1. Introduction
In a rapidly changing world, with unexpected outcomes, the scientific community has to make particular effort to provide a deeper knowledge and understanding of the reality and natural environment surrounding us. In this way, the adoption of new mathematical tools for a better treatment and study of the social, natural and physical phenomena and processes becomes essential. The framework of fractal interpolation makes it possible to enlarge and improve the classical methods of approximation theory. In previous papers, the author defined fractal functions constructed by means of iterated function systems (see, e.g, [
1,
2,
3]). These maps are fractal perturbations of arbitrary continuous functions defined on compact intervals. The new functions interpolate the original mappings on a set of nodes. This approach can be extended to the space of
p-integrable functions, defining the fractal analogues of standard maps in
. A scale vector provides the necessary flexibility to approximate a highly irregular or discontinuous function. If the scale is chosen properly, one can obtain fractal bases of the most used functional spaces, beginning from any basis of these sets. This is done by means of a suitable bounded operator,
, also known as fractal operator ([
1,
2,
3,
4,
5]), transforming systems of ordinary spanning families into their fractal perturbations. In the case of multivariate maps, this operator can no longer be applied to get necessary functions, and some additional tools are required. While it is true that fractal approximation is an active field of research currently, and there is an abundant bibliography about multivariate fractal interpolation functions (see, e.g., [
6,
7,
8,
9,
10,
11,
12,
13,
14,
15,
16,
17]), our approach has some specificities. One of them is that the functions proposed are products of perturbations of classical maps, and consequently they can be as close to them as desired. In this way, the current approach has two main advantages regarding other existing results. The first one is that the functional bases proposed are a generalization of any product basis (classical or not). This fact provides a wide spectrum of maps, in order to choose the optimum for a particular application, extending the analytical, geometric and dynamical possibilities. The second advantage is the addition of properties, unfeasible for the standard known functions, such as non-differentiability, providing irregular maps whose geometric complexity can be quantified by means of the fractal dimension of their traces for instance.
Although the results presented are deeply theoretical, the potential applications of this type of functions include, but are not restricted to, all the applications of the approximation theory and analysis. These are, for instance, Fourier analysis (used extensively in signal theory), bivariate interpolation and numerical analysis in general, study of chaotic systems, graphical design, mechanical engineering, etc. In particular, all the applications involving the standard functions as polynomial, trigonometric, etc. have their counterpart in this fractal field.
The mappings presented own all the advantages of the traditional functions because they include them as particular cases (taking the scale vectors equal to zero). They also provide new non-smooth geometric objects to model complex behaviors. We could mention as inconveniences, in non-smooth cases, the implicit character of their definition that hinders (though does not prevent) punctual evaluations, and the computational demands for an accurate graphical representation.
In this paper, we define fractal bases and fractal frames of , in order to approximate two-dimensional square-integrable maps whose domain is a rectangle. This is accomplished by means of the identification of with the tensor product space
The paper is organized as follows. Iin
Section 2, we introduce the fractal perturbation of a continuous function. In
Section 3, we define fractal frames and bases of
composed of products of fractal functions. We also obtain weaker families, as Bessel, Riesz and Schauder sequences. Additionally, we study some properties of the tensor product of the fractal operators previously mentioned, corresponding to each variable.
2. -Fractal Functions
In this section, we present the basics of the theory of fractal interpolation, initiated by M. Barnsley [
18] and developed further by many authors, for instance, M. A. Navascués [
2]. Consider a partition
of a real compact interval
, and a set of data
. Define subintervals
for
, and the following Iterated Function System (IFS):
, where the mappings
are such that
for some
. The maps
are continuous functions satisfying a Lipschitz condition in the second variable:
for
and
. Additionally,
must satisfy some join-up conditions:
for
. According to [
18] (Theorem 1) and [
19] (Theorem 2 of Section 6.2), the described IFS owns a unique attractor that is the graph of a continuous function
interpolating the given data. By definition,
g is the fractal interpolation function (FIF) of the IFS defined and it is unique satisfying the fractional equation
for
and
.
In this paper, we consider the following mappings:
where the scale factor
is such that
. The coefficients of
are
The functions
are defined by means of two continuous functions
f and
b,
, as
where
and
.
The fractal interpolation function
of this IFS was called by [
2] the
-fractal function of
f with respect to the partition
, the scale vector
, and the map
b.
To define an operator
given by
, the first author considered in [
2] another operator
L such that
If
is linear and bounded, with respect to the supremum norm
or with respect to
-norm
and
,
, then
is also linear and bounded regarding the respective norms.
The functional Equation (
2) satisfied by
is:
for
A consequence of this expression provides a bound of the distance between
f and
where
is defined as
.
Figure 1 represents the graph of an
-fractal function
with respect to the operator
, where
and
. The interval is
,
, the sampling is uniform and
.
Figure 2 shows the graph of
for the operator
, the maps
and
. The interval is
,
the sampling is uniform and
.
can be extended to
, and in this way we obtain fractal perturbations of
p-integrable functions ([
1]). For convenience, in this paper, we denote as
the operator extended to
and
represents the image of
represents the norm of the operator with respect to the mean square norm in
The operator
enjoys many important properties ([
1]). For instance, the following inequality holds:
where
I is the identity and
L is the extension to
of the corresponding operator of
If , then is injective and has a closed range.
If , then is an isomorphism.
If ,
Due to the last item, we can consider that the fractal maps are generalizations of any function.
3. Fractal Frames on the Rectangle
In this section, we analyze the spanning properties of the fractal functions on the rectangle where I and J are compact real intervals. For this purpose, we use the identification with the tensor product space
Let us recall the definition and properties of the tensor product of two Hilbert spaces
and
. One way of introducing the tensorial product is approached by means of linear operators ([
20,
21]). We consider spaces on the field R since we deal in general with spaces of real functions (although the extension to complex functions is straightforward). Thus,
Definition 1. Let , be separable real Hilbert spaces. Their tensor product is defined aswhere is an orthonormal basis of Remark 1. If is an orthonormal basis of then The sum is independent of the basis chosen in every space . The operator is the adjoint of
The space
is Hilbert with respect to the inner product:
that induces the norm
Remark 2. The notation is reserved for the norm of A as linear and bounded operator.
Let us define now the tensor product of two vectors
and
as the operator
where
. The main properties of this product are summarized as:
.
.
.
.
If is an orthonormal basis of and is an orthonormal basis of , then is an orthonormal basis of
Let
denote the dimension of the range of the operator
Let us define
and
then
and
The identification of
with
comes from the equality:
where
and
Thus,
is an integral operator whose kernel is
. Similarly,
has as kernel the sum
The following result can be read in ([
20], Th. 7.16)
Theorem 1. The identification of with the function extends uniquely to an isometric isomorphism of with whose inverse identifies with the operator
Let
The definition of the tensor product of
implies that:
Definition 2. A Riesz basis of a Hilbert space H is a system equivalent to an orthonormal basis , that is to say, there is an isomorphism such that for any m.
Definition 3. A Riesz sequence of a Hilbert space H is a Riesz basis for its closed span . If , then it is a Riesz basis.
In [
3], the following result is proved:
Theorem 2. If is an orthonormal basis of and α is a scale factor such that , then is a Riesz basis of .
Remark 3. This result is also true for a non-orthonormal basis since there is a topological isomorphism such that . With these hypotheses, is also an isomorphism and and are equivalent bases.
Let us recall now the tensor product of two linear operators.
Definition 4. If S is a linear and bounded operator of and T is a linear and bounded operator of thenis defined as:where is linear, bounded and satisfying . The main properties of this tensor product are:
is linear and bounded as operator of .
.
.
.
.
Remark 4. The notation represents the norm as operator of (with respect to in ).
Remark 5. is the operator that maps a function into . is the operator defined in (6) for the second variable and β is the scale vector in the y-direction. Theorem 3. Let and be Riesz bases of and respectively, and let us consider scale vectors α and β satisfying and , then is a Riesz basis of .
Proof. The tensor product is a Riesz basis. In this case, and are topological isomorphisms due to Theorem 2. Taking and in the properties 4 and 6 before Remark 4, we have that is an isomorphism. Then, is a basis equivalent to □
According to Property
of the tensor product of operators:
Moreover (Theorem 3.3 of [
1]),
and
Lemma 1. If then Proof. If
, the inequality (
9) and the condition on
imply that
Moreover, by Equation (
15),
hence (
16)
□
Remark 6. If or , the norm reaches the bound given in the previous Lemma since , improving the limit provided in Theorem 3.7 of the reference [1]. Definition 5. The constant C of a basis of a Banach space H iswhere is the Nth partial sum operator of the expansionfor any that is to say, defined by Proposition 1. In the conditions of Theorem 3, the constant of the basis satisfies the inequality:where C is the constant of the basis Proof. Let us consider
and the expansion of the element
defined as
where
represent the coefficient of
U with respect to the basis
Then,
If
is the
Nth partial sum of this expansion, then
where
is the
Nth projection with respect to the basis
The result follows from the properties of the tensor product of operators, the previous lemma and the inequalities (
13) and (
14). □
Definition 6. A sequence , where H is a Hilbert space, is a frame if there exist constants such that for all For the next results, we need additional properties of the fractal families in .
Proposition 2. If is a frame and , then is a frame of .
Proof. For the right inequality of (
17), let us think that for
,
where
k is the right bound of the frame
(
B in the expression (
17)). For the left inequality, according to Proposition 4.8 and Theorem 4.18 of [
1],
is injective with closed range and
is defined on the range of
. For
,
and
Moreover, since
is a frame,
where
is is the left bound of the frame
(
A in the expression (
17)). Using the inequality (18), the result is proved. □
Proposition 3. If are frames and , then is a frame of .
Proof. In this case,
and
are frames according to the previous proposition. The tensor product of frames is a frame (Theorem 2.3, [
22]) and consequently by Property
of Definition 4
is a frame. □
Definition 7. A sequence of a Hilbert space H is a Bessel sequence if there exists a constant such that for any Theorem 4. If , are frames of and , then is a Bessel sequence for any scale vectors α, β.
Proof. Let us consider any
then
The first inequality is due to the fact that is a frame, and the last equality comes from Property (3) of Definition 4. □
In the following lemma, represents the adjoint of A, and the same notation is used for all the operators concerned.
Lemma 2. If then , and Moreover, if is bounded below, there exists a constant such thatand, if is bounded below, there exists a constant such that Proof. Let us take, for instance, the first element
. It is linear and bounded operator from
to
and
and
where
and
are orthonormal bases of
and
, respectively. The proof of the rest of the inequalities is similar. □
Theorem 5. Let and be orthonormal bases of and , respectively. If is a frame and is bounded below, then is a frame of
The upper frame constant is and the lower is , where M is the lower frame constant of and is such thatfor any Proof. The right inequality of (17) is proved in the previous theorem. The tensor product of orthonormal bases is an orthonormal basis and the constant K of the previous theorem is equal to one.
For the left one, let us consider any
then
and
The previous inequality is due to the fact that
is a frame. Since
is bounded below,
applying the previous lemma. □
Proposition 4. Let and be orthonormal bases. If and then is a frame.
Proof. It is a consequence of Theorem 2 ( is a Riesz basis) and Proposition 2 ( is a frame). □
Corollary 1. Let and be orthonormal bases. If is a frame, then is a frame.
Proof. In this case, () and is bounded below. Applying Theorem 5, one obtains the result. In addition, note that the tensor product of frames is a frame. □
The following lemma can be read in [
22], Theorem 2.6.
Lemma 3. If Q is a linear and bounded invertible operator of H and is a frame of , then the sequence is also a frame of
Theorem 6. If and are frames of and , respectively, then is a frame.
Proof. Let us consider the former lemma for
and
. With the hypothesis on
,
is invertible and consequently
is a frame. Let us see that
.
for every
. This completes the proof. □
Lemma 4. If Q is an invertible, linear and bounded operator of K and is a frame of , then is also a frame.
Proof. It can be read in [
22], Corollary 2.11. □
Theorem 7. If and are frames of and , respectively, then is a frame.
Proof. With the hypothesis on the scale vector , is invertible. Consequently, is as well. Since is a frame, applying the previous lemma, is a frame.
This equality completes the proof. □
Theorem 8. If , and are frames, then is a frame.
Proof. By definition of the tensor product of operators,
Let us consider now that, according to the previous theorem,
Since
is invertible,
is a frame. Then,
is a frame, using Theorem 6. □
Proposition 5. If and then is injective and Proof. With the hypotheses on
and
,
and
are injective with closed range (Proposition 4.8 and Theorem 4.18 of [
1]). Let us denote
Then, if
, due to the properties of the tensor product of vectors:
(The inverse operators are defined on the range of
and
respectively). The continuity and linearity of
imply that
For the other content, if
, then there exists
such that
Since
The equality of ranges (
19) implies that
is a Hilbert space and then closed. To prove the injectivity of
, let us consider that, if
then
according to Properties
and
of the tensor product of operators. □
Corollary 2. If and then is bounded below.
Proof. In these hypotheses,
and
are injective and
and
are isomorphisms on
and
, respectively.
is bounded on
and thus
for any
□
Corollary 3. If and then is surjective.
Proof. However,
is injective and
is closed, thus
is also closed ([
23]). Consequently, the operator quoted is surjective. □
Definition 8. A sequence , where H is a Hilbert space, is a frame sequence if it is a frame for its closed span .
Proposition 6. If and and and are frames, then is a frame sequence.
Proof. According to the definition of frame sequence, we need to prove that there exist constants
such that, for all
,
The right inequality is already proved in Theorem 4. According to Proposition 5, is bounded on the range of
If
, since
then
Hence (Equations (
20) and (
21)), there exists a constant
such that
and the left inequality is proved. □
Definition 9. A sequence of a Hilbert space H is a Riesz sequence if there exist such that for any Theorem 9. Let and be Riesz bases of and , respectively. If and is a Riesz sequence of
Proof. In these hypotheses, is a topological isomorphism on its range, and thus it preserves the bases. □
Definition 10. A sequence of a Banach space is a Schauder sequence if it is a Schauder basis ([24]) for . Theorem 10. If are Riesz bases of , and , then is a Schauder sequence of .
Proof. With the hypotheses given is a topological isomorphism from on and the isomorphisms preserve the bases. □
Proposition 7. If the operators are defined as , where are fixed mappings strictly increasing, continuous and such that and , then 1 belongs to the point spectrum of
Proof. In this case, 1 belongs to the point spectrum of
and
(Proposition 2 of [
2]) and the product of eigenvalues is an eigenvalue of the tensor product. □