Parseval frames from compressions of Cuntz algebras
A row co-isometry is a family (V i) i = 0 N - 1 of operators on a Hilbert space, subject to the relation ∑ i = 0 N - 1 V i V i ∗ = I. As shown in Bratteli et al. (J Oper Theory, 43, 97–143, 2000), row co-isometries appear as compressions of representations of Cuntz algebras. In this paper we will present some general constructions of Parseval frames for Hilbert spaces, obtained by iterating the operators V i on a finite set of vectors. The constructions are based on random walks on finite graphs. As applications of our constructions we obtain Parseval Fourier bases on self-affine measures and Parseval Walsh bases on the interval.
Copyright comment Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
Introduction
Structured bases appear in harmonic analysis, operator theory, and approximation theory, among other areas. The classical example of a structure basis is an exponential (Fourier) basis, which gives rise to Fourier series expansions. A probability measure on is spectral if there exists a sequence of exponential functions that form an orthonormal basis for . Lebesgue measure on the unit (hyper-)cube is spectral; remarkably, Jorgensen and Pedersen initially showed that there are fractal measures which are spectral [[26]]. Wavelet bases [[8]] are another ubiquitous class of structured bases. These arise from the action of a system of unitary operators on —dilations and translations [[12]]—that encode natural operations on the latent space. Wavelets, however, lead a double existence between and , as elucidated by Mallat [[27]]. Wavelet bases in are generated by the iterated action of a finite number of (co-)isometries. These co-isometries give rise to a notion of scale in , and the corresponding scale decomposition is referred to as the cascade algorithm.
In the case of wavelet bases, the co-isometries satisfy what are now known as the Cuntz relations:
Graph
These relations were in fact observed by engineers—albeit without a precise mathematical formulation—in the first half of the twentieth century. Cuntz is credited with the discovery and thorough study of the algebras generated by such systems of co-isometries and the associate representation theory of those algebras [[7]]. The role played by the Cuntz algebras in wavelet theory was described in the work of Bratteli and Jorgensen [[1]–[4]]. Orthonormal wavelet bases (ONB) are constructed from various choices of quadrature mirror filters (e.g. see [[8]]). These filters are in one-to-one correspondence with certain representations of a Cuntz algebra.
The Cuntz relations give an elegant way to understand the geometry of cascade algorithms (for example the Discrete Wavelet Transform). The first identity is used to decompose a vector v into a "cascade" of bits , where i indexes from 0 to . Recovery of v is obtained through the same identity: apply each to the bits and sum it all up. The second relation (orthogonality of operators) tells us that bit interference/overlapping is avoided.
In [[13]] new examples of orthonormal bases and also classic ones were found to be generated by Cuntz algebra representations. More precisely, a multitude of ONBs such as Fourier bases on fractals, Walsh bases on the unit interval, and piecewise exponential bases on the middle-third Cantor set can be gathered under the umbrella of Cuntz algebra representations. Moreover, in [[14]] a connection between the ONB property and irreducibility of a Cuntz algebra representation was made in the particular case of Walsh systems.
However, orthonormal bases are sometimes too restrictive—for example, not all measures are spectral [[26]], and orthogonal wavelets may lack certain desirable properties [[8]]. Non-orthogonal expansions given by frames were introduced by Duffin and Schaffer [[18]] and popularized by [[9]]. A Parseval frame for a Hilbert space H is a family of vectors such that
Graph
Parseval frames arise naturally as images of orthonormal bases under a co-isometry; notably, by the Naimark dilation theorem, all Parseval frames have this form [[22]].
One of our main motivations and applications comes from the harmonic analysis of fractal measures. The study of orthogonal Fourier series on fractal measures began with the paper [[26]], in connection with the Fuglede conjecture [[21]]. Jorgensen and Pedersen proved that, for the Cantor measure on the Cantor set with scale 4 and digits 0 and 2, the set of exponential functions
Graph
is an orthonormal basis of . Many more examples of spectral measures have been constructed since, see, e.g., [[10], [29]]. For the classical middle-third Cantor measure, Jorgensen and Pedersen proved that this construction is not possible [[26], Section 6], so that measure is not spectral. Strichartz [[29]] posed the natural question of whether this measure has a frame of exponential functions. This question is still open! Non-orthogonal (but non-frame) Fourier series expansions for the middle-third Cantor measure were constructed in [[24]].
Motivated by Strichartz's question, in [[28]], weighted Fourier Frames were obtained for the Cantor set by making the set bigger, then constructing a basis for the bigger set, and then projecting the basis onto the Cantor set. Using the same dilation technique, in [[16]], a multitude of Parseval frames of weighted exponential functions and generalized Walsh bases were constructed for self-affine measures and for the unit interval. These constructions lead naturally to reconsidering the Cuntz relations, noting that only the first relation is needed to reconstruct a signal originally decomposed by a cascade algorithm. In such a set-up the operators are called row co-isometries.
In this paper we continue with the philosophy first emphasized in [[22]]—that frames are compressions of orthogonal bases—by considering compressions of Cuntz algebras. Indeed, one key idea is a dilation result from [[5]] (Theorem 2.3 in the next section) which allows one to extend a representation of a row co-isometry (which lacks the orthogonality constraints) to a "genuine" Cuntz algebra representation. We extend this philosophy to the pair Parseval frame, row co-isometries. By doing so, we will present a general framework for the construction of Parseval frames and orthonormal bases from row co-isometries, a framework which includes both of the following settings: (i) Fourier bases on self-affine measures and (ii) Walsh bases on the interval. We show that the new results are effective and can capture and unify previously obtained examples of Parseval frames and ONBs.
In Sect. 2 we include some definitions and notations; in Sect. 3 we state the main results; in Sect. 4, we present the proofs and some related results, and in Sect. 5 we apply the theory to various classes of examples.
Preliminaries and notations
Definition 2.1
Let H be a Hilbert space. A family of N bounded operators on H that satisfy the relations
Graph
is called a family of Cuntz isometries, or a a representation of the Cuntz algebra .
Note that the second relation implies that the operators are isometries with orthogonal ranges, and the first relation implies that the sum of the ranges add up to the whole space.
Such a representation is called irreducible if the only operators A on H, that commute with and , i.e., , , for all , are multiples of the identity , for some . Equivalently, a representation is irreducible if and only if the only closed subspaces K of H which are invariant for the representation, i.e., , , for all , are and .
For a closed subspace K of H we will denote by the orthogonal projection onto K.
Definition 2.2
Let H be a Hilbert space. A family of vectors in H is called a frame, if there exist constants , called the frame bounds such that
Graph
The family of vectors is called a Parseval frame if the frame bounds are equal to 1, .
Theorem 2.3
[[5], Theorem 5.1] Let K be a Hilbert space, and let be bounded operators satisfying
2.1
Graph
Then K can be embedded into a larger Hilbert space H carrying a representation of the Cuntz algebra such that K is cyclic for the representation, and if is the projection onto K we have
2.2
Graph
The system is unique up to unitary equivalence, and if is defined by
Graph
then the commutant is isometrically order isomorphic to the fixed point set , by the map .
Definition 2.4
Let K be a Hilbert space, an integer, and some bounded operators on K, . We say that is a row co-isometry if
Graph
Definition 2.5
The data in Theorem 2.3 will be referred to as the Cuntz dilation corresponding to the row co-isometry .
Definition 2.6
Let denote the empty word. Let be the set of all finite words with digits in , including the empty word. For , we denote by the length of .
For a word , we will denote by and similarly for . Also and .
Definition 2.7
Let be a row co-isometry. Let be a closed subspace of K. Suppose the following conditions are satisfied:
- Vi∗K0⊂K0 for all .
- There exists an orthonormal basis for , with M finite, some maps , , such that , for all and .
- The maps are one-to-one whenever the transitions are possible, in the sense that, for all and all such that and , if then .
Then we say that acts on as a random walk on the graph . M is the set of vertices, the maps indicate an edge with label i, from a vertex to the vertex , and represents a weight for this edge, more precisely is a probability of transition from c to along this edge.
Indeed, since it follows that .
Note that if there are exactly two distinct such that , then the total probability of transition from c to is . We will use the convention that when we say "the probability of transition from c to ," we mean the probability of transition from c to (only) through i, that is, .
- If, for any , there exists in such that and , then we say that the random walk is irreducible. In other words, one can reach from through with positive probability .
- We say that is simple on if the only operators with
- 2.3
Graph
- are , ; recall that denotes the orthogonal projection onto the subspace .
- If, for any and , we have , whenever , we say that is reversing on the random walk.
- If, for any in M, there exists n such that, for all with , we have or , then we say that the random walk is separating.
- A word , , is called a cycle word for , if , and for and . The points , are called cycle points.
- A word is called a loop for if , or and . Note that any loop is of the form , for some cycle words for c.
In Sect. 5, we give several examples of such co-isometries, which include Fourier series on fractal measures, Walsh bases, and a combination of the two.
Definition 2.8
Let be a complete metric space. A map is called a contraction, if there exists a constant , such that
Graph
An iterated function system (IFS) is a finite family of contractions on X. By [[23], Section 3], given an iterated function system on X, there exists a unique closed bounded set such that
Graph
Furthermore is compact. is called the attractor of the iterated function system.
Main results
Structured bases regularly arise as the orbit under the action of a system of isometries applied to a collection of vectors. Since frames are compressions of orthonormal bases, structured frames might analogously arise under the action of a system of co-isometries. In the case of row co-isometries, the question of whether their action generates a frame depends on the action of the co-isometries themselves as well as the structure of their Cuntz dilation. We describe sufficient conditions under which frames are generated by row co-isometries in our main results, Theorems 3.1 and 3.6.
Theorem 3.1
Let be a row co-isometry and let be its Cuntz dilation. Suppose there is a subspace of K such that acts on as an irreducible random walk with orthonormal basis . Assume in addition that is simple on .
Fix a point . Define
Graph
(In particular, .)
For , define to be the set of words that end in exactly n cycle words for c, i.e., words of the form , , cycle words for c.
Let
Graph
Then
- Ec(n) is an orthonormal basis for , for all .
- Hc(n)⊆Hc(n+1) for all .
- ∪n≥0Ec(n)={Sωec:ω∈Ω}.
- Let . Then V is invariant for the Cuntz representation . Also is the Cuntz dilation of the row co-isometry and it is irreducible.
In addition, the following statements are equivalent
- *span¯{Sωec:ω∈Ω}=H.
- The Cuntz representation is irreducible.
- The only operators with
- ∑i=0N-1ViTVi∗=T,
Graph
If is reversing, then we also have:
- For all and all ,
- 3.1
Graph
and the previous statements are equivalent to
- {Vωec:ω∈Ωc(0)} is a Parseval frame for K.
For the second part of the paper, we will impose some extra conditions on the co-isometry .
Assumption 3.2
Let be a row co-isometry. We make the following assumptions
3.2
Graph
, and for all .
3.3
Graph
where .
3.4
Graph
Under these assumptions, one can define a random walk on the set , where the transition from t to is given probability . Note that,
Graph
An important role in the study of the Cuntz dilation associated to the co-isometry and the Parseval frames generated by it, is played by the minimal invariant sets associated to this random walk. We define these and more here.
Definition 3.3
For a point , and , we say that the transition is possible throughi if . We also write or .
Remark 3.4
Note that if , the transition from may be possible through either i or . We will make the convention that if we write " is possible", we mean " is possible through i," to distinguish the paths along i and .
Given a word , we define . We say that the transition is possible (in several steps) through , if all the transitions are possible.
We define
Graph
is the probability of transition from t to through (passing through , ).
Note that the transition is possible in several steps, if and only if .
A subset M of is called invariant, if, for all t in M and , if the transition is possible, then is in M. We define the orbit of to be
Graph
A closed invariant subset M of is minimal if there does not exist any proper closed invariant subset of M. Equivalently, as in Lemma 4.23, for any .
Given an invariant set M, we define the subspaces K(M) of K and H(M) of H by
Graph
Theorem 3.5
Suppose that the assumptions (3.2)–(3.4) hold. Then there are finitely many minimal compact invariant sets. If is the complete list of minimal compact invariant sets, then the spaces are invariant for , and the spaces are invariant for the representation . The spaces are mutually orthogonal and
Graph
The sub-representations on , are irreducible and disjoint. (Recall that two representations , are called disjoint if there is no non-zero intertwining operator , , for all i.)
Theorem 3.6
Assume in addition that the maps are one-to-one and that all the minimal compact invariant sets are finite. Pick a point in for every . Then acts as an irreducible, separating random walk on the spaces . If is reversing on each , then is a Parseval frame for K.
Proofs
Proof of Theorem 3.1
We begin with a general lemma which shows that projections of iterations of the Cuntz dilation isometries are the corresponding iterations of the row co-isometries.
Lemma 4.1
Let be the Cuntz dilation of the system , and the projection onto K. For ,
4.1
Graph
Proof
Let and . Using (2.2) and , we have
Graph
By induction, (4.1) follows.
We will also need the next general result which shows when the vectors are orthogonal.
Lemma 4.2
If is a representation of the Cuntz algebra, , and , then is a prefix of or vice-versa, i.e., there exists a word such that or .
Proof
Let , . If there exists k such that then take the smallest such k, and we have
Graph
This is impossible, therefore is a prefix of or vice-versa.
To prove item (a) in the theorem, we will use the next Lemma:
Lemma 4.3
If and , then there exists a loop for c such that or .
Proof
First, an easy computation shows that, for and ,
4.2
Graph
By Lemma 4.2, either or for some . Assume the first. We prove that is a loop. We have
Graph
This implies that and so is a loop for c.
Lemma 4.4
Suppose for some cycle words and for c and some words and ; then and . Also, every word can be written uniquely as for some cycle words for c, (n could be 0) and some word .
Proof
Suppose . Then, if we read the words from the right, we see that for some non-empty word . We have . Since all these transitions are possible and the maps are one-to-one in this case, it follows that , but this contradicts the fact that is a cycle word for c. Thus and .
Now take an arbitrary word . If it does not end in a cycle word, it is in . If it ends in a cycle word, by the previous statement, it can be written uniquely as , with cycle word. Repeat for and use induction, to obtain the last statement in the lemma.
See Remark 4.5, where we point out why the condition that the maps are injective is important in this context.
Now take in for some . If then, with Lemma 4.3, there exists a loop such that or . Assume that . Since is a loop, , we have that for some cycle words for c. Since ends in n cycle words for c, it follows that ends in cycle words for c, which, by Lemma 4.4, is impossible since . This shows that is an orthonormal basis for .
Remark 4.5
The condition that the maps be one-to-one when the transitions are possible is important in Lemma 4.4, and to guarantee that the vectors with in are orthogonal. Indeed, if we do not assume this condition, it is possible that a word ends both in a cycle word for c and in two cycle words for c. Consider
Graph
A simple check shows that is a row co-isometry on , and if is the standard basis for , then, this co-isometry acts as a random walk on , with , , , , , , , .
The random walk is clearly irreducible and, if with , we obtain that , so the co-isometry is simple on .
Note that both 0 and 10 are cycle words for . Therefore the word ends the cycle word 0 for c, since 1, 01 and 101 are not cycle words for c. At the same time ends in two cycle words for c, .
(b) We will need some key lemmas.
Lemma 4.6
Define the random walk/Markov chain , on M by
Graph
Then
Graph
Also the random walk is recurrent.
Proof
The first equality follows from a simple computation.
Since M is finite and the random walk is irreducible by hypothesis (see Definition 2.7), the random walk is also recurrent (see, e.g., [[19], Theorem 6.4.4] or [[20], Section XV.6]).
Definition 4.7
Let and , . We say that for the first time if , , and for . In particular, if and , then for the first time if and only if is a cycle word for c.
Lemma 4.8
Let . Then
4.3
Graph
Proof
By Lemma 4.6, the random walk on M is recurrent, so . But
Graph
Lemma 4.9
Let . Then
Graph
Proof
Using the Cuntz relations, and that for all and , we have
Graph
Continuing, by induction we get, for ,
4.4
Graph
Note that if we take two different that appear in the first sum, then one can not be the prefix of the other, by the definition of " for the first time". Also if appears in the first sum, and appears in the second, then is not a prefix of because . Using Lemma 4.2, it follows that all the terms in the first sum are mutually orthogonal, and they are orthogonal to all the terms in the second sum. Denote the second sum by . Since the norms are equal, we get
4.5
Graph
From Lemma 4.8,
Graph
therefore and this proves the lemma.
Let and , then with Lemma 4.9
Graph
But so . Thus . (c) is trivial. (d) Using Lemma 4.9, it follows that, for every , , thus . Clearly V is invariant for all the operators . To see that it is also invariant for the operators , let . If , then . If , then , according to the previous argument. Thus V is invariant for the Cuntz representation and it contains .
Since is invariant for all , we have and . Let . Then
Graph
Since spans V it follows that is cyclic for the representation . Thus is the Cuntz dilation of the row co-isometry . By Theorem 2.3, it follows that the Cuntz representation is irreducible, since is simple on .
(e) We assume now that is reversing. If , then , with and cycle words for c.
If is a cycle word for c, then . So
Graph
Let . Then, with (a), we have
Graph
But
Graph
by Lemma 4.8, so we obtain
Graph
But, with (b), we obtain
Graph
Next we prove the equivalences.
(i) (ii). If (i) holds, then and (ii) follows from (d).
(ii) (i). We know from (d) that V is invariant for the representation . Since this is irreducible, we must have which implies (i).
(ii) (iii). Follows from Theorem 2.3.
Assume now that is reversing.
(i) (iv). Follows from (e), since if , then .
(iv) (i). For this last implication we prove first the following
Lemma 4.10
Suppose V and K are closed subspaces of the Hilbert space H and suppose is an orthonormal basis for V. Then is a Parseval frame for K if and only if .
Proof
The sufficiency is well known. Indeed, if then
Assume now is a Parseval frame for K. Let . Then, since is an orthonormal basis for V, we have
Graph
The last equality holds because is a Parseval frame for K. But this implies that so . As k was arbitrary in K, it follows that .
With Lemma 4.1 we have for all . Then, with Lemma 4.10, we get that . V is invariant for the representation therefore we obtain that K is also cyclic for the representation and clearly invariant for all . Therefore is the Cuntz dilation of the row co-isometry . By the uniqueness in Theorem 2.3, it must be isomorphic to the representation . By (d), the Cuntz representation on V is irreducible, so (ii) follows. With an earlier implication this yields (i).
Proposition 4.11
Let be a row co-isometry which acts on the subspace as a random walk which is irreducible and separating. Then is simple on .
Proof
Let T be as in (2.3). Since is invariant for , we have , iterating (2.3) we obtain, for all ,
4.6
Graph
Then, take in M. Since the random walk is separating if we take n large enough, we have that, for any with , either or . Therefore
Graph
So, T has zero off-diagonal entries.
We prove that the diagonal entries are equal. By taking the real and imaginary parts, we can assume that T is self-adjoint. Let such that
Let . Since the random walk is irreducible, there exists such that and . Let . We have
Graph
But then, we must have equalities in all inequalities, and therefore, since , we get that . Thus all the diagonal entries of T are the same, and T is a multiple of the identity, so is simple on .
Remark 4.12
In Proposition 4.13 we show that, in many cases, the family is incomplete in H.
Proposition 4.13
As in Theorem 3.1, let be a row co-isometry and let be its Cuntz dilation. Suppose there is a subspace of K such that acts on as an irreducible random walk with orthonormal basis . Assume in addition that is simple on .
- Suppose for all and all , we have is either 0 or 1. Then M is a single cycle, is reversing, and, for , , and for , with and a loop for c, is a unimodular constant multiple of ; therefore is an orthonormal basis for the space .
- Suppose that there exists such that for two distinct . Then, for all , the family is incomplete in V.
Proof
For (i), if the probabilities are all 0 or 1, it follows that from each point in M there is a unique possible transition to another point in M, and since the random walk is irreducible, it follows that M has to be a cycle.
We prove that is reversing. Indeed let and let be the unique digit such that the transition is possible. We have
Graph
so . Thus is reversing.
Let , and let be the unique cycle word for c. Using the same argument as before, we get that is also reversing on the random walk and so . But this implies that if then . By induction, we obtain the last statements in (i).
For (ii), let . Let be a fixed cycle word for c. Let . We prove that is orthogonal to , for all , . Indeed, if this is not true, let , for some and cycle words for c, then by Lemma 4.3, we have that for some cycle words . But, from Lemma 4.4, we get that and and . So , a contradiction.
Note also that, is orthogonal to all , for all , , by Theorem 3.1.
Consider , which is perpendicular to , but also to all with , . Thus v is orthogonal to the subspace .
We just have to check that v is not zero.
If there exist , , such that the transitions , are possible, then, since the random walk is irreducible, there are possible transitions from and , back to c. Therefore c has at least two distinct cycle words , , and . We take c and in the previous argument to be . Since v is perpendicular to , using the Pythagorean theorem, we have:
Graph
Therefore, for all and so the family is incomplete in H.
Let us provide an application of Theorem 3.1 in the case when a row co-isometry admits periodic points.
Definition 4.14
A non empty word is called irreducible if there is no word such that with . Let be the set of words that do not end in , including the empty word.
Theorem 4.15
Let be the Cuntz dilation of the row co-isometry . Suppose is a unit vector such that for some irreducible word . Then the following are equivalent:
- {Vωv0|ω∈Ωβ} is a Parseval frame.
- {Sωv0|ω∈Ωβ} is ONB in H.
- {Sωv0|ω∈Ω} spans H.
- The representation is irreducible.
- The commutant of the Cuntz dilation is trivial.
- Bσ(K)=C1K.
Proof
We will use the following lemma:
Lemma 4.16
Let with , and then if and only if .
Proof
If then, since , we have
Graph
Therefore, .
For the converse, replace S by .
With the requirements of Theorem 3.1 in mind we construct the following data:
4.7
Graph
and for .
4.8
Graph
4.9
Graph
Note that the maps are trivially one-to-one when the transitions are possible.
Notice that for words containing a digit the transitions are not possible because in this case. Because we see that
Graph
Claim 4.17
M forms an orthonormal basis for .
First we notice the elements in M are unit vectors. Because for all , we have:
Graph
To show, for
4.10
Graph
assume without loss of generality where .
With Lemma 4.16, entails . Using the Cuntz orthogonality relations and , the left-hand side in (4.10) can be rewritten as . If this last term would not vanish then with Lemma 4.3 we get that is a loop for , for some . From (4.8) we have to have , which then gives with Lemma 4.16. Exploiting these fixed point properties for and we have succesively in finitely many steps
Graph
Since the isometries have orthogonal ranges we have that the infinite words are equal:
with a word shorter than . Then , which implies that
Graph
Let . Then there exists m, n in such that . Denote and for all . Then from the last equality above with infinite words, we have for for all . Then for all k:
Graph
So actually has period d, i.e. where . This contradicts the fact that is irreducible. Hence (4.10) holds and the Claim is proved.
Next, we check that the notions introduced in Definition 2.7 are satisfied in this setting.
Claim 4.18
for all . Thus item (i) in Definition 2.7 holds.
From the definition of M above we see that if . For the remaining cases we prove that , thus finishing the Claim.
. Let . Then is the projection onto , and is smaller than the projection . Hence
Graph
which implies . Since is a projection, it follows that
4.11
Graph
Then using the Cuntz relations
Graph
It follows that , . Hence
4.12
Graph
This concludes the Claim.
Notice that item (ii) in Definition 2.7 is satisfied because of Eq. (4.12) and the choice of M, and . Also the random walk is irreducible: if and with say then and , where .
Claim 4.19
is reversing on the random walk .
Suppose for . We need show . Because we have for some . From the definition of the we must have (the case follows similarly). What is needed to show now is
4.13
Graph
From (4.11) with and rewriting using the Cuntz relations we have
Graph
We apply the projection on both sides and use and Lemma 4.1 to obtain (4.13).
Claim 4.20
The random walk is separating.
Let and in M, with . We prove that if is a word with then either or . Suppose by contradiction that for some with both and are non zero. Using (4.9) we have , , and continuing we get for all . As a consequence the following infinite words are equal:
Graph
Concatenating the word on the left hand side we obtain
Graph
Using the same argument as above we get a contradiction with the fact that is irreducible.
With Proposition 4.11, it follows that is simple on .
We are done checking the settings and requirements of Theorem 3.1. We apply its item (d) and finish the proof.
In the particular case when one of the co-isometries admits a fixed point one easily obtains the following characterizations.
Corollary 4.21
Let be the Cuntz dilation of row co-isometry , and unit vector such that . The following are equivalent:
- The family of vectors is a Parseval frame in K.
- The family of vectors is ONB in H.
- The family of vectors spans H.
- The representation is irreducible.
- The commutant of the Cuntz dilation is trivial.
- Bσ(K)=C1K.
Example 4.22
If the Cuntz dilation is not irreducible, it is possible that is not even a frame for its span.
Define the operators on by
Graph
where is the canonical basis in , , , and , , and for , . Thus for all r.
Denote
Graph
with inverses
Graph
Clearly defines a bounded operator on . We compute the adjoints
Graph
Therefore
Graph
Then, we have
Graph
Thus
Also, if then .
Now, for , we have
Graph
Consider the vector , . If , then and , and
Graph
If , then
Graph
If is a frame,then for a frame constant , we have
Graph
Letting , we get , a contradiction.
Proof of Theorem 3.5
We remark first the following: suppose are two different minimal compact invariant sets. Then they have to be disjoint (see Lemma 4.23). Indeed, if they are not, then is compact invariant set which is contained in , and, since is minimal, , and similarly for , which is a contradiction.
We also note, that since the functions and are continuous, and using relation (3.3), we get
Graph
it follows that the maps are also continuous.
Lemma 4.23
For , the closed orbit is compact and invariant. If M is a minimal compact invariant set, and , then . If are distinct compact minimal invariant sets, they are disjoint.
Proof
First, we prove that the entire orbit of a, that is is compact. Let be a common contraction constant for all , . It suffices to show that any sequence in O has a convergent subsequence. Now let , , be a sequence in the orbit of a and fix some . By passing to a subsequence, we may assume that the sequence of lengths , otherwise the orbit would be finite, thus one could select a constant subsequence.
Let be the unique compact attractor of the iterated function sytem , i.e., the unique compact subset with the property that
Graph
See [[23], Section 3] or [[25]] for details.
Because the attractor is compact, one can extract a convergent subsequence of . There is no loss of generality if we relabel the subsequence and still denote it by . Say . Let and N large enough so that and whenever . Using the triangle inequality, we have, for :
Graph
Hence the subsequence is convergent.
Next we show that is invariant. Note first that is invariant: indeed if with possible then with possible. So and . It follows that is possible, hence . Now, if with possible (so ) then one can find a sequence with . Hence the transitions are possible, where are words depending on n. Because and are continuous we obtain and for n large enough. It follows that the transitions are possible, so . Because we obtain .
If then is contained in M. Since M is minimal and is invariant, we get .
If are distinct compact minimal invariant sets, and we assume that they are not disjoint, then for we have , a contradiction.
Lemma 4.24
The minimal compact invariant sets are contained in the attractor of the iterated function system .
Proof
Indeed, let M be a minimal compact invariant set. Let . Pick such that the transitions are possible. Then , for all . Let . Then, if is a common contraction ratio for the maps , we have . Since and is compact, it follows that is has a convergent subsequence to a point in . Then the same subsequence of is convergent to , which means . Then is a compact invariant set contained in M and in . Since M is minimal .
Thus, all minimal compact invariant sets are contained in the compact attractor of the iterated function system .
Lemma 4.25
There exists a constant such that for any two distinct minimal compact invariant sets and , .
Proof
We follow the argument in [[6]]. Since the maps are uniformly continuous on the attractor , there exists such that, if , then for all i.
Suppose now, that there exist and such that . Since , it follows that there exist such that . Then, since , we obtain that . This means that the transitions and are possible, so and . Also . Repeating the argument and using induction, we obtain some labels such that the transitions and are possible for all n. So and . But, using the same argument as in Lemma 4.25, converges to 0, which would mean that and this contradicts the fact that and are distinct minimal compact invariant sets.
Using Lemmas 4.25 and 4.24, all minimal compact invariant sets are cpntained in the compact attractor and they have a distance bigger than between any two of them. Thus, there can be only a finite number of minimal compact invariant sets.
We start with a simple relation that we will use often. For a word and ,
Graph
We have, for and , , if the transition is possible, and if the same transition is not possible. So is invariant for .
The next lemma shows that the spaces are invariant for the Cuntz representation .
Lemma 4.26
If is a subspace of K which is invariant under the operators , , then
Graph
is invariant for the representation .
Proof
Clearly, the space is invariant under the operators , . For a nonempty word , and we have . Also .
Next, we prove that the spaces are mutually orthogonal.
Lemma 4.27
If and are two closed invariant subsets with then
Graph
Proof
Let be a word and assume that the transitions and are possible in several steps. Then and . Also, since the maps are contractions with ratio , if then
Graph
so, if n is large enough then
Graph
and this would yield a contradiction. Thus, for long enough, one of the transitions or is not possible, so or .
Using this, with n large enough, and the co-isometry relation, we have
Graph
Lemma 4.28
Assume that the subspaces and are invariant for the operators , , and are mutually orthogonal, and let and be the subspaces of H as in Lemma 4.26. Then and are orthogonal.
Proof
Given two words and , if there exists i such that then, take the first such i and, for and :
Graph
since the ranges of and are orthogonal.
In the remaining case, is a prefix of (or vice-versa), for some word , then
Graph
because .
Using Lemmas 4.27 and 4.28, we get the next lemma:
Lemma 4.29
If and are two closed invariant subsets with then the subspaces and are orthogonal and the subspaces and are orthogonal.
Thus, the spaces are mutually orthogonal.
To prove the irreducibility of the Cuntz representation on , the disjointness and the fact that the sum of these spaces is H, we will introduce the Ruelle operator:
Definition 4.30
Define the Ruelle transfer operator for functions by
4.14
Graph
Recall that is the space of bounded operators T on K with
Graph
By Theorem 2.3, the commutant of the representation is in bijective correspondence with by the map .
Define the map , from to complex valued functions defined on , by
4.15
Graph
For an operator A in the commutant of the representation , the operator is in , and we define .
Lemma 4.31
The function , maps into the continuous fixed points of the Ruelle transfer operator, is linear, and order preserving. Also, .
Proof
It is clear that, for , the function is continuous. Also, the map is clearly linear and order preserving. We check that is a fixed point for R. We have, using the assumptions (3.2)-(3.3),
Graph
So is a fixed point for R. In particular, taking A to be the identity, since for all t, it follows that .
Lemma 4.32
Let and . Then there exists such that if has length , then
Graph
Proof
Let be the compact attractor of the iterated function system . We claim that there is a such that, if , and , then .
Suppose not. Then there exists a sequence in and a sequence in , such that and . Since is compact, by passing to a subsequence, we may assume that converges to some in . Since converges to 0, we get that also converges to . Since the map is continuous, we get that both and converge to , a contradiction.
Now, take some point in in . Since the maps are contractions, with some contraction ratio , if we take n large enough and in with length , we have, for :
Graph
and since is in , we get that , which implies that .
Lemma 4.33
Let A be in the commutant of . If then .
Proof
We have, for
Graph
Suppose . By Lemma 4.32, for a given , for long enough, . Using we have:
Graph
Then, using the Cauchy-Schwarz inequality
Graph
Since was arbitrary, we get for all , and since these vectors span the entire space, we get on K and therefore on H.
Lemma 4.34
Let F be a compact invariant set and let h be a real valued continuous fixed point of the Ruelle operator. Then the sets
Graph
are compact and invariant.
Proof
Let . We have that, if the transition is possible, then and . If the transition is not possible, then . Therefore
Graph
Thus, we must have equality in the inequality, which means that whenever . Therefore if the the transition is possible, so S is invariant. Similarly for I.
Lemma 4.35
If h is a continuous fixed point of the Ruelle operator then h is constant on minimal compact invariant sets.
Proof
If , then the real and imaginary part of h are also fixed points of the Ruelle operator, as R is order preserving. Hence we may assume that h is real-valued. Let M be a minimal compact invariant set. By Lemma 4.34, the set
Graph
is compact invariant and contained in M. Since M is minimal, it follows that , so h is constant on M.
Lemma 4.36
Let h be a continuous fixed point of the Ruelle operator. If on all minimal compact invariant sets, then on .
Proof
Taking the real and imaginary parts, we can assume h is real valued. Let . By Lemma 4.34, the set
Graph
is compact invariant, so, by Zorn's lemma, it contains a minimal compact invariant set . Then, for all ,
Graph
Similarly
Graph
So h is 0 on . Since was arbitrary, h is 0 everywhere.
Lemma 4.37
Let M be a minimal compact invariant set. Then H(M) is irreducible for the representation .
Proof
Suppose L is a closed subspace of H(M) which is invariant for the representation . Then, let be the projection from H to L, the projection from H to K and the projection from H to H(M). By Theorem 2.3 and Lemma 4.31, the function
Graph
is a continuous fixed point of the Ruelle operator. Also since , we have, with , From Lemma 4.29 we get that on all minimal compact invariant sets different than M. Indeed, if is such a set then, for , , so .
Since , we get that is zero on all minimal compact invariant sets different than M. Also, from Lemma 4.35, is constant c on M. Since on M, we get that on all minimal compact invariant sets. Therefore, by Lemma 4.36, on . This implies, by Lemma 4.33, that . Since and are projections, it follows that or , and this means that L is either or H(M), so H(M) is irreducible.
Lemma 4.38
If is a complete list, without repetitions, of the minimal compact invariant sets, then
Graph
Proof
Let and let be the corresponding projection. We know that commutes with the representation and we consider the fixed point of the Ruelle operator , .
For every and every , since , we have that . Thus is 0 on every minimal compact invariant set. Thus, according to Lemma 4.36, is zero everywhere, and by Lemma 4.33, , which means that .
Lemma 4.39
Let , be two distinct minimal compact invariant sets. Then the restrictions of the representation to and are disjoint.
Proof
Let be an operator that intertwines the two representations. Extend A from H to H, by letting on the orthogonal complement of . Then A commutes with the representation . We claim that the fixed point of the Ruelle operator for all .
If , then and , so, by Lemma 4.29, we get that .
If t is in a minimal compact invariant set M different than , then H(M) is orthogonal to , by Lemma 4.29, so by definition. Therefore in this case too. So, is zero on all the minimal compact invariant sets, therefore is zero everywhere, according to Lemma 4.36, which implies that , by Lemma 4.33.
Proof of Theorem 3.6
Assume now that the maps are one-to-one and that the minimal compact invariant sets are finite. We prove a Lemma.
Lemma 4.40
Assume that all the maps , are one-to-one. Let M be a minimal finite invariant set. Then for any in M, .
Proof
Let . Let be a common contraction ratio for the maps and let n be large enough, so that . Then, if , and the transitions and are possible (in several steps), then and
Graph
so and, since the maps are one-to-one, it follows that , a contradiction. Thus, for one of the transitions and is not possible. Then, using the Cuntz relations
Graph
Lemma 4.40 also shows that the random walk on M is also separating. Since in a minimal finite invariant set M, the orbit of any point is M, it follows that the random walk on M is irreducible. By Proposition 4.11, is simple on K(M) for any minimal finite invariant set.
We know that the Cuntz representation on is irreducible, and therefore, with Theorem 3.1, we have that and so , for any .
The proof that we have a Parseval frame is very similar to the end proof of Theorem 3.1.
Let . We have, for all , using the fact that is reversing on all the spaces :
Graph
But
Graph
by Lemma 4.8, so we obtain
Graph
Then,
Graph
The next proposition gives a sufficient condition for all the minimal compact invariant sets to be finite.
Proposition 4.41
Suppose the Assumptions 3.2 hold. Assume that at least one of the minimal compact invariant sets is finite. Let be the attractor of the iterated function system . Assume that the zero sets are finite, and that all the maps are one-to-one, for all . Then all the minimal compact invariant sets are finite.
Proof
Let be a finite minimal invariant set and let be a minimal compact invariant set. Let . Then, by Lemma 4.23, . Since is finite, we get that , so, there exist such that the transition is possible, so , where .
Since is continuous, there exists such that, if then .
Let c be a common contraction ration for the maps . Let such that . Let . Suppose the transition is possible. Then . Also,
Graph
since , and . Therefore and therefore the transition is possible. This means that is in M. Also,
Graph
By induction, and for all .
But, as , converges to the fixed point of , which is . Since and M is closed, it follows that , which contradicts the fact that .
Thus, the transition is not possible so . Since , we have
Graph
which is a finite set. Since x was arbitrary in M, it follows that M is finite.
Corollary 4.42
In the hypotheses of Theorem 3.5, assume in addition that the operators are Cuntz isometries. If is reversing on K, then for all , , the numbers are either 0 or 1, and is an orthonormal basis for K.
Proof
If is reversing on K, then for , if , we have . But, since is an isometry , so . This implies that for all .
The fact that is an orthonormal basis for K follows from Theorem 3.5 since the Cuntz dilation of is just on K.
Examples
Example 5.1
We start with a general example which combines exponential functions and piecewise constant functions on the attractor of an affine iterated function system.
Definition 5.2
The Hilbert space will be the -space associated to the invariant measure of an affine iterated function system.
Let R be a expansive integer matrix (i.e., all eigenvalues have ). Let , , . Consider the affine iterated function system
Graph
Let be the attractor of the iterated function system , i.e., the unique compact set with the property that
Graph
Let be the invariant measure of the iterated function system, i.e., the unique Borel probability measure such that
Graph
for all bounded Borel functions on . See [[23]] for details.
We say that the measure has no overlap if
Graph
Assume that has no overlap. Define by
Graph
so that for all , .
Now, to construct the co-isometry on the Hilbert space , suppose that there exist some points in , and ( ) such that the matrix
5.1
Graph
has orthonormal columns (so it is an isometry). In other words, for all in B:
5.2
Graph
Define the function on by:
Graph
Here is the characteristic function of the subset A.
Define the operators on by
5.3
Graph
Proposition 5.3
We have the following:
- For and :
- 5.4
Graph
- In particular, for ( ,
- 5.5
Graph
- {Vi}i=0M-1 is a co-isometry on . This shows that the Assumptions 3.2 are satisfied, with , and
- 5.6
Graph
- If, in addition, the matrix in (5.1) is unitary, then the operators are Cuntz isometries.
Proof
We have that, for ,
Graph
In particular, we have that
Graph
which implies (5.4).
Secondly, for , a fixed , and , we have
Graph
If the matrix in (5.1) is unitary, then we have that
Graph
Example 5.4
In this example, we consider the weighted Fourier frames studied in [[15], [28]].
As in Example 5.1, consider an affine iterated function system on ,
Graph
Assume now that there exist some points in , , and some complex numbers , , such that the matrix
5.7
Graph
is an isometry, i.e.,
5.8
Graph
By [[10], Theorem 1.6], we get that the measure has no overlap. Indeed, according to the cited reference, we just have to make sure that the elements in B are not congruent modulo . But this follows, by contradiction, from (5.8).
Note that this corresponds to a special case in Example 5.1, when, for all , we have for all , that is is independent of b. Using Proposition 5.3, we obtain that the isometries are given by
5.9
Graph
and they satisfy the Assumptions 3.2, with
5.10
Graph
where
5.11
Graph
The set is a minimal invariant set, because and , so the only possible transition from 0 is to with probability .
In dimension , since is a trigonometric polynomial, it has finitely many zeros in the attractor of the maps , and therefore we can use Proposition 4.41, to conclude that all minimal invariant sets are finite.
We check that is reversing on every space for all minimal invariant sets , . For this, we will use [[15], Proposition 4.2], which shows that, for every , , for all . We include the statement of that result, because it gives a lot of information about the structure of the minimal finite invariant sets in this situation:
Proposition 5.5
[[15], Proposition 4.2] Assume for all . Let be a non-trivial finite, minimal invariant set. Then, for every two points the transition is possible from t to in several steps. In particular, every point in the set is a cycle point. The set is contained in the interval .
If t is in and if there are two possible transitions and , then .
Every point t in is an extreme cycle point, i.e., and if is a possible transition in one step, then and
5.12
Graph
In particular for all .
Take , and such that . This means that the transition is possible and so and for all . Then, for , we have:
Graph
Also
Graph
So .
Thus, we can apply Theorem 3.5. So, we pick a point in each minimal invariant set . Recall that is the set of all words in that do not end in a cycle word for . We compute , for , and we show that
5.13
Graph
Indeed, using the fact that for all , and , take , and we have:
Graph
Then
Graph
The relation (5.13) then follows by induction.
Thus, with Theorem 3.5 and Corollary 4.42, we obtain
Corollary 5.6
[[16], Theorem 1.6] In dimension , let , be all the minimal finite invariant sets, and pick for each . The family of weighted exponential functions
Graph
is a Parseval frame for .
Corollary 5.7
[[11], Theorem 8.4] In dimension , suppose that the matrix
Graph
is unitary. Let , be all the minimal invariant sets, and pick for each . The family of exponential functions
Graph
is an orthormal basis for .
Example 5.8
In this example, we show that, in higher dimensions, it is possible to have minimal compact invariant sets which are infinite. Take
Graph
One can take
Graph
and all , so that the matrix in (5.7) is unitary.
We have
Graph
We note that the set is invariant. Indeed, . So, if then the second component of is and so
Graph
If , then the second component of is so . Thus the only possible transitions from are
Graph
Let be the attractor of the iterated function system , , i.e., the unique compact subset of such that
Graph
We claim that is a minimal compact invariant set.
If we compute the fixed points of and , which are and respectively, the interval is invariant for both and , which implies that .
Note that, if , then
Graph
if and only if . Thus, the only points for which not both transitions are possible, could be and . But these points are outside the interval so they are not in . Therefore, for all points in , both transitions are possible. But the closed orbit of any point in the attractor is the attractor itself, thus is minimal compact invariant.
Remark 5.9
A lot of information about the structure of the minimal invariant sets in higher dimensions can be found in [[6]].
Example 5.10
We present here an example of a Cuntz representation, where a cycle point has two cycle words and therefore the family is incomplete in the Hilbert space that corresponds to its minimal invariant set (by Proposition 5.5).
We consider an affine iterated function system on , as in Definition 5.2, determined by the scaling matrix , and digits where . Let , . We will order the set for the sake of indexing the matrices to follow. Take
Graph
(this appears also in [[28]]).
Graph
is unitary where denote the coordinates of . To simplify notation, we will often identify (x, 0) with x. Define, for ,
Graph
and .
Then from Proposition 5.3, since the matrix above is unitary, the operators form a representation of the Cuntz algebra . If we denote , then we have that the have the form:
Graph
We now find the compact minimal invariant sets. Assume is a compact minimal invariant subset of . Then if , then there must exist a possible transition for all n. By compactness of , there exists a convergent subsequence as , since the second component of is , which converges to 0 as . Since is minimal . Thus , for some (note that any transition from a point in leads to ). Now we calculate:
Graph
However, we see that is independent of b, so we denote
Graph
Therefore we may write
Graph
Thus, must be invariant for the maps and the weights . Note that in particular, since , and for , it suffices to see that is invariant for the and . We will show that or .
With Proposition 5.5, we know that, for any , we must have for all . So . Also, . Also, if and the transition to is possible, then is in so it must be of the same form . Note that only if x is of the form for some . Thus, if for some then the transition is possible and is not in . This means that c cannot be in . So we only have to check . is the trivial invariant set. Also is invariant. We have the possible transitions and . Since and are not in it follows that neither are nor .
In the case of , we see that has two distinct cycle words, 1 and 3, 0. Indeed and . Thus by Proposition 4.13, we know that .
Example 5.11
We use Theorem 3.5 to provide a class of Parseval frames as in [[17]], Theorem 3.11. Let A be a matrix such that (hence ) and the first row is constant , . With and define
Graph
Note that this corresponds to the Example 5.1 when , , for all and are the entries of the matrix A. Indeed, the attractor of the iterated function system is [0, 1] and the invariant measure is the Lebesgue measure on [0, 1].
By Proposition 5.3, the Assumptions 3.2 are satisfied, with
Graph
We will show that the only compact minimal invariant set is then apply Theorem 3.5.
The set is invariant because the only possible transition from 0 is with probability .
To show this is the only compact, minimal invariant set suppose by contradiction there is a compact minimal invariant set with some , . We argue that necessarily thus , contradicting the minimality of . If then because there must be at least one possible transition since . Continuing in this fashion we get for all . By compactness and invariance , as desired.
Since the only cycle word for 0 is 0, with Theorem 3.5, we obtain:
Corollary 5.12
[[17], Theorem 1.4] The family of functions
Graph
is a Parseval frame for .
Acknowledgements
We would like to thank professor Deguang Han for very helpful conversations and to the anonymous referee for a very careful and thorough review of our paper.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
[
References
1
Bratteli O, Evans DE, Jorgensen PET. Compactly supported wavelets and representations of the Cuntz relations. Appl. Comput. Harmon. Anal.. 2000; 8; 2: 166-196. 1743534. 10.1006/acha.2000.0283. 0960.42013
2
Bratteli O, Jorgensen PET. Isometries, shifts, Cuntz algebras and multiresolution wavelet analysis of scale N\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$N$$\end{document}. Integral Equ. Oper. Theory. 1997; 28; 4: 382-443. 1465320. 10.1007/BF01309155. 0897.46054
3
Bratteli, O., Jorgensen, P.: Wavelets through a looking glass. In: Applied and Numerical Harmonic Analysis. Birkhäuser Boston Inc., Boston (2002). (The world of the spectrum)
4
Bratteli, O., Jorgensen, P.E.T.: Wavelet filters and infinite-dimensional unitary groups. In: Wavelet Analysis and Applications (Guangzhou, 1999), Volume 25 of AMS/IP Stud. Adv. Math., pp. 35–65. Amer. Math. Soc., Providence (2002)
5
Bratteli O, Jorgensen PET, Kishimoto A, Werner RF. Pure states on Od\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\cal{O} }_d$$\end{document}. J. Oper. Theory. 2000; 43; 1: 97-143
6
Cerveau D, Conze J-P, Raugi A. Ensembles invariants pour un opérateur de transfert dans Rd\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${ R}^d$$\end{document}. Bol. Soc. Brasil. Mat. (N.S.). 1996; 27; 2: 161-186. 1418931. 10.1007/BF01259358. 0883.42028
7
Cuntz Joachim. Simple C∗\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$C^*$$\end{document}-algebras generated by isometries. Commun. Math. Phys.. 1977; 57; 2: 173-185. 467330. 10.1007/BF01625776. 0399.46045
8
Daubechies, I.: Ten Lectures on Wavelets, Volume 61 of CBMS-NSF Regional Conference Series in Applied Mathematics. Society for Industrial and Applied Mathematics (SIAM), Philadelphia (1992)
9
Daubechies I, Grossmann A, Meyer Y. Painless nonorthogonal expansions. J. Math. Phys.. 1986; 27; 5: 1271-1283. 836025. 10.1063/1.527388. 0608.46014
Dutkay DE, Haussermann J, Lai C-K. Hadamard triples generate self-affine spectral measures. Trans. Am. Math. Soc.. 2019; 371; 2: 1439-1481. 3885185. 10.1090/tran/7325. 1440.42030
Dutkay DE, Jorgensen PET. Iterated function systems, Ruelle operators, and invariant projective measures. Math. Comp.. 2006; 75; 256: 1931-1970. 2240643. 10.1090/S0025-5718-06-01861-8. 1117.28008electronic
Dai, X., Larson, D.R.: Wandering vectors for unitary systems and orthogonal wavelets. Mem. Am. Math. Soc. 134(640):viii+68 (1998)
Dutkay, D.E., Picioroaga, G., Song, M.-S.: Orthonormal bases generated by Cuntz algebras. J. Math. Anal. Appl. 409(2):1128–1139 (2014)
Dutkay DE, Picioroaga G, Silvestrov S. On generalized Walsh bases. Acta Appl. Math.. 2019; 163: 73-90. 4008697. 10.1007/s10440-018-0214-x. 1423.42049
Dorin Ervin Dutkay and Rajitha Ranasinghe. Weighted Fourier frames on fractal measures. J. Math. Anal. Appl.. 2016; 444; 2: 1603-1625. 3535778. 10.1016/j.jmaa.2016.07.042. 1345.28016
Dorin Ervin Dutkay and Rajitha Ranasinghe. Weighted Fourier frames on self-affine measures. J. Math. Anal. Appl.. 2018; 462; 1: 1032-1047. 3771287. 10.1016/j.jmaa.2017.12.055. 1390.28015
Dorin Ervin Dutkay and Rajitha Ranasinghe. Parseval frames of piecewise constant functions. Oper. Matrices. 2020; 14; 2: 317-331. 4118598. 1464.42023
Duffin RJ, Schaeffer AC. A class of nonharmonic Fourier series. Trans. Am. Math. Soc.. 1952; 72: 341-366. 47179. 10.1090/S0002-9947-1952-0047179-6. 0049.32401
Durrett R. Probability: Theory and Examples, Volume 31 of Cambridge Series in Statistical and Probabilistic Mathematics. 20104: Cambridge; Cambridge University Press. 10.1017/CBO9780511779398
Feller, W.: An Introduction to Probability Theory and its Applications, vol. I, 3rd edn. Wiley, New York-London-Sydney (1968)
Fuglede Bent. Commuting self-adjoint partial differential operators and a group theoretic problem. J. Funct. Anal.. 1974; 16: 101-121. 470754. 10.1016/0022-1236(74)90072-X. 0279.47014
Han, D., Larson, D.R.: Frames, bases and group representations. Mem. Am. Math. Soc. 147(697):x+94 (2000)
Hutchinson John E. Fractals and self-similarity. Indiana Univ. Math. J.. 1981; 30; 5: 713-747. 625600. 10.1512/iumj.1981.30.30055. 0598.28011
Herr J, Weber ES. Fourier series for singular measures. Axioms. 2017; 6; 2: 7
Jorgensen, P.E.T.: Harmonic Analysis, Volume 128 of CBMS Regional Conference Series in Mathematics. American Mathematical Society, Providence (2018). Smooth and non-smooth, Published for the Conference Board of the Mathematical Sciences
Jorgensen PET, Pedersen S. Dense analytic subspaces in fractal L2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$L^2$$\end{document}-spaces. J. Anal. Math.. 1998; 75: 185-228. 1655831. 10.1007/BF02788699. 0959.28008
Mallat Stephane G. Multiresolution approximations and wavelet orthonormal bases of L2(R)\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$L^2({ R})$$\end{document}. Trans. Am. Math. Soc.. 1989; 315; 1: 69-87. 0686.42018
Picioroaga Gabriel, Weber Eric S. Fourier frames for the Cantor-4 set. J. Fourier Anal. Appl.. 2017; 23; 2: 324-343. 3622655. 10.1007/s00041-016-9471-0. 1362.42066
Strichartz Robert S. Mock Fourier series and transforms associated with certain Cantor measures. J. Anal. Math.. 2000; 81: 209-238. 1785282. 10.1007/BF02788990. 0976.42020
]
Reported by Author; Author; Author; Author