0

Given the following equation

$$\frac{J_{n - 1} (u)}{uJ_n (u)} - \frac{K_{n-1}(w)}{wK_n(w)} = 0$$

(where $J$ is the Bessel function of the first kind, $K$ is the modified Bessel function of the second kind, $u, w \in \mathbb{R}$, $u \neq w$ in general)

it should be proved that it is equivalent to

$$\frac{J_{n - 1} (u)}{uJ_{n - 2} (u)} + \frac{K_{n-1}(w)}{wK_{n-2}(w)} = 0$$

by using some recurrence relations (as the ones presented in this document).

By comparing the two (allegedly equivalent) equations, it should be

$$\frac{J_{n - 1} (u)}{J_{n} (u)} = \frac{J_{n - 1} (u)}{J_{n - 2} (u)}$$

$$\frac{K_{n - 1} (w)}{K_{n} (w)} = - \frac{K_{n - 1} (w)}{K_{n - 2} (w)}$$

If it is right, how to proceed now?

The equality should be proved for any integer $n \geq 0$ and at least for any non-negative real values $u,v$. It is used to describe the optical fibers propagation (in documents like this, to obtain (17.104) from (17.103)).

BowPark
  • 1,366
  • The last two formulas cannot be valid for general $n,u,w$ because they would imply $J_{n-2}(u)=J_n(u);$ and $K_{n-2}(w)=-K_n(w),,$ which is obviously wrong. Please make clear what is intended: Do you seek for special values $n,u,w$ or shall the formulas be valid for $n \in \mathbb{N}$ etc? – gammatester Feb 05 '16 at 11:17
  • @gammatester I just edited the question to (hopefully) make it more clear. – BowPark Feb 05 '16 at 15:04
  • 1
  • @Lucian Thank you. I think it is not directly useful here, but it is certainly a similar problem. – BowPark Feb 05 '16 at 21:17

0 Answers0