1
$\begingroup$

Define $\mathfrak{m}$ as the union of the minor arcs of the form $|\alpha-\frac{a}{q}|\leq 1/qQ$, with $(a,q)=1$ and $Q_0<q\leq Q$, with $Q_0\geq N/Q$, for a certain $N\geq Q$ large.

Is it possible to find a sharp upper bound for
$$\int_{\mathfrak{m}}|\sum_{n\leq N}\omega(n)e(n\beta)|^2d\beta$$

for the additive function $\omega(n)$, which counts the number of distinct prime factors of an integer $n$?

For sharp upper bound I mean nothing that uses Parseval's identity. The idea should be to insert the definition of $\omega(n)$ in the exponential sum and then use the structure of the minor arcs in question to obtain a saving compared to the straightforward application of Parseval's identity.

Thanks in advance!

$\endgroup$
2
  • $\begingroup$ I solved it and the answer is going to appear in a forthcoming paper. $\endgroup$ Sep 10, 2019 at 21:15
  • $\begingroup$ cool:) any references to this exponential sum? (maybe your paper). $\endgroup$
    – tomos
    Nov 2, 2023 at 12:37

1 Answer 1

1
$\begingroup$

Not an answer just a small comment: if the singular series coming from the problem of evaluating $$\sum _{n-n=0}h(n)h(n')$$ with the circle method is equal to $$\int _0^1\left |\sum_{n\leq N}h(n)e(n\beta)\right |^2d\beta$$ then you can write your minor arc contribution as $$\left (\int _0^1-\int _\mathfrak M\right )\left |\sum_{n\leq N}h(n)e(n\beta)\right |^2d\beta ,$$ so you just need to do the major arcs. (Maybe you know this already, but just in case.)

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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