2
$\begingroup$

Let $E$ be a Banach space, $T:E\rightarrow E$ a continuous, norm-bounded, and nonlinear mapping., and $\{x_n\}_{n\in\mathbb N}$ such that $$x_{n+1}=T(x_n),\:\forall n\in \mathbb{N}:=\{0,1,\cdots\}.$$ Let $$X_n=\overline{\text{Conv}}\{x_n,x_{n+1}\cdots\}.$$

Let $X_{\infty}=\bigcap_{n=0}^{+\infty}X_n$. We assume that $X_{\infty}$ is not empty, and compact.

So, I'm wondering, what we can or cannot say about the compactness of the sequence $\{x_n\}_{n\in \mathbb{N}}$.


Note that:

This problem/question is under the framework of Measure of non-compactness wich acts only on norm-bounded sets.

So, here bounded means that there exist $M>0$ such that: $$\left \| T(x) \right \|\leq M,\;\forall x\in E.$$

I don't think that this condition is useful in this question.

$\endgroup$

1 Answer 1

1
$\begingroup$

In general it may fail to be compact. Consider $E:=L_2(\mathbb{R})$, and $x_n:=\chi_{[n,n+1]}$. Clearly, for any $n\in\mathbb{N}$, all functions in the set $\overline{\text{co}}\{x_k:k\ge n\}$ have support in $[n,+\infty)$, and in the intersection we get a compact nonempty set, the singleton $X_\infty=\{0\}$. However no subsequences of $x_n$ converges strongly, since it already converges weakly to $0$, and $\|x_n\|_2=1$.

In general, in any first countable topological space, the compactness of a sequence $(x_n)_n$ is equivalent to $\bigcap_{n\ge k}\overline{\{x_n:n\ge k\}}$ being non-empty, since the latter is the set $\text{Lim}(x_n)$ of the limits of all converging subsequences . In a Banach space, this set may be any separable closed subset.

$\endgroup$
6
  • $\begingroup$ Oh, and I forgot to mention that here $T$ is the shift $u\mapsto u(\cdot-1)$, of course (it's linear, therefore nonlinear. One can make it uniformly bounded by truncation, if needed) $\endgroup$ Oct 27, 2020 at 21:39
  • $\begingroup$ What does "it's linear, therefore nonlinear" mean? $\endgroup$
    – LSpice
    Oct 27, 2020 at 21:52
  • 1
    $\begingroup$ Ah. I assumed "nonlinear" meant "it cannot be linear", but "not necessarily linear" seems like a more reasonable reading. $\endgroup$
    – LSpice
    Oct 28, 2020 at 3:18
  • 1
    $\begingroup$ "Nonlinear" usually means "not necessarily linear", like in "nonlinear Fredholm operator", etc. $\endgroup$ Oct 28, 2020 at 6:49
  • 1
    $\begingroup$ Yes, nonlinaer here means "not necessarily linear" and not "it cannot be linear". $\endgroup$
    – Motaka
    Oct 28, 2020 at 9:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.