4
$\begingroup$

This is a question I asked myself today...

$ $

Do you know if it is possible to build a strictly-increasing sequence $(u_n)_{n\in\mathbb{N}^\star}$ of positive integers such that $\displaystyle\sum_{n=1}^{+\infty}\frac{1}{u_n}<+\infty$ and such that for any given $0<\varepsilon<1$, one has $\displaystyle\sum_{n=1}^{+\infty}\frac{1}{u_n^\varepsilon}=+\infty$ ?

$ $

This is looking for a Dirichlet series (with reciprocal of $u_n$ coefficients) with abscissa of convergence $\sigma=1$ and that has a finite limit at $s=1$, does that exist ? It would mean that the corresponding L-function (if the series was extendable around $s=1$) would have no pole at $s=1$. This is tough because the L-function would not be in the Selberg class, and those are hard to study...

$\endgroup$
3
$\begingroup$

You can take $u_n = \lceil n \log^2 n\rceil$: $$ \sum_{n=1}^\infty \frac{1}{n \log^2 n} < \infty $$ but $$ \sum_{n=1}^\infty \frac{1}{n^\varepsilon \log^{2\varepsilon} n} = \infty $$ for every $\varepsilon < 1$.

$\endgroup$
  • $\begingroup$ Oh that's totally true ! I tried $u_n=n\log(n)$ (the PNT makes sure $p_n$ works), but I didn't try that ! Thanks ! $\endgroup$ – Anthony Dec 7 '18 at 2:01
  • $\begingroup$ @Anthony Glad this helped! $\endgroup$ – Clement C. Dec 7 '18 at 2:02

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.