I'm trying to solve the following problem. Let $f_0 \in L^p(\mathbb{R})$ and let $f_n(x)=f_0(x+n)$. Show that $f_n$ converges weakly to zero in $L^p(\mathbb{R})$.

I know that if $(x_n)$ is a sequence in $X$, then we have the following.
$\forall f \in X^*, f(x_n)$ converges to $f(x)$ iff $(x_n)$ converges to $x$ in the weak topology.

I know the dual space of $L^p$ is $L^q$ where $\frac{1}{p}+\frac{1}{q}=1$.

I've been playing around with some of the well-known properties of $L^p$ spaces but haven't made much progress.


You can argue by density. Given $g\in L^q$, choose a compactly supported $h$ such that $\|g-h\|_q<\epsilon$.

Then you have $$ \int gf(x+n)\,dx=\int h f(x+n)\,dx+\int (g-h) f(x+n)\,dx. $$ The first integral approaches zero as $n\to \infty$ because you are integrating over some bounded interval (the support of $h$) $$ \int h f(x+n)\,dx=\int\limits_{-L}^L h f(x+n)\,dx\le C\int\limits_{-L}^L f(x+n)\,dx= $$ $$ =C\int\limits_{-L+n}^{L+n} f(x)\,dx\le C_1\left[\int\limits_{-L+n}^{L+n} f(x)^p\,dx\right]^{1/p}\to 0\quad\textrm{as} n\to\infty $$ as it is the "tail" of a convergent integral ($C$ is the max of $|h|$ and $C_1$ depends on $L$ here as you are using Holder's inequality).

As for the second integral, it is bounded by $$ \|(g-h)\|_q \|f(x+n)\|_p\le \epsilon\|f\|_p $$ by Holder's inequality. Choosing first $\epsilon$ small and then $n$ large proves your claim.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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