Uniform estimates on paraproducts. (Q1809919): Difference between revisions
From MaRDI portal
ReferenceBot (talk | contribs) Changed an Item |
Set OpenAlex properties. |
||
Property / OpenAlex ID | |||
Property / OpenAlex ID: W2030117458 / rank | |||
Normal rank |
Latest revision as of 08:46, 30 July 2024
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Uniform estimates on paraproducts. |
scientific article |
Statements
Uniform estimates on paraproducts. (English)
0 references
2002
0 references
One defines a paraproduct form as a multilinear form \[ \Lambda(f_1,\dots,f_n)=\int_{\mathbb R} \sum_{k\in \mathbb Z} \prod_{i=1}^n Q_{i,k} f_i (x)\, dx. \] Here, \(Q_{i,k}f=f\ast \phi_{i,k}\) is `adapted to \(\{\xi:| \xi| \leq 2^k\}\)' meaning that \(\hat{\phi}_{i,k}\) has support in \(\{| \xi| \leq 2^k\}\) while the derivatives \(\phi^{(\alpha)}\) satisfy \(\| \varphi_{i,k}^{(\alpha)}\| _\infty \leq 2^{-k\alpha}\) for all \(\alpha\) up to some given order. One also imposes the moment conditions \(\hat{\phi}_{i,k}(0)=0\) for all \(i,k\). Then \(\Lambda\) is continuous from \(\prod L^{p_i}\) to \(\mathbb C\) provided \(\sum 1/p_i=1\), as standard Calderón-Zygmund theory shows. This can be proved with weaker moment conditions. However, the full complement of moment conditions allows the authors to extend the sum over \(k\) defining \(\Lambda\) to sums over all \((j_1,\dots, j_n)\in \Omega\) where \(\Omega\) is defined by the conditions \(j_i-j_{i'}\geq m(i,i')\) for a set of integers \(m(i,i')\), \(1\leq i<i'\leq n\). With these conventions, the authors prove (as a special case) that if \(n\geq 2\) and \(\Omega\) is as above, then \[ \left| \int_{\mathbb R} \sum_{(j_1,\dots,j_n )\in \Omega}\prod_{i=1}^n Q_{j_i,i} f_i \right| \leq C_{p,n} \prod_{i=1}^n \| f_i\| _{L^{p_i}}. \] As before, \(\sum 1/p_i=1\). The estimate is uniform in that the constant \(C_{p,n}\) depends only on the choice of the \(p_i\) and on \(n\), but not on the particular integers \(m(i,i')\). Technically, the proof uses the language of graph theory to organize the frequency contributions of the various terms. The proof also uses induction on \(n\).
0 references
paraproduct
0 references
multilinear form
0 references