Uniform families of ergodic operator nets (Q1955754): Difference between revisions

From MaRDI portal
RedirectionBot (talk | contribs)
Removed claim: reviewed by (P1447): Item:Q588018
RedirectionBot (talk | contribs)
Changed an Item
Property / reviewed by
 
Property / reviewed by: Wilfried Hazod / rank
 
Normal rank

Revision as of 22:58, 19 February 2024

scientific article
Language Label Description Also known as
English
Uniform families of ergodic operator nets
scientific article

    Statements

    Uniform families of ergodic operator nets (English)
    0 references
    0 references
    18 June 2013
    0 references
    Let \(\mathcal{L}(X)\) denote the algebra of bounded linear operators on a Banach space \(X\) and let \(G\ni g\mapsto S_g\in\mathcal{L}(X)\) denote a bounded continuous representation of a -- mostly amenable -- semi-topological semigroup \(G\). Generalizing, e.g., Cesaro means and Abel means (for \(G=\mathbb{Z}_+\)) and continuous analogues as, e.g., \(G=\mathbb{R}_+\ni s\mapsto (1/s)\int_0^s S(t)\,dt\) and various other means, the author investigates `strong, resp., weak ergodic nets' and `uniform families of ergodic nets': let \(S:=\{S_g:g\in G\} \subseteq \mathcal{L}(X)\), then a net \(\left(A_\alpha^S\right)_{\alpha\in \mathcal{A}}\) is strongly right ergodic if (1) \(A_\alpha^S\in\overline{co}(S)\) for all \(\alpha\in\mathcal{A}\) (\(\mathcal{A}\) denoting a fixed directed set) and (2) \(A_\alpha^S x - A_\alpha^S S_g x\to 0\) (for all \(x\in G, g\in G\)). Analogously, left ergodicity and weak ergodic nets are defined. The author's definition differs slightly from the `classical' one (cf., e.g., the monograph by \textit{U. Krengel} [Ergodic theorems. With a supplement by Antoine Brunel. Berlin-New York: Walter de Gruyter (1985; Zbl 0575.28009)]), however, relying on a result of the pioneer work of \textit{R. J. Nagel} [Ann. Inst. Fourier 23, No. 4, 75--87 (1973; Zbl 0252.47003)] it is shown in Theorem 1.4 that the existence of weak and strong ergodic nets are equivalent conditions. Uniform families of ergodic nets are families \(\left\{ A_\alpha^{S_i}: \alpha\in\mathcal{A}, i\in I\right\}\), where \(S_i, i\in I\), are bounded representations of \(G\) which are uniformly bounded and satisfy the ergodicity condition (2) and the convexity condition (1) uniformly for \(i\in I\). (The precise formulation is too complicated for a review.) A typical example is \(G=\mathbb{Z}_+\), \(\mathcal{A}=\mathbb{N}\) and \(I=\mathbb{T}:=\{\lambda\in\mathbb{C}:|\lambda|=1\}\). Let \(S\in \mathcal{L}(X)\), \(||S||\leq 1\). Then \(\left\{(1/N)\sum_0^{N-1} (\lambda S)^n : N\in\mathbb{N}, \lambda\in\mathbb{T}\right\}\) satisfies the above mentioned conditions, hence is a uniform family of ergodic nets (Wiener-Wintner theorem). If \(\left\{A_\alpha^{S_i}: \alpha\in\mathcal{A}, i\in I\right\}\) is a uniform family of ergodic nets, then, for all \(i\in I\) and \(x\in X\), \(A_\alpha^{S_i}x\) converges, however, this convergence need not to be uniform. In Theorem 2.6, the author obtains sufficient conditions for uniform convergence.
    0 references
    mean ergodic theorem
    0 references
    amenable semigroup
    0 references
    Wiener-Wintner theorem
    0 references
    ergodic net
    0 references

    Identifiers