0
$\begingroup$

Suppose we consider in $\mathbb R^n$, then how to show $\Vert f \Vert_{L^{p}} \leq C\Vert \nabla^{s}f \Vert_{L^{q}}^{\alpha}$, where $s>0$ is noninteger and $\alpha \in (0,1)$?

$\endgroup$
  • $\begingroup$ Can you be more precise: what are $p$ and $q$ and what do you mean by $\nabla^s f$? $\endgroup$ – Piotr Hajlasz Nov 21 '18 at 14:31
2
$\begingroup$

Such an inequality cannot be true unless $f=0$. That can be proved by a standard homogeneity argument. Suppose that $\Vert \nabla^sf\Vert_p<\infty$. Replacing $f$ by $tf$, where $t>0$ we have $$ t\Vert f\Vert_p=\Vert tf\Vert_p\leq C\Vert \nabla^s(tf)\Vert_p^\alpha=t^\alpha C\Vert \nabla^sf\Vert_p^\alpha, \quad \Vert f\Vert_p\leq t^{\alpha-1} C\Vert \nabla^sf\Vert_p^{\alpha}, $$ Since $\alpha-1<0$, letting $t\to\infty$, the right hand side will converge to $0$ and hence $\Vert f\Vert_p=0$, $f=0$.

$\endgroup$
1
$\begingroup$

I guess that your question is ill-formulated: you have to respect the homogeneity on both sides of the inequality. Let us say that for $f$ in the Schwartz space you always have $$ \Vert f\Vert_{W^{t,q}(\mathbb R^n)}\lesssim \Vert \vert D\vert^s f \Vert_{L^{p}(\mathbb R^n)}\quad\text{(and also}\quad \Vert f\Vert_{L^{n/(n-1)}(\mathbb R^n)}\lesssim \Vert D f\Vert_{L^{1}(\mathbb R^n)}), $$ provided $ \frac{s-t}{n}=\frac{1}{p}-\frac{1}{q},\quad 1<p<q<+\infty. $ Applying the second inequality above to $f=u^2$, you obtain for $p>1$, $$ \Vert u\Vert_{L^{2n/(n-1)}(\mathbb R^n)}^2\lesssim \Vert uD u\Vert_{L^{1}(\mathbb R^n)}\lesssim \Vert u\Vert_{L^{p'}(\mathbb R^n)} \Vert D u\Vert_{L^{p}(\mathbb R^n)} \lesssim \Vert u\Vert_{L^{p'}(\mathbb R^n)} \Vert \vert D\vert u\Vert_{L^{p}(\mathbb R^n)}, $$ so that $$ \Vert u\Vert_{L^{2n/(n-1)}(\mathbb R^n)} \lesssim \Vert u\Vert_{L^{p'}(\mathbb R^n)}^{1/2} \Vert \vert D\vert u\Vert_{L^{p}(\mathbb R^n)}^{1/2}, $$ an inequality resembling yours. Choosing $f= u^\alpha$ leads to more general interpolation inequalities.

$\endgroup$

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

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