4
$\begingroup$

Prove that if $\displaystyle \sum_{n=1}^ \infty a_n$ converges, and {$b_n$} is bounded and monotone, then $\displaystyle \sum_{n=1}^ \infty a_nb_n$ converges.

No, $a_n, b_n$ are not necessarily positive numbers.

I've been trying to use the Dirichlet's Test, but I have no way to show that $b_n$ goes to zero. If I switch $a_n$ and $b_n$ in Dirichlet's Test, I can show $a_n$ goes to zero, but then I'm having trouble showing that $\displaystyle \sum_{n=1}^ \infty \left\lvert a_{n+1}-a_{n}\right\rvert$ converges (because $a_n$ isn't necessarily monotone).

$\endgroup$
  • 1
    $\begingroup$ Use abel partial summation. $\endgroup$ – Math Wizard Feb 5 '16 at 0:31
  • 1
    $\begingroup$ I have, that is what allowed me to show that the partial sums of ∑anbn converge. But, I still need to satisfy the other condition(s). It's these other conditions am having difficulty with. $\endgroup$ – Jill_Johnson Feb 5 '16 at 0:56
  • $\begingroup$ Since {$b_n$} is bounded and monotone, it has a Max/Min (both cases are analogous). So let $M$ be a lower/upper bound (which is finite) for {$b_n$}, then $M\cdot\sum a_{n}$ converges... shouldn't be hard from there, or I could be missing something? $\endgroup$ – mm8511 Feb 5 '16 at 1:25
  • $\begingroup$ @mm8511 With this argument, you don't show convergence of the series, only that it's bounded. $\endgroup$ – Clement C. Feb 5 '16 at 1:41
7
$\begingroup$

Outline: Indeed, the key is use Dirichlet's test (a.k.a. Abel's summation at its core) as you intended:

$$\begin{align} \sum_{n=1}^N a_n b_n &= \sum_{n=1}^N (A_n-A_{n-1}) b_n = A_Nb_N + \sum_{n=1}^{N-1} A_n b_n -\sum_{n=1}^{N-1} A_n b_{n+1} \\ &= A_Nb_N + \sum_{n=1}^{N-1} \underbrace{A_n}_{\text{bounded}} \underbrace{(b_n -b_{n+1})}_{\text{constant sign}} \end{align}$$ where $A_n \stackrel{\rm def}{=} \sum_{n=1}^{N} a_n$ and $A_0=0$.

Now, this does not quite work: the issue boils down to the fact that at the end of the day you cannot rely on $b_n\xrightarrow[n\to\infty]{} 0$; since indeed it is not necessarily true. But you need this for the argument to go through.

To circumvent that, observe that $(b_n)_n$ is a bounded monotone sequence, and therefore is convergent. Let $\ell\in\mathbb{R}$ be its limit.

You can now define the sequence $(b^\prime_n)_n$ by $b^\prime_n \stackrel{\rm def}{=} b_n-\ell$. This is a monotone bounded sequence converging to $0$, and $$ \sum_{n=1}^N a_n b^\prime_n = \sum_{n=1}^N a_n b_n - \ell \sum_{n=1}^N a_n. $$ The second term is a convergent series by assumption on $(a_n)_n$, so showing convergence of the series $ \sum a_n b_n$ is equivalent to showing convergence of the series $\sum a_n b^\prime_n$. Apply your idea (Abel's summation) on the latter.

$\endgroup$
  • $\begingroup$ That's definitely what I was looking for. Thank you so much. $\endgroup$ – Jill_Johnson Feb 5 '16 at 2:15
  • $\begingroup$ Glad this was helpful... that's a nice question. $\endgroup$ – Clement C. Feb 5 '16 at 2:16

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.