On modular firmly nonexpansive mappings in the variable exponent sequence spaces \(\ell_{p(\cdot)}\) (Q2659599)

The authors consider the question of existence of fixed points of certain mappings on subsets of the sequence space \[ \ell_{p(\cdot)} = \Big\{(x_n)_{n\in\mathbb{N}}\colon \sum_{n=0}^{\infty} \frac{1}{p(N)} |\lambda x_n|^{p(n)} < \infty \;\text{for some}\; \lambda>0\Big\}. \] A convex modular is a function \(\rho\colon \ell_{p(\cdot)} \to [0,\infty]\) satisfying the conditions \begin{enumerate} \item \(\rho(x)=0\) if and only if \(x=0\) \item \(\rho(\gamma x) = \rho(x)\) for \(|\gamma|=1\) \item \(\rho(\alpha x + (1-\alpha) y) \leq \alpha \rho(x) + (1-\alpha) \rho(y)\) for \(\alpha\in [0,1]\) \end{enumerate} for all \(x,y\in\ell_{p(\cdot)}\). A mapping \(T\colon D\to \ell_{p(\cdot)}\) with \(D\subset\ell_{p(\cdot)}\) is called \(\rho\)-nonexpansive if it satisfies \[ \rho(Tx-Ty) \leq \rho(x-y) \] for \(x,y\in D\). It is called \(\lambda\)-\(\rho\)-firmly nonexpansive for \(\lambda\in (0,1)\) if it satisfies \[ \rho(Tx-Ty) \leq \rho((1-\lambda)(x-y)+\lambda (Tx-Ty)) \] for all \(x,y\in D\). The main results of this paper are two fixed point theorems for \(\lambda\)-\(\rho\)-firmly nonexpansive mappings on certain subsets of \(\ell_{p(\cdot)}\) under the additional assumption that \(\rho\) is strictly convex.