An Integral Inequality With Applications (Q3335984)

From MaRDI portal

!

This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:

scientific article; zbMATH DE number 3867635
Language Label Description Also known as
default for all languages
No label defined
    English
    An Integral Inequality With Applications
    scientific article; zbMATH DE number 3867635

      Statements

      0 references
      1984
      0 references
      C*-convex functions
      0 references
      integral inequality
      0 references
      exponential integrability
      0 references
      Sobolev functions
      0 references
      An Integral Inequality With Applications (English)
      0 references
      Let \(1\leq q<\infty\), \(p^{-1}+q^{-1}=1\), \(f\in L^ q({\mathbb{R}}^+)\) with \(f\geq 0\) and \(\| f\|_ q\leq 1,\) where \({\mathbb{R}}^+=[0,\infty).\) If \(\phi \geq 0\) is locally integrable, define the functions \(\psi\), F on \({\mathbb{R}}^+\) by \[ \psi(x)=(\int^{x}_{0}\phi^ p dr)^{1/p},\quad F(x)=\int^{x}_{0}f\phi dr, \] and let \(\Phi:{\mathbb{R}}^+\to {\mathbb{R}}^+\) be nonincreasing. The integral inequality considered by the author is \[ (1)\quad \int^{\infty}_{0}\Phi \{N(\psi(x))- N(F(x))\}dN(\psi(x))\geq C_{N,q}\int^{\infty}_{0}\Phi(t)dt. \] The main result is that a finite constant \(C_{N,q}\) exists such that (1) is valid for all \(\phi\), \(\Phi\), f satisfying the above conditions, provided \(N:{\mathbb{R}}^+\to {\mathbb{R}}^+\) is a \(C^*\)-convex function. The definition of \(C^*\)-convex functions is somewhat complicated, but the author shows (Theorem 1) that if \(N\in C^ 1({\mathbb{R}}^+)\) is convex with \(N(0)=0, N(x)\geq 0,\) then N is \(C^*\)-convex precisely when there exist constants \(c,d(0<c<d<\infty)\) such that \[ (N^{-1}(\ell s))^ 2\geq [N^{-1}((\ell -1)s)][N^{-1}((\ell +c)s)]\quad for\quad \ell>d,\quad 0<s<\infty. \] The functions \(N(x)=e^ x-1\), \(e^{x^ 2}- 1\), \(x^ p(p\geq 1)\) are \(C^*\)-convex, for example. The author's result (Theorem 3) is actually somewhat more general than (1), but we omit details here. Using the more general version, an application is given consisting of an extension of a result of \textit{J. Moser} [Indiana Univ. Math. J. 20, 1077-1092 (1971; Zbl 0203.437)] on exponential integrability of certain spaces of Sobolev functions. As a second application, by taking \(\Phi(t)=e^{-t},\) and making the change of variable \(x=\{N^{-1}(\log(1/u))\}^ p,\) (1) leads to a generalization of an inequality of \textit{M. Jodeit} jun. [Stud. Math. 44, 545-554 (1972; Zbl 0244.26010)]. See also \textit{C. J. Neugebauer} [Proc. Am. Math. Soc. 82, 560-564 (1981; Zbl 0483.26009)].
      0 references

      Identifiers

      0 references
      0 references
      0 references
      0 references
      0 references
      0 references