Haar frame characterizations of Besov-Sobolev spaces and optimal embeddings into their dyadic counterparts (Q6160011)

From MaRDI portal
scientific article; zbMATH DE number 7700693
Language Label Description Also known as
English
Haar frame characterizations of Besov-Sobolev spaces and optimal embeddings into their dyadic counterparts
scientific article; zbMATH DE number 7700693

    Statements

    Haar frame characterizations of Besov-Sobolev spaces and optimal embeddings into their dyadic counterparts (English)
    0 references
    0 references
    0 references
    0 references
    0 references
    23 June 2023
    0 references
    The authors investigate the norm characterization for elements in Besov spaces \(B^s_{p,q}(\mathbb R)\) and Triebel-Lizorkin spaces \(F^s_{p,q}(\mathbb R)\) in terms of expressions involving their Haar coefficients or suitable variations thereof. The paper deals with the range of parameters \((s,p,q)\) in which the Haar system is not an unconditional basis. This complements previous work of the authors, e.g., [\textit{G.~Garrigós} et al., J. Geom. Anal. 31, No.~9, 9045--9089 (2021; Zbl 1478.46032)], where a complete description was given for the parameter range where the Haar system forms an unconditional basis or a Schauder basis. The main aim of the paper under review is to see that the range of those characterizations previously shown in terms of Haar coefficients can be extended to other parameters provided that they doubly oversample with Haar coefficients obtained by a shift, namely, if \(h_{j,\mu}(x)= h(2^jx-\mu)\) for \(j=0,1,2, \dots\) and \(\mu\in \mathbb Z\), where \(h=\chi_{[0,1/2)}- \chi_{[1/2,1)}\) and \(h_{-1,\mu}=\chi_{[\mu,\mu+1)}\) stands for the Haar system in \(\mathbb R\), they consider \(\tilde h_{j,\nu}(x)= h(2^jx-\nu/2)\) for \(j=0,1,2, \dots\) and \(\nu\in \mathbb Z\) and \(\tilde h_{-1,\nu}=\chi_{[\nu,\nu+1)}\) and look at the extended Haar system \(\{\tilde h_{j,\nu}: h\ge -1,\,\nu\in \mathbb Z\}\). They use the notation \(c_{j,\mu}(f)= 2^j |\langle f, \tilde h_{j,2\mu}\rangle|+ 2^j |\langle f, \tilde h_{j,2\mu+1}\rangle|\) for \(j=0,1,2, \dots\) and \(c_{-1,\mu}(f)=\langle f, \chi_{[\mu,\mu+1)}\rangle\) and get characterizations of the norms in Besov and Triebel-Lizorkin spaces using these sequences, extending some previously known results which simply used the coefficients \(2^{j}\langle f, h_{j,\mu}\rangle\). They also show that, in the case \(1/p<s<1\), the classical Besov space \(B^s_{p,q}\) is a closed subset of its dyadic counterpart. They also provide equivalent norms for the Sobolev space \(W^1_p(\mathbb R)\) for \(1<p<\infty\) and some optimal inclusions between \(B^s_{p,q}\) and \(F^s_{p,q}\) and their dyadic counterparts.
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    Haar system
    0 references
    Haar frames
    0 references
    Sobolev spaces
    0 references
    Besov spaces
    0 references
    Triebel-Lizorkin spaces
    0 references
    dyadic versions of function spaces
    0 references
    wavelets
    0 references
    splines
    0 references
    0 references