Translation-Invariant Operators on Lebesgue Spaces

  • Context: Undergrad 
  • Thread starter Thread starter Euge
  • Start date Start date
  • Tags Tags
    Operators
Click For Summary

Discussion Overview

The discussion revolves around the properties of translation-invariant bounded linear operators on Lebesgue spaces, specifically focusing on the implications of such operators existing between different \( L^p \) spaces. The scope includes theoretical aspects and definitions related to translation invariance.

Discussion Character

  • Technical explanation
  • Conceptual clarification

Main Points Raised

  • One participant asserts that if a nonzero, translation-invariant bounded linear operator \( T : L^p(\mathbb{R}^d) \to L^q(\mathbb{R}^d) \) exists, then it must follow that \( q \ge p \).
  • Another participant questions the definition of translation invariance, asking if it means that \( T(f(x+c))=T(f(x)) \) for any constant \( c \).
  • A further clarification is provided regarding the definition of translation invariance, stating that for \( y \in \mathbb{R}^d \) and the operator defined as \( (\tau_y f)(x) := f(x + y) \), the requirement is that \( T\circ \tau_y = \tau_y\circ T \) for all \( y \in \mathbb{R}^d \).
  • Another participant reiterates the same definition of translation invariance, emphasizing the operator's behavior under translation.

Areas of Agreement / Disagreement

Participants express uncertainty regarding the definition of translation invariance, with some seeking clarification while others provide definitions. The discussion does not reach a consensus on the implications of the existence of such operators.

Contextual Notes

The discussion includes assumptions about the properties of translation-invariant operators and their implications, which may depend on the definitions used. There are unresolved aspects regarding the implications of the operator's existence on the relationship between \( p \) and \( q \).

Euge
Gold Member
MHB
POTW Director
Messages
2,072
Reaction score
245
Show that if there exists a nonzero, translation-invariant bounded linear operator ##T : L^p(\mathbb{R}^d) \to L^q(\mathbb{R}^d)## where ##1 \le p, q < \infty##, then necessarily ##q \ge p##.
 
  • Like
Likes   Reactions: Greg Bernhardt
Physics news on Phys.org
Does translation invariant mean that ##T(f(x+c))=T(f(x))## for any constant c?
 
  • Like
Likes   Reactions: Euge
Office_Shredder said:
Does translation invariant mean that ##T(f(x+c))=T(f(x))## for any constant c?
If ##y\in \mathbb{R}^d## and ##(\tau_y f)(x) := f(x + y)## for all ##x\in \mathbb{R}^d##, we require ##T\circ \tau_y = \tau_y\circ T## for all ##y\in \mathbb{R}^d##.
 
Euge said:
If ##y\in \mathbb{R}^d## and ##(\tau_y f)(x) := f(x + y)## for all ##x\in \mathbb{R}^d##, we require ##T\circ \tau_y = \tau_y\circ T## for all ##y\in \mathbb{R}^d##.

Thanks. I tried to Google translation invariant operator but the first hit was a proof of this result, so I stopped looking.
 
  • Like
Likes   Reactions: Euge
If ##f : \mathbb{R}^d \to \mathbb{R}## is a smooth and compactly supported, the supports of ##f## and ##\tau_y f## are disjoint for all sufficiently large ##|y|##. Hence ##\|f + \tau_yf\|_p = (\|f\|_p^p + \|\tau_yf\|_p^p)^{1/p} = 2^{1/p}\|f\|_p## for ##|y| \gg 1##. By translational invariance of ##T##, a similar argument shows ##\|T(f + \tau_yf)\|_q = 2^{1/q} \|Tf\|_q## for ##|y| \gg 1##. Boundedness of ##T## forces ##2^{1/q} \|Tf\|_q \le \|T\| 2^{1/p} \|f\|_p##. Replacing ##f## with ##f + \tau_yf## and repeating the analysis produces the inequality ##2^{2/q} \|Tf\|_q \le \|T\| 2^{2/p} \|f\|_p##. Repeating the process, we find ##2^{N/q} \|Tf\|_q \le \|T\| 2^{N/p} \|f\|_p## for all positive integers ##N##. By density of smooth compactly supported functions in ##L^p## the same inequalities hold true for all ##f\in L^p##.

Since ##T## is nonzero, there is a nonzero ##f\in L^p## such that ##\|Tf\|_q > 0##. The inequality $$\|T\| \ge 2^{N(1/p - 1/q)} \frac{\|Tf\|_q}{\|f\|_p}\quad (N = 1,2,3,\ldots)$$ forces ##1/p - 1/q \le 0## (otherwise ##\|T\| = \infty##). In other words, ##q \ge p##.
 

Similar threads

  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 3 ·
Replies
3
Views
3K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 1 ·
Replies
1
Views
3K
  • · Replies 16 ·
Replies
16
Views
2K
  • · Replies 1 ·
Replies
1
Views
2K