$$\sum_{n>0} \frac{\mu(n)}{n^s}$$

Sum from 1 to infinity of The Möbius function$/n^s$, i.e., Möbius function/Riemman-zeta function?

Sorry, I forgot to mention that the way that I am suppose to tackle this question is with the information that: L(s,f[convolution]g) = L(s,f)*L(s,g).

The problem was that I couldn't use the mobius inversion theorem or anything to try and find the mobius function in terms of a convolution? Any help? thanks.

  • $\begingroup$ Just a minor point: what you have written is not "Möbius/Zeta," although it sort of looks like it. $\endgroup$ – angryavian Jan 11 '14 at 18:32

Hints for you to complete and prove:

Put $\;\mathcal P:= \{p\in\Bbb N\;;\;p\;\text{is a prime}\}\;$ :

$$\text{Re}(s)>1\;:\;\;\zeta(s)=\prod_{p\in\mathcal P}\left(1-\frac1{p^s}\right)^{-1}\implies$$

$$\frac1{\zeta(s)}=\prod_{p\in\mathcal P}\left(1-\frac1{p^s}\right)=1-\sum_{p\in\mathcal P}\frac1{p^s}+\sum_{p,q\in\mathcal P\,,\,p\neq q}\frac1{p^sq^s}-\sum_{p_1,p_2,p_3\in\mathcal P\,,\,i\neq j\implies p_i\neq p_j}\frac1{p_1^sp_2^sp_3^s}+\ldots$$



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.