검색
Article Search

JMB Journal of Microbiolog and Biotechnology

OPEN ACCESS eISSN 0454-8124
pISSN 1225-6951
QR Code

Article

Kyungpook Mathematical Journal 2021; 61(2): 371-381

Published online June 30, 2021

Copyright © Kyungpook Mathematical Journal.

Riesz and Tight Wavelet Frame Sets in Locally Compact Abelian Groups

Arvind Kumar Sinha* and Radhakrushna Sahoo

Department of Mathematics, National Institute of Technology Raipur, G.E. Road Raipur, Chhatisgarh-492010, India
e-mail : aksinha.maths@nitrr.ac.in and radhakrushnasahoo92@gmail.com

Received: September 30, 2019; Revised: December 25, 2020; Accepted: January 11, 2021

In this paper, we attempt to obtain sufficient conditions for the existence of tight wavelet frame sets in locally compact abelian groups. The condition is generated by modulating a collection of characteristic functions that correspond to a generalized shift-invariant system via the Fourier transform. We present two approaches (for stationary and non-stationary wavelets) to construct the scaling function for L2(G) and, using the scaling function, we construct an orthonormal wavelet basis for L2(G). We propose an open problem related to the extension principle for Riesz wavelets in locally compact abelian groups.

Keywords: wavelet frame sets, Riesz wavelets, tight wavelet frame sets, translational and multiplicative tilings, spectral set

Duffin and Schaeffer [10] introduced the notion of frames, and Daubechies et al. [9] the notion of wavelet frames. Among frames, tight wavelet frames, due to their numerical stability, play an essential role in the series representation of a function and signal transmission. Other types of wavelets with simple structure have been established on a Lebesgue measurable set in Rn with a non-negative and finite measure. The most fundamental one is the Haar wavelet Ψ(x)=1[0,1/2]1[1/2,1], introduced by Alfred Haar in 1910. It is a one-dimensional prototypical illustration of a wavelet. The Shannon wavelet, whose Fourier transform is 1[1,1/2)(1/2,1], is another typical example. The Shannon wavelet is an essential tool for analyzing and reconstructing functions. We concentrate on Shannon-type wavelets whose Fourier transform is supported in the finite measure domain. We study Shannon wavelet's Fourier transforms in the Shannon set Θ=[2π,π][π,2π). We take the orthonormal wavelet Ψ=2sinh(2x1)sinh(x) with Ψ^=1Θ, where Ψ^ is a Fourier transform of Ψ and 1Θ is an indicator of set Θ. For more examples and constructions of wavelet sets in Rn see [2, 3, 7, 23, 24]. Wang [29] addressed the existence of wavelet sets with the documentation of spectral sets. The characterization of wavelets has been generalized to n-dimensional Euclidean spaces. The dilation set is defined on any countable subset of non-singular matrices, whereas the translation set is defined on any countable subset in Rn.

Grepstad and Lev [14, 15] introduced the concept of multi-tiling and Riesz bases on a bounded Riemann measurable set in Rn. Shah et al. [25] provided complete characterizations of orthogonal families, tight frames, and orthonormal bases of Gabor systems on local fields of positive characteristic. More results in this direction can be found in [5, 6, 26, 28]. Iosevich et al. [19] gave the concept of tight wavelet frame sets in finite vector spaces. This paper continues the line of research for locally compact abelian (LCA) groups. Mayeli [21] gave the concept of Riesz wavelets, tiling, and spectral sets in LCA groups. Further motivated by works of [1, 4, 12, 20, 27], we introduce the sufficient condition for the existence of tight wavelet frame sets in LCA groups. We also construct an orthonormal wavelet basis for L2(G).

Some LCA groups, such as the rational p-adic additive group Q {p}, have no discrete subgroup in the field of p-adic rational numbers. Consequently, such groups do not apply to the classical concepts of wavelet sets. Thus the finite abelian group does not possess any wavelet set in the conventional sense. Iosevich et al. [19] defined tilings of tight wavelet frame sets in finite vector space. We introduce the concept of tilings of tight wavelet frame sets in infinite LCA groups. We assume that G is an infinite group and admits a discrete subgroup. For the notion of wavelet sets in Qp and a non-commutative setting we refer [3] and ]8], respectively. For Riesz wavelets' characterizations generated by multiresolution analysis in the finite cases, see [4, 16, 22].

The main objective of this work is to investigate the characterization of tight wavelet frame sets for infinite LCA groups equipped with a Haar measure. We present two approaches to construct (for stationary and non-stationary wavelets) the scaling function for L2(G). We construct an orthonormal wavelet basis for L2(G). We propose an open problem related to the extension principle for Riesz wavelets in locally compact abelian groups.

Now we give some well-known definitions and results.

Definition 2.1.([19])

A subset {xi}iZ of a Hilbert space H is said to be a frame if there exists a sequence of vectors {xi:iZ} in H and two positive real numbers PQ such that for every xH

PxH2 iZ|x,xiH|2QxH2.

P and Q are the lower and upper bounds of the frame. If P = Q , then the frame is called a tight wavelet frame and if P = Q = 1 , then the frame is called a Parseval frame.

Definition 2.2.([21]) A collection of Haar-measurable wavelet sets {Θl}l=1r in Rn is called a Riesz wavelet set if there exists a countable set SGL(n,R) and ΨlRn such that for Ψ^l=1Θl, where GL(n,R) is a general linear group of order n. The collection of elements

ω= l=1r{Δ(s)1/2Ψl(s(x)α):αR,sS},

is called a Riesz wavelet basis for L2(Rn), where Δ is a unitary operator over R and r is a fixed positive integer.

The main feature of frames is their redundancy; for example, they play an essential role in robustness. In Hilbert's space, the frame behaves as a simplified, visualized characterization of sets. Engineers and applied mathematicians have used frames for frequency-domain analysis of discrete-time systems. Frames are more attractive in the space L2(G) as it is closed to the mother wavelet.

Let α,β,γR be given. We define the linear operators: modulation Mα, dilation Dβ and translation Tγ, for a function fL2(G) as follows:

Mαf(x)=e2πiαx,  Dβf(x)=|β|1/2f(βx),  Tγf(x)=f(xγ),

If {Mα,Dβ,Tγ} is a frame in L2(G), then (α,β,γ) generates a wavelet frame in L2(G). The Fourier transform of a function fL2(G) is defined by f^(ω)=G f(x)e2πiωxdx. The Fourier transform of modulation Mα, dilation Dβ and translation Tγ, is defined by Mαf^=Mαf^, Dβf^=D1/βf^ and Tγf^=Tγf^ respectively.

Definition 2.3.([21]) Let ΨL2(Rn), n ≥ 1, then Ψ is called a wavelet for L2(Rn) if there exist D GL(n,R) and countable sets TRn, such that the family

{|det(M)|1/2Ψ(Mxt):MD,tT}

is an orthonormal basis for L2(Rn). The sets D and T are called dilations and translation sets, respectively.

The size of D and T of orthonormal wavelet frames for L2(Rn) have been studied by many authors, see [17, 18, 29].

Definition 2.4.([21]) If G is an LCA group equipped with a Haar measure and G^ its dual. A collection of measurable subsets Θk, 1kn, of G^ is

called a (Riesz) wavelet collection of sets for G if there is a countable subset A of Aut(G) and countable subsets TkG (1kn) such that

k=1n{Ψk,j,r:1kn,jTk,rA}

is a (Riesz) orthogonal basis for L2(G), Ψk,j,r(x)=(r)1/2Ψk(r(x)λ), Ψ^k=1Θk When n = 1, we may say Θ=Θ1 is a (Riesz) wavelet set.

Wavelet sets and minimally supported frequency wavelets were introduced by Fang and Wang [11] and studied exclusively by Hernandez et al. [17, 18]. A measurable set ΘRn with non-zero finite measure is called a wavelet set, if for the function ΨL2(Rn) with Ψ^=1Θ and the set ω of (2.2) becomes an orthogonal basis for L2(Rn). A simple way to establish a wavelet frame in Rn is to choose a function whose Fourier transform is the indicator of a measurable set. The classical interpretation of wavelet frame sets obtains dilations and translations of the family (2.4). A set ΘRn is called a wavelet frame set with respect to dilations and translations if for the function ΨL2(Rn) with Ψ^=1Θ, the family (2.4) becomes a frame in L2(Rn).

Wang [29] studied the characterization of wavelet sets. Shah and Debnath [26] developed tight wavelet frames using the unitary extension principle in local fields and gave sufficient conditions for the formation of tight wavelet frame sets with a limited number of refinable functions. Wavelet sets were discussed by many authors [5, 13]. Mayeli [21] gave the concept of tilings of Riesz wavelets in LCA groups. Let G be an infinite LCA group equipped with Haar measure and C a set of complex numbers. A linear function Ψ:GC is said to be a wavelet frame for L2(G) if there exists a subset AAut(G) and a countable subset T □ G such that the family

{Ψ(αxλ):αA,λT}

is an orthonormal basis with respect to wavelet frames for L2(G). Note that in continuous cases, for a non-singular matrix M, the dilation factor c=|detM|n/2 makes the dilation map f|det(M)|1/2f(Mx) an isometry. If c=1 then the dilation of function in Rn belongs to a discrete group. To overcome this problem in the discrete case, we take both stationary and non-stationary wavelet settings of LCA groups to define the class of automorphism in G. In this way, the dilation of a function in Rn can be obtained in a co-compact group as well as a discrete group. In [2, 19, 26] the relationship between spectral sets and translational tilings were discussed. A measurable set Θ with positive measure is called a spectral set if there exists a countable set T such that the collection of exponentials {e2πiδx:δT} forms an orthonormal basis for L2(Θ). In this case, the pair (Θ,T) is said to be a spectral pair.

Proposition 2.5.([29]) Let D ∈ GL(n,R), TRn and ΘRn be a positive finite Lebesgue measure. If {Mt(Θ):MD} is a tiling of Rn and (Θ,T) is a spectral pair, then Ψ=1ˇΘ is a wavelet with respect to the dilation set D and the translation set T. Conversely, if Ψ=1ˇΘ is a wavelet with respect to the dilation set D and the translation set T and 0 ∈ T, then {Mt(Θ):MD} is a tiling of Rn and (Θ,T) is a spectral pair.

Definition 2.6.([21]) Let Θ be a subset of Rn, then Θ is said to be a multiplicative tiling set for Rn corresponding to a collection of n×n non-singular matrices M if {α(Θ):αM} is a set partition for Rn i.e.,

αM1α(Θ)(x)=1,xRn.

Equivalently, Rn= αMα(Θ) where for any αα' the two sets α(Θ) and α'(Θ) are disjoint in measure.

Definition 2.7.([21]) Let Θ be a subset of Rn, then Θ is said to be a translational tiling set for Rn corresponding a countable set LRn if

* lL1Θ(xl)=1,xRn.

Equivalently, Rn= lLΘ+l, where for each pair of independent elements l and l', the sets Θ +l and Θ+l' disjoint in measure.

Let f be a function on an infinite LCA group G; then the Fourier coefficient is defined by

f^(ξ)= nGanχn(ξ)¯, ξG,

where χn(ξ)=e2πin.ξα is the characteristic function, α is in G, n.ξ is an inner product and we have an=f(n). Here n is a variable. The function f^ is the Fourier transform of f. The Fourier transform of the function f ∈ G is defined as

f(x)= nGbnχn(x),

where bn=f^(n). The Fourier transform F:L2(G)L2(G) given by ff^ is a unitary map. For any hL2(G), we shall denote by hˇ the inverse Fourier transform of h.

Let Aut(G) be the set of all automorphism of G. A linear function Ψ:L2(G)C is a wavelet frame if there exists a set of automorphism AAut(G) and a subset TG such that the linear map

{Ψ(αxβ):αA,βT}

is an orthonormal basis concerning a wavelet frame for L2(G).

When αAut(G) and β ∈ G is associated with dilation and translation operators σα and τβ defined by

σαΨ(x)=Ψ(αx)andτβΨ(x)=Ψ(xβ)

respectively.

Definition 3.1. A measurable subset ΘRn with finite and positive measure is called a wavelet set if there is a function ΨL2(Rn) with Ψ^=1Θ, such that the family 2.4 is an orthonormal basis for L2(G) .

Definition 3.2. Let J be a subset of an infinite LCA group G, then J is called a multiplicative tiling for G if there exists a set of group automorphism A in Aut(G) such that J tiles G multiplicatively by A, i.e.,

G= αAα(J)

for αα' the union of two sets α(J) and α'(J) are disjoint.

Definition 3.3. Let J be a subset of an infinite LCA group G, then J is called a translational tiling set for G if there exists a subset TG such that

G= βT(J+β),

and for ββ' the union of the two sets (J+β) and (J+β') is disjoint.

As a result, if J is a multiplicative tiling set for an infinite LCA group G and 0 ∈ G, then α(0)=0 is a natural requirement for any group automorphism α. We say a set J has spectrum K if the characteristic {χk}kK is an orthonormal basis for L2(J). In this case, we say J is a spectral and (J, K) is a spectral pair.

Proposition 3.4. Let G be an infinite LCA group, C a set of complex numbers and Ψ:L2(G)C, then for a certain group automorphism αAut(G) and βG we have

σατβΨ^(n)=χα1β(n)¯Ψ^(α*n),

where α*=(αt)1=(α1)t the inverse transpose of α.

If Ψ^=1J, where J is a spectral set, then

σατβΨ^(n)=χα1β(n)¯1αt(J)(n).

Proof. Let αAut(G) and β ∈ G. By applying a Fourier transform and modifying the variable using the definition of Fourier transforms, for all n ∈ G we have,

τβΨ^(n)= mGΨ(mβ)χn(m)¯  = mGΨ(m)χn(m+β)¯  =χn(β)¯ mGΨ(m)χn(m)¯  =χn(β)¯Ψ^(n)  =χβ(n)¯Ψ^(n)

and

σαΨ^(n)= mGΨ(αm)χn(m)¯  = mGΨ(m)χn(α1m)¯  = mGΨ(m)χα*n(m)¯  =Ψ^(α*n).

Now a combination of both equations (3.1) and (3.2) gives the assertion of the first part of Proposition 3.4.:

σατβΨ^(n)=τβΨ^(α*n)=χα* n(β)¯Ψ^(α*n)=χα1 β(n)¯Ψ^(α*n).

For the proof of the second part of proposition 3.4., we take the equality 1J(α*n)=1αt(J)(n), then the proposition also holds.

{\bf Proposition 3.5.} Let AAut(G) and TG. The family

{σατβΨ:αA,βT}

is an orthonormal basis for L2(G) if and only if the family

{χα1β(n)¯Ψ^(α*n):αA,βT}

is an orthonormal basis for L2(G). Here n is a variable.

The proof of Proposition 3.5. is straightforward using a Fourier transform and Proposition 3.4.

Propostion 3.6. There exists no non empty subset JG such that Ψ=1ˇJ, the inverse Fourier transform of the indicator function 1J, is a generator of Parseval wavelet frame for L2(G).

Proof. Now we have a contradictory argument to prove this proposition.

Suppose that there exists a non-empty set J and a subset AAut(G) and a set TG such that the set

{σατβ1ˇJ:αA,βT}

is a Parseval frame for L2(G). By Proposition 3.5., we get that

{χα*n(β)¯1αt(J)(n):αA,βT}={χα1β(n)¯1αt(J)(n):αA,βT}

is a Parseval frame for L2(G)=L2(G^), where G^ is a dual group of an infinite LCA group G. Let h^=1{0}L2(G) be a indicator function for the set {0}. Then

1=h^2=αA,βT|h,σατβΨ|2  =αA,βT|h^,σατβΨ^|2  =αA,βT|nG h ^ (n)1 αt(J)(n) χ α* n(β)¯|2  =αA,βT|1 αt(J)(0)|2  =(T)αA,βT|1 αt(J)(0)|2

where # stands for number of elements. Two cases are taken into consideration here:

Case one: If 0 ∈ J, then 0αt(J) for all α ∈ A. The above formula thus indicates that 1=(A)(T). So there must be only one element of the wavelet system, let A={α} and T={β}. Then all the vectors in L2(G) must be constant multiple of σατβΨ. We show that this is not the case if J ≠ G. Let g≠ 0 in L2(G) such that (g^)αt(J) is empty. There is no such function since J is not the whole group G. Then there is no constant ϵ such that g=ϵσατβΨ. This indicates that there is not just one element to the Parseval wavelet frame when JG, thus (A)(T)>1.

Case two: If 0 does not belong to J. By the equation in (3.4) we get 1=0, which is impossible. This completes the proof.

Therefore, the above Proposition 3.6. shows that there does not exist any Parseval wavelet frame for L2(G) generated by Ψ=1ˇJ.

Definition 3.7. Let J and K be two subsets of an LCA group G, and we say (J, K) is a tight wavelet frame spectral pair if the set of characteristic {χk:kK} is a tight wavelet frame for L2(G).

Proposition 3.8. Let J and K be two subsets of G and (J, K) be a spectral pair. Suppose that J is a multiplicative tiling set concerning a set of automorphisms AAut(G). The following hold true:

  • (i) (J,K) is a tight wavelet frame spectral pair with the frame bound 0<PQ<.

  • (ii) For all α ∈ A, (α(J),(α1)t(K)) is a tight wavelet frame spectral pair with the frame bound 0<PQ<.

  • (iii) The family {1α(J)χ(α1)t(k):αA,kK} is a tight wavelet frame for L2(G) with the frame bound 0<PQ<.

  • (iv) If 0 not in J, then {(J)1/21α(J)χ(α1)t(k):αA,kK} is an orthonormal basis for L2(G).

Proof. Suppose J has a spectrum K. Then {(J)1/2χk:kK} is an orthonormal basis for L2(G).

Proof of (i): Note that a projection map ff1J of L2(G) onto L2(G), therefore the image of the orthonormal basis {(J)1/2χk:kK} by this projection map is a Parseval frame for L2(G). This shows the statement (i).

Proof of (ii): Let H be a subgroup of LCA group G and {χk:kK} be a frame for L2(H) with the frame bound 0<PQ<. Then {χ(α1)t(k):kK} is a frame for L2(α(H)) with the unified frame bounds P and Q. To prove this, we define the map Fα:L2(H)L2(α(H)) by ffα1. The image of {χk:kK} under Fα is {1α(H)χ(α1)t(k):kK} and this map is unitary. Therefore, {1α(H)χ(α1)t(k):kK} forms a frame for L2(α(H)) with the same frame bounds.

Proof of (iii): By the supposition on the multiplicative tiling property of J we get

L2(G)= αAL2(α(J))

To complete (iii), we shall prove the following instead.

Let Y be a measurable set and I be an index set such that Y= αAYα are disjoint. Suppose that for all α ∈ A, L2(Yα) has a tight wavelet frame {fm,α:mIα} with the frame bound S. We are saying that the family {fm,α:αA,mIα} is a tight wavelet frame for L2(Y) with the frame bound S. To show this, let hL2(Y). Since {Yα:αA} is a partition for Y, then h=αAhα, hα=h1Yα and we get

h2=αAhα L2(Yα)2  =αA S 1mI |hα,f m,α L2 (Yα)|2  =S1αA,mI α |h,1Yαfm,αL2(Y)|2.

This completes the proof of (iii).

Proof of (iv): If J is a spectral set in with 0 does not belong to J; therefore, facts can be derived explicitly from the fact that any Parseval frame with normalized frame components is an orthonormal basis.

Now, we propose an open problem about the extension principle for Riesz wavelets in LCA groups.

Let M be a non-singular matrix with dilation factor c=|detM|1/2 and L a subset of an LCA group G, then for j ≥ 0, we define the 2πMjl shift operator

Tjl:L2(G)L2(G)such that,Tjlf=f(.2πMjl),fL2(G),lL.

Our aim is to construct scaling functions Φjr r=1,2,m in L2(G) and for the Riesz wavelets Ψjr, r=1,2,ρj in L2(G), where m and ρj are positive integers, such that the collection

{Φjr:j0,r=1,2,m}{TjlΨjr:j0,r=1,2,ρj,lL}

forms a Riesz wavelet orthonormal basis for L2(G).

In the way defined above, what type of scaling function can show the extension principle for Riesz wavelets in LCA groups?

  1. D. Barbieri, E. Hernandez and A. Mayeli, Tiling by lattices for locally compact abelian groups, C. R. Math., 355(2017), 193-199.
    CrossRef
  2. J. J. Benedetto and R. L. Benedetto, A wavelet theory for local fields and related groups, J. Geom. Anal., 14(3)(2004), 423-456.
    CrossRef
  3. J. J. Benedetto and M. T. Leon, The construction multiple dyadic minimally supportrd frequency wavelets on Rd, The functional and harmonic analysis of wavelets and frames, 43--74, Contemp. Math. 247, Amer. Math. Soc., 1999.
    CrossRef
  4. M. Bownik, Riesz wavelets and generalized multiresolution analysis, Appl. Comput. Harmonic. Anal., 14(3)(2003), 181-194.
    CrossRef
  5. M. Bownik and M. S. Jakobsen, On Wilson bases in L2(Rd), SIAM J. Math. Anal., 49(5)(2017), 3999-4023.
    CrossRef
  6. M. Bownik and J. Jasper, Existence of frames with prescribed norms and frame operator, Appl. Numer, Harmon. Anal., 4(2015), 103-117.
    CrossRef
  7. D. Chen, On the splitting trick and wavelet frame packets, SIAM J. Math. Anal., 31(4)(2000), 726-739.
    CrossRef
  8. B. Currey and A. Mayeli, Gabor fields and wavelet sets for the Heisenberg group, Monatsh. Math., 162(2011), 119-142.
    CrossRef
  9. I. Daubechies, A. Grossmann and Y. Meyer, Painless nonorthogonal expansions, J. Math. Phys., 27(1986), 1271-1283.
    CrossRef
  10. R. J. Duffin and A. C. Schaeffer, A class of nonharmonic Fourier series, Trans. Amer. Math. Soc., 72(1952), 341-366.
    CrossRef
  11. X. Fang and X. H. Wang, Construction of minimally supported frequency wavelets, J. Fourier Anal. Appl., 2(1996), 315-327.
    CrossRef
  12. S. S. Goh, B. Han and Z. Shen, Tight periodic wavelet frames and approximation orders, Appl. Comput. Harmon. Anal., 31(2011), 228-248.
    CrossRef
  13. S. S. Goh and K. M. Teo, Extension principles for tight wavelet frames of periodic functions, Appl. Comput. Harmon. Anal., 25(2)(2008), 168-186.
    CrossRef
  14. S. Grepstad and N. Lev, Multi-tiling and Riesz bases, Adv. Math., 252(2014), 1-6.
    CrossRef
  15. S. Grepstad and N. Lev, Riesz bases, Meyer’s quasicrystals, and bounded remainder sets, Trans. Amer. Math. Soc., 370(2018), 4273-4298.
    CrossRef
  16. B. Han and R.-Q. Jia, Characterization of Riesz bases of wavelets generated from multiresolution analysis, Appl. Comput. Harmon. Anal., 23(2007), 321-345.
    CrossRef
  17. E. Hernandez, X. Wang and G. Weiss, Smoothing minimally supported frequency wavelets, I, J. Fourier Anal. Appl., 2(1996), 329-340.
  18. E. Hernandez, X. H. Wang and G. Weiss, Smoothing minimally supported frequency wavelets, II, J. Fourier Anal. Appl., 3(1997), 23-41.
    CrossRef
  19. A. Iosevich, C. K. Lai, and A. Mayeli, Tight wavelet frame sets in finite vector spaces, Appl. Comput. Harmon. Anal., 46(2019), 192-205.
    CrossRef
  20. E. A. Lebedeva and J. Prestin, Periodic wavelet frames and time-frequency localization, Appl. Comput. Harmon. Anal., 37(2)(2014), 347-359.
    CrossRef
  21. A. Mayeli, Riesz wavelets, tiling and spectral sets in LCA groups, Complex Anal. Oper. Theory, 13(2019), 1177-1195.
    CrossRef
  22. C. Monico and M. Elia, Note on an additive characterization of quadratic residues modulo p, J. Comb. Inf. Syst. Sci., 31(2006), 209-215.
  23. L. S. Pontryagin, Topological groups, Princeton Univ. Press, 1946.
  24. S. P. Raja, Bezier and B-spline curves-a study and its application in wavelet decomposition, Int. J. Wavelets Multiresolut. Inf. Process., 18(4)(2020), 2050030, 38 pp.
    CrossRef
  25. F. A. Shah, O. Ahmad and N. A. Sheikh, Orthogonal gabor systems on local fields, Filomat, 31(2017), 5193-5201.
    CrossRef
  26. F. A. Shah and L. Debnath, Tight wavelet frames on local fields, Analysis, 33(2013), 293-307.
    CrossRef
  27. F. A. Shah and A. Wahid, Wavelet packets on locally compact abelian groups, An. St. Univ. Ovidius Constanta Ser. Mat., 18(2)(2010), 223-239.
  28. S. H. Srivastava and F. A. Shah, AB-wavelet frames in L2(Rn), Filomat, 33(2019), 3587-3597.
    CrossRef
  29. Y. Wang, Wavelets, tiling and spectral sets, Duke Math. J., 114(1)(2002), 43-57.
    CrossRef