Information Technology Reference
In-Depth Information
Λ=
()
y
P
(),
nn
h
y
y R
2
. (4)
2
nZ
0 CC
<≤<∞
(ii) There exist constants
such that, for any constant matrix se-
1
2
{ n nZ
P
qucnce
, we have,
2
h
(5)
CP Py y CP
{ }
() ()
{ }
1
n
n
n
2
n
2
nZ
2
{}
P
{ n nZ
P
Where
denotes the norm of sequence
2
hD in space
(), ()
x
x
We consider any functions
LR with the inner product:
2
2
()
, $
() x
−⋅
ω
h
()
ω
=
h
xe dx
Dh
,
=
R xxx
D h
( )
( )
.
2
2
R
As usual, $ ()
h
ω
h
() LR
x
2
2
()
denotes the Fourier transform of any function
.
2 Vector-Valued Multiresolution Analysis
Λ
Definition 3. Let
Ω
be a separable Hilbert space and
be an index set. We recall
{
}
that a sequence
ι Γ∈Λ⊆Ω is a frame for
Ω
if there exist two positive constants
A ,
A such that
2
2
2
∀∈Ω
ξ
,
A
ξ
ξ
,
Γ ≤
A
ξ
, (6)
1
ι
2
ι
∈Λ
A ,
A are called frame bounds. A sequence
2
where
, and
ξ
=
ξ ξ
,
h is
an exact frame if it ceases to be a frame when any one of its elements is removed.
If
{:
h
ι ι∈Λ ⊆Ω
}
AA
=
. A frame {:
ι ι∈Λ
}
is a tight one if we can choose
1
2
AA
==
1
, then it follows from (1) that
,
1
2
∀∈Ω
ξ
=
ξ
,
Γ Γ
ι
ι
ι
∈Λ
A sequence {
ι Γ∈Λ⊆Ω
:
}
is a Bessel sequence if (only) the upper inequality of
()
()
1 follows. If only for all
h
∈⊂Ω
U
1 holds, the se-
, the upper inequality of
quence {:
ι Γ∈Λ⊆Ω
}
is a Bessel sequence with respect to (w.r.t.) U . More-
over, we assume that U
⊂Ω
D
U
is a closed subspace, if for all
, there exist two
2
2
A ,
A
>
0
2
A
D
|
D
,
Γ ≤
|
A
D
real numbers
such that
, the se-
2
1
ι
2
ι∈Λ
quence {:
is called an U − subspace frame. In this section, we intro-
duce the notion of vector-valued multiresolution analysis.
Γ∈Λ
ι ι
}
⊆Ω
Definition 4. A vector-valued multiresolution analysis of
LRC is a nested se-
2
(,
s
)
quence of closed subspaces {
X
}
such that (i)
XX j Z
+
1 ,
;
(ii)
j
j Z
j
j
I
U
LRC , where O denotes ss
×
jZ XO
=
{}
and
jZ X
is dense in
2
(,
s
)
zero
j
j
h
()
yX yX j Z
+
h
( )
,
∈⇔ ∈ ∀∈
matrix; (iii)
; (iv) there is a vector-valued
j
j
1
Search WWH ::




Custom Search