9
$\begingroup$

Let $ c_{n,k} $ be the Simsun permutations$^1$ defined by the following relations: $\displaystyle c_{n,0} = 1, \hspace{0.1cm} (n \ge 1);$ $$ c_{n,k} = (k+1) c_{n-1,k} +(n-2k+1) c_{n-1,k-1}, \hspace{0.5cm} (1 \leq k \leq \lfloor n/2 \rfloor);$$ and $ c_{n,k} = 0, \hspace{0.1cm} ( k> \lfloor n/2 \rfloor). $

Now, let $n=2p.$ I am trying to find the value of \begin{eqnarray} A_p:=\sum_{k=0}^{p} c_{2p,k} \hspace{0.1cm} \frac{(2p-2k)!}{2^{p-k} (p-k)!}. \end{eqnarray} or at least a sharp upper bound for it. We know$^2$ that $$ \sum_{k=0}^{p} c_{2p,k} \hspace{0.1cm} 2^{2p-k} = (2p+1)!.$$

From this identity we can easily obtain the bound $$ A_p \leq 2^{-p} p!(2p+1)! $$ for $ A_p $ which is a big upper bound.

Additionally, we know$^2$ that $ \sum_{k=0}^{p} c_{2p,k} = T_{p+1}$, where $ T_n= \frac{2^{2n}(2^{2n}-1) |B_{2n}|}{2n} $ is the sequence of tangent numbers$^3$ (defined by the Bernoulli numbers $B_n$), appearing in the Taylor series expansion of tan($x$): $$\text{tan}(x)=\sum_{n=1}^{\infty} T_n \frac{x^{2n-1}}{(2n-1)!}.$$

Motivation: In a part of my research (in quantum statistical mechanics) I need to show convergence of a series. I have reduced the initial problem to finding the value of $ A_p $, or at least a good upper bound for $ A_p $.

Any hint or idea would be greatly appreciated! Thanks in advance!

  1. For Andre and Simsun permutations see here and here.
  2. See here for the paper "increasing trees and alternating permutations" by G. Kuztensov, I. Pak, and A.E. Postnikov. In this paper, the Andre permutations are denoted by $ d_{n,k} $.
  3. See here.
$\endgroup$
3
  • $\begingroup$ I just noticed (by computer) that $f(x) = \sum_{p,k} c_{p,k} t^k \frac{x^p}{p!}$ seems to satisfy $f'' - 3t f^2 + (2t-1)f = 0$. I have no idea whether that might be helpful, though. $\endgroup$ Feb 20, 2016 at 21:52
  • $\begingroup$ @Martin Thank you very much Martin! Indeed, I need to prove the convergence of the series $$ \sum_{p=0}^{\infty} \frac{x^{2p}}{(2p)!} A_p, $$ which resembles to your function $f(x)$. Then I need to see if I can obtain more information from your proposed equation. Thanks again! I'll be back soon! $\endgroup$
    – user83150
    Feb 22, 2016 at 12:50
  • $\begingroup$ The last series diverges, considering some special cases in my initial research question. $\endgroup$
    – user83150
    Aug 23, 2016 at 18:05

0

Your Answer

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