
c
er
t
i
f
i
e
s
t
h
at
s
i
g
nal
s
c
an
be
r
e
c
ov
er
ed
f
r
om
f
ew
er
s
am
pl
e
s
t
han N
y
qu
i
s
t
r
at
e
.
O
n
t
hi
s
p
aper
,
t
he
c
om
pr
es
s
i
v
e
s
e
ns
i
ng
t
ec
hni
qu
e
i
s
a
ppl
i
ed
f
or
dat
a
c
om
pr
e
s
s
i
on
o
n
o
ur
w
e
at
her
m
oni
t
or
i
n
g
s
y
s
t
em
.
O
n
t
hi
s
w
eat
her
m
oni
t
or
i
n
g s
y
s
t
em
,
c
om
pr
es
s
i
o
n us
i
ng c
om
p
r
es
s
i
v
e
s
en
s
i
n
g w
i
t
h f
ew
e
r
s
am
pl
es
o
r
m
eas
ur
em
ent
s
m
ea
ns
m
i
ni
m
i
z
i
ng
s
ens
i
ng
and
ov
er
al
l
e
ne
r
gy
c
os
t
.
O
ur
f
oc
u
s
on
t
hi
s
p
aper
l
i
e
s
i
n
t
h
e
s
e
l
ec
t
io
n
of
m
a
t
r
ix
f
o
r
r
epr
e
s
ent
at
i
o
n ba
s
i
s
un
der
w
h
i
c
h
t
he w
e
at
her
dat
a ar
e
s
p
ar
s
el
y
r
epr
e
s
ent
ed
.
R
es
ul
t
s
f
r
om
s
i
m
ul
at
i
on
s
how
t
hat
t
he
us
i
ng
of
D
C
T
(
D
i
s
c
r
et
e C
os
i
ne
T
r
an
s
f
or
m
)
as
r
epr
e
s
ent
at
i
o
n b
as
i
s
has
a b
et
t
er
per
f
or
m
an
c
e o
n
w
eat
her
d
at
a r
e
c
ov
er
y
c
om
par
ed w
i
t
h ot
h
er
t
r
an
s
f
or
m
m
et
hod
s
s
u
c
h a
s
W
al
s
h
-
H
adam
ar
d
T
r
an
s
f
or
m
(
W
H
T
)
and D
i
s
c
r
et
e
W
av
el
e
t
T
r
ans
f
or
m
(
D
W
T
)
.
Ke
y
w
o
rd
s
:
c
om
pr
es
s
i
v
e
s
en
s
i
ng,
D
C
T
,
r
e
pr
es
ent
a
t
i
o
n ba
s
i
s
,
w
eat
her
m
o
ni
t
or
i
ng
C
o
p
y
r
i
g
h
t
©
20
16 U
n
i
ver
si
t
a
s A
h
mad
D
ah
l
an
.
A
l
l
r
i
g
h
t
s r
eser
ved
.
1
.
I
n
tr
o
d
u
c
ti
o
n
W
e
at
her
m
oni
t
or
i
ng
p
l
a
y
s
an
i
m
por
t
ant
r
ol
e
i
n
h
um
an
l
i
f
e.
W
eat
her
m
oni
t
or
i
n
g s
y
s
t
em
ai
m
s
t
o c
ol
l
ec
t
and des
c
r
i
be t
he s
t
at
e of
t
he w
eat
h
er
on one r
egi
on,
s
uc
h a
s
t
e
m
per
at
ur
e,
hum
i
di
t
y
,
pr
es
s
ur
e,
r
ai
n f
al
l
,
s
ol
ar
r
adi
at
i
on,
pr
ec
i
p
i
t
at
i
on
,
w
i
n
d di
r
ec
t
i
on an
d
s
peed,
et
c
.
O
bj
ec
t
i
v
e
of
w
e
at
her
m
oni
t
or
i
ng
ar
e
pr
o
v
i
de
w
eat
her
or
c
l
i
m
at
e
d
at
a,
ho
ur
l
y
,
d
ai
l
y
,
or
m
ont
hl
y
.
T
hes
e w
eat
her
and c
l
i
m
at
e dat
a c
an be us
e
d f
or
a
v
ar
i
et
y
of
us
es
,
s
uc
h as
ear
l
y
w
ar
ni
ng
s
y
s
t
em
on
t
er
r
es
t
r
i
al
on m
ar
i
ne
.
D
at
a f
r
o
m
w
ea
t
her
m
oni
t
or
i
n
g s
y
s
t
em
c
an be
us
ed
a
l
s
o t
o
ev
a
l
u
at
e
c
l
i
m
at
e pat
t
er
ns
a
nd l
ong
t
er
m
s
f
or
c
as
t
s
.
W
e
at
her
m
oni
t
or
i
n
g s
y
s
t
e
m
c
ons
i
s
t
s
o
f
a nu
m
ber
o
f
w
eat
h
er
s
t
at
i
ons
.
O
ne
w
eat
h
er
s
t
at
i
o
n has
s
om
e w
e
at
h
er
s
ens
or
s
t
o m
eas
ur
e and
r
ec
or
d w
eat
her
p
ar
am
et
er
s
.
E
x
am
pl
e of
w
eat
h
er
m
oni
t
or
i
ng s
y
s
t
em
c
onf
i
gur
at
i
on f
r
o
m
one w
eat
h
er
s
t
at
i
o
n c
an be s
ee
n on F
i
g
ur
e 1.
W
e
at
her
par
am
et
er
s
t
hat
m
eas
ur
ed
i
n
a
w
e
at
h
er
s
t
at
i
on
c
a
n
be
s
t
or
ed
i
n
a
bui
l
t
i
n
d
at
a
l
og
ge
r
or
t
r
ans
m
i
t
t
ed
t
o
a
bas
e
s
t
at
i
o
n
on
r
em
ot
e
l
oc
at
i
o
n
u
s
i
ng
a
c
om
m
uni
c
at
i
on
l
i
nk
.
I
f
t
he
dat
a
ar
e
s
t
or
ed
i
n
a d
at
a
l
o
gg
er
,
w
e c
an
do
w
nl
oad
t
h
em
t
o a
c
om
put
er
at
a
l
at
er
t
i
m
e i
f
w
e
ne
ed f
or
f
ur
t
her
pr
oc
es
s
i
ng [
1]
.
F
ig
ur
e
1
.
T
he c
onf
i
g
ur
at
i
on
of
t
he
w
e
at
her
m
oni
t
or
i
ng
s
y
s
t
em
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NI
K
A
I
S
S
N
:
1
693
-
6
930
C
ompr
es
s
i
v
e
S
e
ns
i
n
g A
l
g
o
r
i
t
hm
f
or
D
at
a
C
o
mpr
es
s
i
on
on W
eat
her
Mon
i
t
or
i
ng
…
(
R
i
ka
S
u
st
i
ka
)
975
T
her
e
i
s
a
pr
obl
em
on t
hi
s
w
e
at
h
er
m
oni
t
or
i
ng
s
y
s
t
e
m
.
D
at
a
t
hat
ar
e
c
ol
l
ec
t
e
d
f
r
o
m
a
num
ber
of
s
ens
or
nodes
ar
e s
o hu
ge
but
t
he s
y
s
t
em
c
ons
t
r
ai
ned b
y
l
i
m
i
t
ed
po
w
er
s
u
ppl
y
,
m
e
m
o
r
y
,
pr
oc
es
s
i
ng
per
f
or
m
anc
e,
and c
om
m
uni
c
at
i
on ba
nd
w
i
dt
h.
T
hr
ee op
er
at
i
o
ns
m
ai
nl
y
r
es
pons
i
b
l
e
f
or
ener
g
y
c
on
s
u
m
pt
i
on
ar
e
s
ens
i
n
g
or
s
a
m
pl
i
ng,
c
om
put
at
i
o
n,
an
d
c
om
m
uni
c
at
i
on
[
2]
.
W
hen dat
a
t
r
ans
m
i
t
t
ed o
n t
h
e n
et
w
or
k
,
ener
g
y
c
o
ns
um
pt
i
on i
s
dom
i
n
at
ed b
y
r
ad
i
o
c
o
m
m
uni
c
at
i
on
.
T
he
ener
g
y
c
o
ns
um
pt
i
on
of
r
adi
o
c
om
m
uni
c
at
i
on
i
s
di
r
ec
t
l
y
pr
o
por
t
i
o
na
l
t
o
t
he
num
ber
of
bi
t
s
of
dat
a
t
hat
t
r
ans
m
i
t
t
ed
o
n
t
h
e
n
et
w
or
k
[
2]
.
C
om
pr
es
s
i
on
i
s
a
k
ey
t
o
o
v
er
c
om
e
t
hi
s
pr
o
bl
em
.
T
he m
ai
n obj
ec
t
i
v
e
of
c
o
m
pr
es
s
i
on on t
hi
s
s
y
s
t
em
i
s
t
o r
educ
e
ener
g
y
c
ons
um
pt
i
on b
y
r
ed
uc
i
n
g n
um
ber
of
bi
t
s
t
hat
n
eed
t
o
p
r
oc
es
s
and t
r
ans
m
i
t
on t
he
s
y
s
t
em
.
A
l
ar
ge
num
ber
of
c
om
pr
es
s
i
on t
ec
hn
i
qu
es
h
av
e b
een
pr
op
os
ed
i
n
t
h
e
l
i
t
er
at
ur
es
[
2
,
3
].
O
ne
of
c
o
m
pr
es
s
i
on
t
ec
hni
qu
e
t
h
at
i
n
t
er
es
t
i
ng
f
or
m
an
y
r
es
ear
c
her
s
no
w
ada
y
s
i
s
c
o
m
pr
es
s
i
v
e
s
ens
i
ng
t
ec
hn
i
qu
e.
C
om
pr
es
s
i
v
e
s
e
ns
i
ng
i
s
ne
w
ap
pr
oac
h
i
n
s
i
gn
al
pr
oc
es
s
i
ng,
par
t
i
c
u
l
ar
l
y
f
or
dat
a
ac
qu
i
s
i
t
i
on
[
4]
.
O
n
t
r
adi
t
i
on
al
dat
a
ac
qu
i
s
i
t
i
on
,
s
i
gna
l
c
an
be
r
ec
o
ns
t
r
u
c
t
ed
f
r
o
m
t
he
s
am
p
l
e
s
t
hat
ar
e
t
ak
i
ng
at
a
r
at
e
gr
ea
t
er
t
ha
n
N
y
q
u
i
s
t
r
at
e
(
2x
f
m
a
x
)
.
O
n
c
om
pr
es
s
i
v
e
s
ens
i
n
g,
a
s
i
gna
l
c
an be r
ec
ov
er
ed f
r
om
f
ar
f
ew
er
s
am
pl
es
t
han N
y
qu
i
s
t
r
at
e,
as
l
o
ng as
t
he s
i
gna
l
i
s
s
par
s
e or
ap
pr
ox
i
m
at
el
y
s
p
ar
s
e.
I
t
w
i
l
l
o
nl
y
pr
oc
es
s
t
h
e l
ar
ge c
oef
f
i
c
i
ent
s
and
di
s
r
egar
d t
he z
er
o
c
oef
f
i
c
i
ent
s
[6
,
7]
.
T
h
is
c
om
pr
es
s
i
on
al
gor
i
t
hm
r
educ
es
t
he
s
i
z
e
of
dat
a
s
ent
an
d
dec
r
eas
es
t
h
e
s
t
or
age r
equ
i
r
em
ent
of
t
h
e s
y
s
t
em
.
T
hi
s
w
i
l
l
a
l
s
o l
e
s
s
en t
he dat
a t
r
ans
m
i
s
s
i
on ac
t
i
v
i
t
y
.
T
hi
s
c
r
i
t
er
i
a
i
s
s
ui
t
abl
e f
or
c
om
pr
es
s
i
on t
y
p
e
t
h
at
i
s
nee
ded
i
n
w
ea
t
her
m
oni
t
or
i
ng s
y
s
t
e
m
.
O
n
t
h
i
s
pa
per
,
w
e
e
v
a
l
u
at
e
c
om
pr
es
s
i
v
e
s
e
ns
i
n
g
al
g
or
i
t
hm
f
or
i
m
pl
e
m
ent
at
i
on
o
n
w
eat
h
er
m
oni
t
or
i
n
g s
y
s
t
e
m
.
B
y
us
i
ng
c
om
pr
es
s
i
v
e
s
ens
i
ng,
num
ber
of
m
eas
ur
em
ent
c
an be
r
educ
ed a
nd num
ber
of
w
e
at
her
da
t
a t
ha
t
be c
ol
l
ec
t
e
d
bec
om
e
f
ew
er
,
w
i
t
hou
t
s
ac
r
i
f
i
c
i
ng qu
al
i
t
y
of
t
he r
ec
ons
t
r
uc
t
e
d s
i
gna
l
.
O
ne
of
t
he pr
obl
em
s
on i
m
pl
em
ent
i
ng C
S
o
n
w
e
at
her
m
oni
t
or
i
n
g
s
y
s
t
em
i
s
w
e do
n’
t
k
no
w
t
he k
i
nd of
r
epr
es
ent
at
i
on
bas
i
s
und
er
w
hi
c
h t
h
e w
e
at
her
dat
a be
s
par
s
el
y
r
e
pr
es
ent
ed.
A
s
i
gn
a
l
m
a
y
n
ot
s
par
s
e
i
n
a
p
ar
t
i
c
ul
ar
dom
ai
n
(
l
i
k
e
t
i
m
e
dom
ai
n)
,
but
i
t
c
an b
e s
par
s
e
or
c
om
pr
es
s
i
bl
e
b
y
t
r
ans
f
or
m
i
ng i
t
t
o
s
o
m
e s
ui
t
abl
e b
as
i
s
.
T
her
e
i
s
no s
y
s
t
em
at
i
c
w
a
y
of
s
el
ec
t
i
n
g t
hi
s
k
i
nd of
m
at
r
i
x
.
W
e
c
an
k
now
t
h
e s
ui
t
a
bl
e bas
i
s
b
y
t
r
i
al
s
an
d ex
per
i
enc
e
[5
].
O
n t
hi
s
pap
er
,
w
e
e
v
al
uat
ed p
er
f
or
m
anc
e of
w
e
at
he
r
s
i
gna
l
r
ec
ons
t
r
uc
t
e
d us
i
n
g c
om
pr
es
s
i
v
e
s
ens
i
ng
w
hen
r
e
pr
es
ent
at
i
on
bas
i
s
i
s
us
i
ng
D
i
s
c
r
et
e
C
os
i
ne
T
r
ans
f
or
m
(
D
C
T
)
,
D
i
s
c
r
et
e
W
al
s
h
H
adam
ar
d T
r
ans
f
or
m
(
W
H
T
)
,
and D
i
s
c
r
et
e
W
av
el
et
T
r
ans
f
or
m
(
D
W
T
)
.
T
he
obj
ec
t
i
v
e
i
s
t
o
s
el
ec
t
w
hat
k
i
nd
of
r
epr
es
ent
at
i
o
n
bas
i
s
t
h
at
m
a
k
e s
i
gna
l
r
ec
ons
t
r
uc
t
i
o
n er
r
or
i
s
m
i
ni
m
um
.
2.
C
o
m
p
r
e
s
s
i
v
e
S
e
n
s
i
n
g
(C
S
)
C
S
is
an
al
t
er
nat
i
v
e
s
am
pl
i
ng
t
heor
y
w
hi
c
h
as
s
er
t
s
t
h
at
c
er
t
ai
n
s
i
g
nal
c
an
b
e
r
ec
ov
er
e
d
f
r
o
m
f
ar
f
ew
er
s
am
pl
es
t
han S
h
ann
on
-
N
y
qu
i
s
t
s
am
p
l
i
n
g us
es
[
3
,
6]
.
A
c
c
or
d
i
ng
t
o S
han
non
–
N
y
q
ui
s
t
t
heor
em
,
t
he s
am
pl
i
n
g
r
at
e
m
us
t
be at
l
eas
t
t
w
i
c
e
t
h
e m
ax
i
m
u
m
f
r
equen
c
y
pr
es
ent
i
n
t
he s
i
gn
al
[
6
,
8]
.
O
ne of
t
he k
e
y
of
t
he C
S
i
s
t
hat
t
h
e s
i
gna
l
i
s
s
par
s
e or
c
o
m
p
r
es
s
i
bl
e [
6]
.
A
s
i
gna
l
i
s
c
a
l
l
ed s
par
s
e
i
f
i
t
has
o
nl
y
a f
e
w
s
i
gn
i
f
i
c
ant
c
om
ponent
s
a
nd a
gr
e
at
er
num
ber
of
i
ns
i
g
ni
f
i
c
ant
c
om
ponent
s
.
C
S
on
l
y
f
oc
us
es
on
l
ar
ge
or
non
z
er
o c
o
ef
f
i
c
i
ent
.
A
di
s
c
r
et
e
s
i
g
na
l
of
l
engt
h
n
i
s
s
ai
d
t
o
be
-
s
par
s
e,
i
f
c
ont
ai
ns
(
at
m
os
t
)
non
z
er
o
ent
r
i
es
w
i
t
h
≪
.
A
s
i
g
na
l
m
a
y
not
l
ook
s
par
s
e
i
n
a
p
ar
t
i
c
u
l
ar
dom
ai
n,
b
ut
i
t
c
a
n
be
s
par
s
e
or
c
o
m
pr
es
s
i
bl
e
b
y
t
r
ans
f
or
m
i
ng
t
o
s
om
e
s
ui
t
ab
l
e
b
as
i
s
,
e.
g.
,
D
C
T
,
F
our
i
er
,
or
W
av
el
et
bas
i
s
.
F
or
m
an
y
na
t
ur
a
l
s
i
gn
al
s
t
h
er
e ar
e ad
equ
at
e b
as
es
an
d di
c
t
i
on
ar
i
es
i
n
w
h
i
c
h s
i
g
nal
s
of
i
nt
er
es
t
bec
om
e
s
par
s
e
or
ap
pr
ox
i
m
at
el
y
s
par
s
e.
A
s
i
gna
l
i
s
s
ai
d
t
o
b
e
c
om
pr
es
s
i
bl
e,
i
f
t
her
e
i
s
a
bas
i
s
i
n
w
hi
c
h
t
h
e s
i
gna
l
has
a
ppr
ox
i
m
at
el
y
s
par
s
e r
epr
e
s
ent
at
i
on
[
7]
.
S
i
g
na
l
of
i
nt
er
es
t
c
an b
e
ex
pr
es
s
ed i
n r
epr
es
e
nt
at
i
o
n bas
i
s
as
:
=
Ψ
(
1)
W
h
er
e
Ψ
i
s
t
r
ans
f
or
m
at
i
on bas
i
s
Ψ
= {
Ψ
1
,
Ψ
2
,
Ψ
3
,…
,
Ψ
N
}
a
nd
is
-
s
par
s
e v
ec
t
or
,
t
h
at
r
epr
es
ent
pr
oj
ec
t
i
o
n c
oef
f
i
c
i
ent
s
of
on
Ψ
.
A
m
ai
n i
dea
i
n t
h
e c
ur
r
ent
C
S
t
he
or
y
i
s
abou
t
ho
w
t
o ac
qui
r
e
a s
i
gn
al
.
T
he ac
qu
i
s
i
t
i
o
n of
s
ig
n
a
l
of
l
engt
h
i
s
done
b
y
m
eas
ur
i
ng
pr
oj
ec
t
i
ons
of
ont
o s
ens
i
ng v
ec
t
or
s
{
=
1
,
2
,
…
,
}
[
7]
.
O
n
t
he m
at
r
i
x
not
at
i
o
n,
t
h
e s
ens
i
ng pr
oc
es
s
i
s
d
es
c
r
i
bed
b
y
[6
-
8]
:
=
Φ
(
2)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SSN
:
1
6
9
3
-
6
930
T
E
L
KO
M
NI
K
A
V
o
l.
14
,
N
o
.
3,
S
ept
em
ber
2016
:
9
74
–
98
0
9
76
W
h
er
e
∈
i
s
t
he
s
i
g
na
l
t
o
be
s
ens
ed,
an
d
Φ
is
×
m
eas
ur
e
m
ent
m
at
r
i
x
,
and
∈
is
m
eas
ur
e
m
ent
v
ec
t
or
.
A
n
onl
i
n
ear
a
l
g
or
i
t
hm
i
s
u
s
ed i
n
C
S
at
r
ec
ei
v
er
s
i
de
t
o r
ec
ons
t
r
uc
t
or
i
g
i
n
al
s
i
g
nal
[
3]
.
T
hi
s
nonl
i
n
ear
r
ec
ons
t
r
uc
t
i
on
al
gor
i
t
hm
r
equi
r
es
k
no
w
l
e
dge
of
a r
e
pr
es
ent
at
i
on
bas
i
s
Ψ
.
Meas
ur
em
ent
v
ec
t
or
,
c
an
be ex
pr
es
s
ed
i
n
r
epr
es
en
t
a
t
i
on
bas
i
s
as
:
=
ΦΨ
(
3)
Θ
=
ΦΨ
is
×
di
m
ens
i
ona
l
r
ec
ons
t
r
uc
t
i
on
m
at
r
i
x
,
and
w
e
w
ou
l
d
h
av
e
:
=
Θ
S
(
4)
R
ec
ons
t
r
uc
t
i
on
al
gor
i
t
hm
i
n C
S
i
s
f
i
n
di
n
g
a s
p
ar
s
e
v
ec
t
or
S
s
at
i
s
f
i
ed
(
4)
ex
ac
t
l
y
or
appr
ox
i
m
at
el
y
w
i
t
h
gi
v
en
and
Θ
.
i
s
an
und
et
er
m
i
ne
d
l
i
n
ear
s
y
s
t
em
w
ho
h
as
m
or
e
unk
now
ns
t
ha
n e
quat
i
o
ns
.
F
i
nd
i
n
g a s
p
ar
s
e s
ol
ut
i
on
of
=
Θ
S
is
a
n
i
ll
-
p
os
ed
pr
ob
l
em
[
7]
.
S
ol
ut
i
on f
or
t
hi
s
pr
o
bl
em
i
s
done
us
ua
l
l
y
b
y
m
i
ni
m
i
z
i
ng
l
0
, l
1
,
or
l
2
n
or
m
ov
er
s
ol
ut
i
on s
pac
e.
O
n
t
hi
s
p
aper
,
w
e
us
e
l
1
m
i
ni
m
i
z
at
i
on
t
o
s
ol
v
e
t
he
pr
ob
l
em
.
T
he
c
om
pr
es
s
i
v
e
s
e
ns
i
ng
al
gor
i
t
hm
s
t
hat
r
ec
ons
t
r
uc
t
t
h
e s
i
gn
al
bas
ed on m
i
ni
m
i
z
i
ng l
1
r
e
f
er
r
ed t
o as
B
as
i
s
P
ur
s
u
i
t
(
B
P
)
[
10]
.
B
y
us
i
ng
l
1
m
i
ni
m
i
z
at
i
on
or
B
a
s
i
s
P
ur
s
u
i
t
,
s
i
gn
al
c
an
be
ex
ac
t
l
y
r
ec
o
v
er
ed
f
r
om
m
eas
ur
em
ent
s
b
y
s
o
l
v
i
n
g a s
i
m
pl
e c
o
nv
ex
opt
i
m
i
z
at
i
o
n pr
o
bl
em
t
hr
ough l
i
n
ear
pr
o
gr
am
m
i
ng [
10]
.
M
in
im
iz
e
‖
‖
1
; S
u
b
j
e
c
t to
:
ΦΨ
=
(
5)
O
nc
e t
he s
o
l
ut
i
o
n l
1
of
(
5)
i
s
f
ound,
r
ec
ons
t
r
uc
t
e
d s
o
l
ut
i
on
m
y
be ex
pr
es
s
ed as
:
∗
=
Ψ
∗
(
6)
3.
F
i
l
te
r
i
n
g
A
l
g
o
r
i
th
m
s
O
n t
hi
s
r
es
ear
c
h,
w
e o
nl
y
f
oc
us
on t
he m
oni
t
or
i
ng of
w
eat
h
er
s
i
gn
al
at
a s
i
ng
l
e
l
oc
at
i
o
n.
W
e
di
v
i
ded
our
s
i
m
ul
at
i
on
us
i
ng
t
w
o s
t
a
ges
,
s
i
gna
l
ac
qui
s
i
t
i
o
n a
nd s
i
gna
l
r
ec
o
ns
t
r
uc
t
i
on.
3.
1
.
S
ig
n
a
l
A
c
q
u
i
s
i
ti
o
n
O
n t
hi
s
w
e
at
h
er
m
oni
t
or
i
ng
s
y
s
t
em
,
s
i
gnal
ac
qu
i
s
i
t
i
on i
s
a pr
oc
es
s
f
or
s
a
m
pl
i
ng w
eat
h
er
s
i
gna
l
t
h
at
m
eas
ur
e
w
e
at
her
c
ond
i
t
i
on
par
am
et
er
s
on a
r
egi
on.
F
r
om
t
hi
s
s
i
gna
l
ac
q
ui
s
i
t
i
o
n
s
t
age,
w
e
w
i
l
l
ge
t
m
eas
ur
em
ent
v
ec
t
or
Y
t
h
at
w
i
l
l
b
e us
ed f
or
s
i
gna
l
r
ec
ons
t
r
uc
t
i
on
on t
h
e
r
ec
ei
v
er
s
i
de
.
T
he di
agr
am
on t
h
i
s
s
i
g
nal
ac
qu
i
s
i
t
i
on
s
t
age c
a
n be
s
een
on F
i
gur
e
2
[7
,
10]
.
F
i
gur
e
2.
S
i
g
na
l
ac
q
ui
s
i
t
i
o
n
s
t
age
S
t
eps
on t
hi
s
s
t
ag
e ar
e
[
7
,
10]
:
1.
R
ead t
he or
i
g
i
n
al
s
i
gna
l
(
)
f
r
om
w
i
r
el
es
s
w
e
at
her
m
oni
t
or
i
ng s
y
s
t
em
.
2.
G
ener
at
e
m
at
r
i
x
t
r
ans
f
or
m
(
D
C
T
,
W
H
T
,
or
D
W
T
)
f
or
s
par
s
i
f
y
i
n
g t
h
e s
i
g
na
l
.
3.
T
r
ans
f
or
m
t
he s
i
gna
l
us
i
ng
det
er
m
i
ned m
at
r
i
x
t
r
ans
f
or
m
.
4.
Mak
e c
oef
f
i
c
i
ent
t
hr
es
ho
l
d
i
ng o
n t
r
ans
f
or
m
do
m
ai
n f
or
s
par
s
i
t
y
en
hanc
em
ent
[
9]
.
C
oef
f
i
c
i
ent
s
s
m
al
l
er
t
ha
n
t
h
r
es
hol
d
i
n
g
v
al
ue
(
δ
)
s
et
t
o
z
er
o.
5.
A
pp
l
y
i
n
v
er
s
e of
m
at
r
i
x
t
r
ans
f
or
m
t
o t
he or
i
g
i
na
l
s
i
gna
l
t
o g
et
,
t
hat
s
par
s
e
i
n
t
r
ans
f
or
m
do
m
ai
n,
bu
t
no
t
s
par
s
e i
n
t
i
m
e dom
ai
n.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NI
K
A
I
S
S
N
:
1
693
-
6
930
C
ompr
es
s
i
v
e
S
e
ns
i
n
g A
l
g
o
r
i
t
hm
f
or
D
at
a
C
o
mpr
es
s
i
on
on W
eat
her
Mon
i
t
or
i
ng
…
(
R
i
ka
S
u
st
i
ka
)
977
6.
G
ener
at
e
m
eas
ur
e
m
ent
m
at
r
i
x
Φ
(
us
i
ng
r
an
dom
pr
oj
ec
t
i
on
m
at
r
i
x
)
,
and
t
h
en
g
et
t
he
m
eas
ur
e
m
ent
v
ec
t
or
.
T
hi
s
m
eans
t
hat
t
he m
eas
ur
em
e
nt
v
ec
t
or
w
as
ob
t
ai
ned
b
y
s
a
m
pl
i
n
g s
i
g
nal
r
andom
l
y
.
3.
2
.
S
i
g
n
a
l
R
e
c
o
n
s
t
r
u
c
ti
o
n
T
he s
i
gnal
r
ec
o
ns
t
r
uc
t
i
o
n
i
s
s
t
age t
o c
ons
t
r
uc
t
t
he
des
i
r
ed
w
eat
h
er
s
i
gn
al
f
r
om
m
eas
ur
e
m
ent
v
ec
t
or
.
B
l
oc
k
di
agr
am
of
t
hi
s
s
t
age c
an
be s
ee
n on
F
i
g
ur
e 3.
F
i
gur
e 3.
S
i
g
na
l
r
ec
ons
t
r
uc
t
i
on s
t
age
I
nput
s
f
or
r
ec
ons
t
r
uc
t
i
o
n a
l
gor
i
t
hm
ar
e m
eas
ur
e
m
ent
v
ec
t
or
,
m
eas
ur
e
m
ent
m
at
r
i
x
(
Φ
)
,
and r
e
pr
es
ent
at
i
on
bas
i
s
(
Ψ
)
.
T
he s
t
eps
on t
hi
s
s
t
ag
e ar
e:
1.
D
et
er
m
i
ne r
ec
ons
t
r
uc
t
i
on
m
at
r
i
x
(
Θ
)
f
r
o
m
k
now
n
Φ
and
Ψ
.
2.
F
i
nd
i
ng s
par
s
e v
ec
t
or
(
)
ex
ac
t
l
y
or
ap
pr
ox
i
m
at
el
y
w
i
t
h gi
v
en
an
d
Θ
,
us
i
ng
B
as
i
s
P
ur
s
ui
t
al
gor
i
t
hm
.
3.
R
ec
ons
t
r
uc
t
t
h
e
w
e
at
h
er
s
i
gna
l
f
r
o
m
gi
v
en
and
Ψ
.
.
4.
S
i
m
u
l
a
ti
o
n
R
e
s
u
l
t
T
hi
s
s
ec
t
i
on ev
a
l
uat
es
t
he
ef
f
ec
t
i
v
en
es
s
of
C
S
i
m
pl
em
ent
at
i
on on
w
e
at
h
er
m
oni
t
or
i
ng
s
y
s
t
em
.
F
or
t
he e
v
a
l
u
at
i
on
w
e
us
ed r
eal
da
t
as
et
w
i
t
hout
c
om
pr
es
s
i
on f
r
o
m
w
i
r
el
es
s
w
e
at
h
er
s
t
at
i
o
n t
hat
de
v
e
l
op
ed b
y
I
n
dones
i
an I
ns
t
i
t
ut
e of
S
c
i
e
nc
es
(
LI
P
I
)
.
T
he f
i
r
s
t
dat
as
et
i
s
f
or
hu
m
i
di
t
y
and t
h
e s
ec
ond on
e i
s
f
or
a
i
r
t
em
per
at
ur
e.
D
at
a t
ak
en on J
anuar
y
20
15 i
n
B
an
du
ng s
t
at
i
on.
T
he
dat
a
r
ead
o
nc
e
i
n
ev
er
y
t
w
o m
i
nut
es
.
O
n s
i
m
ul
at
i
on
,
w
e
us
ed
20
48
s
am
pl
es
.
P
i
c
t
ur
e
of
t
h
e
or
i
gi
na
l
s
i
g
na
l
s
am
pl
es
c
an
be s
ee
n o
n F
i
gur
e 4
.
(a
)
hum
i
di
t
y
(b
)
t
em
per
at
ur
e
F
i
gur
e 4.
O
r
i
g
i
n
al
w
eat
her
s
i
gna
l
F
or
C
S
r
ec
ons
t
r
uc
t
i
on
w
e
us
e
B
as
i
s
P
ur
s
ui
t
as
a
s
t
a
ndar
d
r
ec
ons
t
r
uc
t
i
on
al
gor
i
t
hm
,
b
y
us
i
ng
l
1m
agi
c
t
oo
l
b
ox
f
r
om
C
al
t
ec
h [
11
]
.
R
a
ndo
m
pr
o
j
ec
t
i
on m
at
r
i
x
w
as
us
ed f
or
t
h
e
m
eas
ur
e
m
ent
.
F
or
t
he
s
par
s
i
f
i
c
at
i
on,
w
e
e
v
al
uat
ed
t
hr
ee
r
epr
es
en
t
at
i
on
bas
i
s
,
D
i
s
c
r
et
e
C
o
s
i
ne
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SSN
:
1
6
9
3
-
6
930
T
E
L
KO
M
NI
K
A
V
o
l.
14
,
N
o
.
3,
S
ept
em
ber
2016
:
9
74
–
98
0
978
T
r
ans
f
or
m
(
D
C
T
)
,
D
i
s
c
r
et
e
W
al
s
h H
adam
ar
d T
r
a
ns
f
or
m
(
W
H
T
)
,
and D
i
s
c
r
et
e
W
av
el
et
T
r
ans
f
or
m
(
D
W
T
)
.
O
n w
a
v
e
l
et
t
r
a
ns
f
or
m
,
w
e us
e
d D
au
bec
hi
es
d4
w
av
el
et
(
db
4
w
av
e
l
et
)
.
R
es
ul
t
s
ar
e
pr
es
en
t
ed
on
t
w
o p
ar
t
s
.
T
he f
i
r
s
t
one
pr
es
ent
s
t
h
e s
par
s
i
t
y
a
na
l
y
s
i
s
of
hu
m
i
di
t
y
and
t
em
per
at
ur
e
dat
a.
T
hi
s
r
es
ul
t
i
s
us
e
d
f
or
det
er
m
i
ni
ng
num
ber
of
m
eas
ur
em
ent
or
m
eas
ur
e
m
ent
m
at
r
i
k
s
Φ
.
F
or
s
par
s
i
t
y
enh
anc
em
ent
,
w
e
us
ed
c
o
ef
f
i
c
i
ent
t
hr
es
hol
di
ng
on
t
r
ans
f
or
m
do
m
ai
n.
T
hr
es
ho
l
di
ng
v
a
l
u
e
(
δ
)
t
hat
i
s
us
e
d
on
t
h
i
s
s
i
m
ul
at
i
on
i
s
4.
S
p
a
rs
i
t
y
(r)
o
f
t
he
dat
a c
a
n b
e s
een
on T
ab
l
e
1
.
T
abl
e 1.
D
at
a s
p
ar
s
i
t
y
us
i
n
g di
f
f
er
ent
bas
i
s
T
r
an
sf
o
r
m
s
S
p
ar
si
t
y
(
r
)
H
um
i
di
t
y
T
em
p
er
at
u
r
e
DCT
99
20
W
HT
141
37
DW
T
111
25
F
r
o
m
T
abl
e 1 w
e c
an s
ee
t
hat
D
C
T
s
par
s
i
f
i
es
t
he dat
a bet
t
er
t
han ot
her
t
r
ans
f
or
m
s
.
N
um
ber
o
f
m
eas
ur
e
m
ent
m
i
s
t
a
k
en bas
ed on t
hi
s
s
par
s
i
t
y
a
na
l
y
s
i
s
.
F
r
om
T
abl
e
1
,
ma
x
i
mu
m
s
par
s
i
t
y
i
s
14
1.
B
as
ed
on
pr
ev
i
ous
r
es
e
ar
c
h,
num
ber
of
m
eas
ur
em
ent
m
m
u
s
t
be h
i
gh
er
or
equa
l
w
i
t
h
4
×
[3
,
7]
.
B
as
e
d on
t
hi
s
a
na
l
y
s
i
s
,
w
e t
ak
e 600
as
num
ber
of
m
eas
ur
e
m
ent
.
T
he
s
ec
ond
par
t
of
t
he
r
es
ul
t
s
i
nc
l
u
des
t
h
e
per
f
or
m
a
nc
e
of
C
S
o
n
r
ec
o
v
er
i
ng
w
eat
h
er
dat
a
us
i
ng
D
C
T
,
W
H
T
and
D
W
T
as
r
epr
es
ent
at
i
o
n
bas
i
s
on
s
par
s
i
f
i
y
i
ng
w
eat
h
er
s
i
gna
l
.
R
ec
ov
er
ed
s
i
g
na
l
as
r
es
ul
t
s
f
r
o
m
s
i
m
ul
at
i
o
n c
a
n
be
s
een
o
n F
i
g
ur
e 5
a
nd F
i
gur
e
6.
F
r
om
t
he
pi
c
t
ur
es
,
w
e
c
an
s
e
e
t
hat
w
hen
c
om
par
i
n
g
f
i
gur
es
of
t
he
r
ec
o
ns
t
r
uc
t
ed
s
i
gna
l
s
w
i
t
h
t
he
or
i
g
i
n
al
s
i
gna
l
,
t
he r
ec
ons
t
r
uc
t
e
d s
i
gna
l
i
s
al
m
os
t
c
oi
nc
i
d
e w
i
t
h
t
he or
i
gi
na
l
s
i
gna
l
ev
en t
h
ough t
h
er
e
ar
e
s
o
m
e er
r
or
s
.
I
t
s
m
ean t
h
at
w
e
c
an
r
ec
ov
er
s
i
gna
l
w
i
t
h s
m
al
l
n
um
ber
of
m
eas
ur
em
ent
or
s
a
m
pl
i
n
g r
at
e.
H
o
w
m
uc
h
er
r
or
bet
w
e
en or
i
gi
na
l
s
i
g
nal
a
nd r
ec
o
ns
t
r
uc
t
ed s
i
g
n
al
i
s
t
o
l
er
a
bl
e
depe
nd
on t
he p
ur
pos
e
of
t
he
w
e
at
h
er
m
oni
t
or
i
n
g s
y
s
t
em
,
ar
e t
he da
t
a c
r
i
t
i
c
al
or
not
.
(a
)
hum
i
di
t
y
(b
)
t
em
per
at
ur
e
F
i
gur
e
5.
R
ec
ons
t
r
uc
t
ed
w
e
at
her
s
i
gna
l
T
he
r
es
ul
t
s
on
F
i
gur
e
5
al
s
o
s
ho
w
t
h
at
t
hr
e
e
r
ep
r
es
ent
at
i
on
b
as
i
s
t
hat
us
ed
f
or
s
par
s
i
f
y
i
ng
t
he s
i
gn
al
gi
v
e
di
f
f
er
ent
per
f
or
m
anc
e on s
i
gna
l
r
ec
ons
t
r
uc
t
i
on.
E
r
r
or
b
et
w
ee
n or
i
gi
na
l
s
i
gna
l
and
r
ec
ons
t
r
uc
t
e
d
s
i
gna
l
us
i
ng
c
om
pr
es
s
i
v
e
s
ens
i
ng
a
l
gor
i
t
hm
w
as
c
o
unt
e
d a
nd
w
e
s
u
m
m
ar
i
z
ed
i
t
us
i
ng R
oot
Mean
S
quar
e
E
r
r
or
(
R
M
S
E
)
par
am
et
er
.
=
1
∑
(
(
)
)
2
=
0
,
ℎ
(
)
=
(
)
−
̂
(
)
(
7)
R
S
M
E
f
r
om
our
s
i
m
ul
at
i
o
n i
s
pr
es
ent
ed
on
T
abl
e 2
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NI
K
A
I
S
S
N
:
1
693
-
6
930
C
ompr
es
s
i
v
e
S
e
ns
i
n
g A
l
g
o
r
i
t
hm
f
or
D
at
a
C
o
mpr
es
s
i
on
on W
eat
her
Mon
i
t
or
i
ng
…
(
R
i
ka
S
u
st
i
ka
)
979
T
abl
e 2.
R
MS
E
of
s
i
gna
l
r
e
c
ons
t
r
uc
t
i
on
T
r
an
sf
o
r
m
s
RM
S
E
H
um
i
di
t
y
T
em
p
er
at
u
r
e
DCT
0.
8155
0.
3654
W
HT
0.
9592
0.
4319
DW
T
1.
5856
0.
3828
T
he r
es
ul
t
s
ar
e
a
l
s
o
pr
es
en
t
ed
on
a
di
a
gr
am
on F
i
gur
e
6.
T
hi
s
f
i
g
ur
e s
h
o
w
s
t
h
at
i
n
c
as
e
of
hu
m
i
di
t
y
s
i
gna
l
,
s
i
gn
al
r
ec
ons
t
r
uc
t
i
on us
i
ng D
C
T
as
r
epr
es
ent
at
i
on b
as
i
s
s
ho
w
s
t
he bes
t
per
f
or
m
anc
e,
and
D
W
T
i
s
t
he
w
or
s
t
.
O
n t
em
per
at
ur
e
s
i
g
na
l
,
D
C
T
al
s
o s
h
o
w
s
t
he
bes
t
per
f
or
m
anc
e.
O
n
t
h
i
s
t
em
per
at
ur
e s
i
gna
l
,
W
H
T
i
s
t
he
w
or
s
t
.
F
i
gur
e
6.
R
MS
E
of
s
i
g
nal
r
ec
ons
t
r
uc
t
i
o
n
B
y
c
om
par
i
ng t
he
s
i
m
ul
at
i
o
n r
es
ul
t
s
of
hum
i
di
t
y
dat
a a
nd t
em
per
at
ur
e d
at
a,
w
e
c
an s
e
e
t
hat
t
he
r
ec
ov
er
y
of
t
em
per
at
ur
e
d
at
a
i
s
m
uc
h
c
l
os
er
t
o
t
h
e
or
i
g
i
na
l
s
i
gna
l
or
er
r
or
r
ec
ov
er
y
i
s
s
m
al
l
er
t
h
an
hum
i
di
t
y
da
t
a.
T
hat
i
s
bec
a
us
e
r
epr
es
e
nt
at
i
o
n
b
as
i
s
us
i
ng
D
C
T
i
s
m
or
e
s
ui
t
ed
f
or
r
ec
ov
er
i
ng s
i
gn
al
l
i
k
e t
em
per
at
ur
e s
i
gna
l
.
E
v
en
t
he r
e
c
ons
t
r
uc
t
ed s
i
gna
l
of
hum
i
d
i
t
y
dat
a c
an
not
per
f
ec
t
l
y
s
am
e w
i
t
h
t
he
or
i
g
i
na
l
s
i
g
na
l
,
t
h
e o
v
er
a
l
l
r
ec
o
v
er
i
ng
i
s
not
ba
d.
6
.
C
o
n
c
l
u
s
i
o
n
I
n t
hi
s
pa
per
,
C
S
has
i
m
pl
em
ent
ed f
or
dat
a c
o
m
pr
es
s
i
on on
w
e
at
h
er
m
oni
t
or
i
ng s
y
s
t
em
.
T
hi
s
C
S
a
l
g
or
i
t
hm
i
s
us
e
d
t
o
r
e
duc
e
t
he
n
um
ber
of
s
a
m
pl
es
r
eq
ui
r
e
d t
o
r
ec
o
v
er
t
he
s
am
pl
ed
s
i
gna
l
an
d
s
o
r
educ
e
t
he
e
ner
g
y
c
o
ns
um
pt
i
on
of
t
he
s
ens
or
i
n
b
ot
h
s
am
pl
i
n
g
an
d
t
r
ans
m
i
s
s
i
on.
W
e
s
how
C
S
c
an
r
educ
e t
h
e num
ber
of
s
am
pl
es
r
equi
r
ed t
o
be
t
ak
en i
n t
h
e
w
e
at
h
er
m
oni
t
or
i
ng
s
y
s
t
em
l
oc
at
ed
i
n
B
an
du
ng
I
n
don
es
i
a.
F
r
om
s
i
m
ul
at
i
o
n
w
e
pr
oof
ed
t
hat
D
C
T
i
s
t
he
bes
t
t
r
ans
f
or
m
c
odi
ng t
o
b
e i
m
pl
em
ent
ed as
r
epr
es
ent
at
i
on
bas
i
s
f
or
hum
i
di
t
y
a
nd t
em
per
at
ur
e
dat
a,
as
par
t
of
w
ea
t
her
dat
a
.
R
ef
er
en
ces
[1
]
P
S
us
m
i
t
ha
,
G
S
ow
m
y
abal
a.
D
es
i
gn
and I
m
p
l
em
ent
a
t
i
on
of
W
e
at
her
M
oni
t
or
i
n
g and
C
ont
r
ol
l
i
n
g
S
y
s
te
m
.
I
nt
er
n
at
i
onal
J
o
ur
na
l
of
C
om
put
er
A
p
p
li
c
at
i
on
s
.
201
4;
97(
3
)
:
19
-
22.
[2
]
MA
R
az
z
aque,
C
hr
i
s
B
l
ea
k
l
ey
,
S
i
m
o
n D
ob
s
on
.
C
om
pr
es
s
i
o
n i
n
W
i
r
el
es
s
S
en
s
or
N
e
t
w
or
k
s
:
A
S
ur
v
ey
and C
o
m
par
at
i
v
e E
v
al
u
at
i
o
n.
A
C
M
T
r
ans
a
c
t
i
o
ns
on
S
e
ns
or
N
et
w
or
k
s
.
20
13;
1
0(
1
)
.
[3
]
S
aad Q
,
R
ana M
B
,
W
a
f
a I
,
M
uqadd
as
n N
,
S
u
ngy
oun
g Le
e.
C
om
pr
e
s
s
i
v
e S
s
en
s
i
ng
:
F
r
o
m
T
heor
y
t
o
A
ppl
i
c
at
i
on
s
,
a S
ur
v
ey
.
J
our
na
l
of
C
om
m
uni
c
at
i
on
s
an
d N
et
w
or
k
s
.
201
3;
1
5(
5)
:
443
-
4
56.
[4
]
W
e
n
Y
aw
C
hung
,
J
o
c
el
y
n
F
l
or
es
C
i
l
l
av
er
de.
I
m
pl
em
ent
at
i
on of
C
om
pr
e
s
s
i
v
e
S
en
s
i
n
g
A
l
gor
i
t
hm
f
or
W
i
r
el
e
s
s
S
s
en
s
or
N
et
w
or
k
E
n
er
gy
C
on
s
er
v
at
i
on
.
I
nt
er
n
at
i
o
n
al
E
l
e
c
t
r
on
i
c
C
o
nf
er
e
nc
e
on
S
ens
or
and
A
ppl
i
c
at
i
on
s
.
20
14.
[5
]
X
i
aopei
W
u
,
M
i
ngy
an Li
u.
In
-
s
i
t
u
S
oi
l
M
oi
s
t
ur
e S
ens
i
ng:
M
eas
ur
em
ent
S
c
h
edu
l
i
n
g an
d
E
s
t
i
m
at
i
on
u
s
i
ng
C
o
m
p
r
es
s
i
v
e S
e
ns
i
ng
.
P
r
oc
e
edi
ngs
of
t
he
11t
h
I
nt
er
nat
i
on
al
C
o
nf
er
e
nc
e on
I
nf
or
m
at
i
on
P
r
oc
e
s
s
i
ng i
n S
en
s
or
N
et
w
or
k
s
.
20
12:
1
-
12.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SSN
:
1
6
9
3
-
6
930
T
E
L
KO
M
NI
K
A
V
o
l.
14
,
N
o
.
3,
S
ept
em
ber
2016
:
9
74
–
98
0
980
[6
]
E
m
m
a
nuel
C
a
ndè
s
,
M
i
c
hael
W
a
k
i
n
.
A
n I
nt
r
od
uc
t
i
on t
o
C
om
pr
es
s
i
v
e S
a
m
pl
i
ng.
I
E
EE Si
g
n
a
l
P
r
oc
e
s
s
i
ng M
aga
z
i
ne
.
2008
;
2
5(
2)
:
2
1
-
30.
[7
]
W
u
-
S
hen
g Lu.
C
om
pr
es
s
i
v
e
S
ens
i
ng a
nd S
p
ar
s
e S
i
gn
al
P
r
oc
e
s
s
i
ng
.
U
ni
v
e
r
s
i
t
y
of
V
i
c
t
or
i
a.
C
an
ada.
2010.
[8
]
J
i
an
hua Z
hou,
S
i
w
ang Z
ho
u,
Q
i
ang F
an.
M
at
hem
a
t
i
c
s
A
ppr
o
ac
h
i
n C
o
m
pr
e
s
s
e
d S
ens
i
ng
.
T
el
k
om
ni
k
a.
20
13;
11(
9)
:
543
5
-
5440.
[9
]
I
d
a
W
, Ta
ti
L
, A
n
dr
i
y
an B
S
,
H
endr
aw
an.
S
par
s
i
t
y
P
r
op
er
t
i
e
s
of
C
om
pr
es
s
i
v
e V
i
deo
S
am
p
l
i
n
g
G
ener
at
e
d by
C
oef
f
i
c
i
ent
T
hr
e
s
hol
di
n
g.
T
e
l
k
om
ni
k
a
.
20
14;
1
2(
4)
:
8
97
-
90
4.
[
10]
S
hul
i
n
y
an,
C
hao
W
u
,
W
e
i
D
ai
,
M
ous
t
af
a
G
hane
m
,
Y
i
k
e
G
uo
.
E
n
v
i
r
onm
ent
a
l
M
oni
t
or
i
n
g v
i
a
C
om
pr
es
s
i
v
e S
ens
i
ng.
P
r
oc
ee
di
ng
s
of
t
he S
i
x
t
h I
nt
er
nat
i
onal
W
o
r
k
s
hop
on K
n
ow
l
edge D
i
s
c
ov
er
y
f
r
om
S
ens
or
D
at
a.
201
2:
61
-
68
.
[
11]
J
us
t
i
n R
o
m
ber
g.
l1
-
m
agi
c
.
w
w
w
.
a
c
m
.
c
al
t
ec
h
.
ed
u
/
l1
m
ag
ic
.
Evaluation Warning : The document was created with Spire.PDF for Python.