Topological properties of the sequence spaces \(S^{\nu }\) (Q2497366)
From MaRDI portal
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Topological properties of the sequence spaces \(S^{\nu }\) |
scientific article |
Statements
Topological properties of the sequence spaces \(S^{\nu }\) (English)
0 references
4 August 2006
0 references
The authors define and study topological properties of sequence spaces based on the distribution of wavelet coefficients. More precisely, let \(\nu\) be a nondecreasing right-continuous function defined on \({\mathbb R}\) and taking values in \(\{-\infty\}\cup [0,d].\) Assume that there exists \(s_0\) such that \(\nu(\alpha) = -\infty\) if \(\alpha < s_0\) and \(\nu(\alpha)\in {\mathbb R}\) if \(\alpha \geq s_0.\) In [J. Fourier Anal. Appl. 10, 221--246 (2004; Zbl 1075.42014)], \textit{S. Jaffard} considered a space \(S^{\nu}\) consisting of those distributions whose wavelet coefficients, with respect to a fixed wavelet basis, satisfy that for all \( \alpha \in {\mathbb R},\;\epsilon > 0,\;C > 0\) there is \(J\) such that, for every \(j\geq J,\) the cardinality of the set \[ E_j(C,\alpha):= \{k\;:\;| c_{jk}| \geq C2^{-\alpha j}\} \] is less than or equal to \(2^{(\nu(\alpha)+\epsilon)j}.\) He proved that the space of distributions thus defined does not depend on the chosen wavelet basis, under the assumption that the basis satisfy some mild assumptions. In the present paper, the emphasis is put on the underlying sequence space. More precisely, let \(S^{\nu}\) denote the vector space of all sequences \(\overrightarrow{c}:=(c_{jk})\) (\(j\geq 0\), \(k\in \{0,\ldots, 2^j-1\}^d\), \(d\) being the dimension of the space where the wavelets are defined) satisfying the above condition or, equivalently, \[ \nu_{\overrightarrow{c}}(\alpha) \leq \nu(\alpha)\;\forall \alpha \in {\mathbb R}. \] Here, \[ \nu_{\overrightarrow{c}}(\alpha):=\lim_{\epsilon \to 0^+}\Big(\limsup_{j\to +\infty}\Big(\frac{\text{ ln}(\sharp E_j(1,\alpha + \epsilon)(\overrightarrow{c}))}{\text{ ln} 2^j}\Big)\Big). \] The authors aim to study this space from the point of view of functional analysis. They prove that the spaces of type \(S^{\nu}\) can be endowed with a (unique) complete metric which makes \(S^{\nu}\) a topological vector space. Also, a complete characterization of the compact subsets is obtained, as well as the continuity of operators which appear in a natural way when dealing with wavelet coefficients. Under the additional assumption that \(\nu\) is a concave function, it is proved that \(S^{\nu}\) equals to a countable intersection of Besov spaces. Moreover, the concavity of \(\nu\) is also necessary for the equality between \(S^{\nu}\) and the intersection of Besov spaces just mentioned.
0 references
distribution of wavelet coefficients
0 references
sequence spaces
0 references