3
$\begingroup$

I'm studying harmonic analysis by myself, and I read some online notes that introduce the Bernstein inequality. One of them mention a reversed form of the Bernstein inequality, which is stated below:

Let $\mathbb{T} = \mathbb{R} / \mathbb{Z} = [0,1]$ be the one-dimensional torus. Assume that a function $f \in L^{1}(\mathbb{T})$ satisfies $\hat{f}(j) = 0$ for all $|j| < n$ (vanishing Fourier coefficients). Then for all $1 \leq p \leq \infty$, there exists some constant $C$ independent of $n,p$ and $f$, such that $$\|f'\|_{p} \geq Cn\|f\|_{p}$$

It seems that an easier problem can be obtained by replacing $f'$ with $f''$ in the above inequality. The easier problem is addressed in the MO post below:

Does there exist some $C$ independent of $n$ and $f$ such that $ \|f''\|_p \geq Cn^2 \| f \|_p$, where $1 \leq p\leq \infty$?

However, it seems that the trick of convex Fourier coefficients used in the post above no longer applies to the harder problem (lower bounding the norm of the first derivative). Any suggestions/ideas?

$\endgroup$
1

1 Answer 1

1
$\begingroup$

Let me define the Fourier multiplier $\Omega_N (D)$ by $$ (\Omega_N(D) u)(x)=\text{Fourier}^{-1}\bigl(\hat u(\xi)\mathbf 1 (\vert \xi\vert \ge N)\bigr). $$ We define then the operator $A_N(D)=D\Omega_N(D)$ (also a Fourier multiplier). We have obviously $$ \Vert A_N(D) u\Vert_{L^2}\ge N \Vert \Omega_N(D) u\Vert_{L^2},\quad\text{i.e.}\quad \Vert D u_N\Vert_{L^2}\ge N \Vert u_N\Vert_{L^2}\quad\text{ with $u_N=\Omega_N(D) u$.} $$ Let $p$ be in $(1,+\infty)$. We have $$ u_N= \underbrace{\vert D\vert^{-1} N\Omega_N(D)}_{ \substack{\text{bounded operator on } L^p}}N^{-1}\vert D\vert u_N, $$ so that $$ \Vert u_N\Vert_{L^p}\le C_p\Vert N^{-1}\vert D\vert u_N\Vert_{L^p} \le N^{-1}\tilde C_p\Vert D u_N\Vert_{L^p}, $$ where the last inequality is due to the $L^p$ continuity of the Hilbert transform (the Fourier multiplier $\text{sign} D$) since $ \vert D\vert=D \text{ sign} D $. Above we have used that $\vert D\vert^{-1} N\Omega_N(D)$ is the Fourier multiplication by $$ \vert \xi\vert^{-1} N\mathbf 1 (\vert \xi\vert \ge N), $$ and the Hörmander-Mihlin multiplier theorem gives $L^p$ boundedness. This does not work for $p=1,\infty$.

$\endgroup$

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.