On \(q\)-analogues of quadratic Euler sums (Q2220991): Difference between revisions

For \(\{q, a\}\subseteq\mathbb{R},\;0<q<1\), let \([a, q]:= (q^{a} - 1)(q-1)^{-1}\), the authors define the following ``\(q\) - analogues'' of some classical functions: \[Li_{k}(y):=\sum_{m=1}^{\infty}[m, q]^{-k}y^{m},\; k\in\mathbb{N},\; y\in (-1. 1);\] \[H_{l}(y, a):=\sum_{m=1}^{\infty}[m+a, q]^{-l}y^{m+a},\; l\in\mathbb{N},\; y\in (-1. 1);\] \[f(N; p, z):=\sum_{j=1}^{N}[j, q]^{-p}z^{j},\;\{p, N\}\subseteq\mathbb{N},\; z\in\mathbb{R},\; -1\leq z\leq 1;\] \[S(\mathrm{k}, {\text x}; l, y):=\sum_{m=1}^{\infty}[m, q]^{-l}y^{m}\prod_{j=1}^{n}f(m; k_{j}, x_{j}),\] \[ l\in\mathbb{N},\; y\in (-1. 1),\; \mathrm{k}\in\mathbb{N}^{n},\; \mathrm{x}\in\mathbb{R}^{n},\;\vert x_{i}\vert \leq 1\;\;{\text{for}}\;\;1\leq i\leq n.\] The authors prove a few identities satisfied by those functions and deduce several relations between special values of the Riemann zeta function, the polylogarithms, and the Euler sums, as a consequence of those identities. Their identities, however, are much too complicated to be cited here.