MathOverflow is a question and answer site for professional mathematicians. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Let $s>n/2, \; f \in W^{s,2}(\Bbb R^n)$ . Then how can I show that there is an embedding into the space of uniformly bounded, continuous functions, that is, $$ |f(x)| \leqslant C\| f \|_{W^{s,2}}$$ for almost all $x \in \Bbb R^n$ ?

I think the general Sobolev embedding theorem cannot be applied in this case because the domain is $\Bbb R^n$ which is open.

share|cite|improve this question
You should try to do it yourself. It's a good homework exercise. – Deane Yang Oct 13 '12 at 23:19
can you please explain what you call a "uniformly bounded continuous function"? or do you mean bounded uniformly continuous functions? How is the norm you consider different from that of $C_b(\mathbb R^n)$? – Delio Mugnolo Oct 13 '12 at 23:24
@DelioMugnolo It means that uniformly bounded and continuous functions, $f \in C_b \cap C^0$ I think. – Andrew Oct 13 '12 at 23:27
A global $W^{s,2}$ bound implies more than the Sobolev theorem for a bounded domain, because it not only implies continuity but also implies that the function has to decay fast enough at infinity. – Deane Yang Oct 13 '12 at 23:38
up vote 3 down vote accepted

This is very standard, but perhaps buried in fancier things:

The Fourier transform $\hat{f}$ of $f$ is in $L^2$ for the weight $(1+|x|^2)^s$. Since $s>n/2$, the constant function $1$ is in that weighted $L^2$ space, so by Cauchy-Schwarz-Bunyakowsky, $\int_{\mathbb R^n} |\hat{f}| = \int |\hat{f}|\cdot 1\le |\hat{f}|_s \cdot |1|_t$, where $t=-s$ [to avoid TeX bug?] and the subscript denotes the weight. Thus, $\hat{f}$ is in $L^1$. Thus, by Fourier inversion and Riemann-Lebesgue, $f=\hat{\hat{f}}$ is uniformly continuous and goes to $0$ at infinity.

share|cite|improve this answer
Dear Paul: presumably $1$ has the weight $-s$ instead of the weight $s$? – Willie Wong Nov 5 '12 at 10:29
Oops, thanks, Willie. – paul garrett Nov 5 '12 at 13:20

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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