I would like to find solutions of the following differential equation:

$ \sum_{1}^{\infty} a_n f(nx) + f''(x)+ x^2 f(x)=\lambda f(x)$

For example in space of function from $\mathbb R^*$ to $\mathbb C$

If we modify the sum in the differential equation by posing $g(t)=f(e^{t})$, and make a Fourier transform like advise here (for an equation with infinite sum but without derivation) it seems it does not work.

Which method would you suggest ?


If the numbers $b_k=\sum_{n=1}^\infty a_nn^k$ are all fnite, you can find a power series solution: plug $$f(x)=\sum_{k=0}^\infty c_k x^k,$$ and you obtain a recurrency which determines $c_k$: $$c_{k+2}=\frac{1}{(k+1)(k+2)}\left((\lambda-b_k)c_k-c_{k-1}\right).$$ So you can set $c_0$ and $c_1$ arbitrarily, and then determine all $c_k$. To obtain a convergent series you need further assumptions on $a_n$ which will imply a growth estimate of $b_k$.

  • $\begingroup$ Yes, if the $a_n$ decrease extremely rapidely it can provide a solution, but in the case the $a_n$ do not decrease faster than $n^{-k}$ ? for example if for all n we have $a_n=1$ ? Do you think there is a way to obtain a solution using convolution or Fourier transform? $\endgroup$ – Bertrand Dec 20 '14 at 17:56
  • 1
    $\begingroup$ You need some assumption on $a_n$, other wise you have to explain what the first term of your equation means. $\endgroup$ – Alexandre Eremenko Dec 20 '14 at 19:20
  • $\begingroup$ Alexander, first I would like to thank you for your help! Condition on $a_n$ I am interested in can be all $a_n$ are lower than a constant. In this case if there is a solution f then asymptotic of f at infinity is at least in $x^a$ with $a>1$. in fact if it is not possible to find an explicit solution showing there is a solution and finding an asymptotic a infinity will be already a good result for me. Do you know a method to find an asymptotic ? $\endgroup$ – Bertrand Dec 21 '14 at 8:54
  • $\begingroup$ @Bertrand If $a_n$ are just bounded and $f$ grows, how do you understand the sum? (formally the series diverges in this case). $\endgroup$ – fedja Dec 25 '14 at 11:58
  • $\begingroup$ Fedja, Mistake from my side, of course I was meaning f is $x^{-a}$ with $a>1$. $\endgroup$ – Bertrand Dec 28 '14 at 9:28

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.