Uniformly Lipschitzian mappings in modular function spaces (Q5945239)

Let \(\Omega\) be a nonempty set and \(\Sigma\) be a nontrivial \(\sigma\)-algebra of subsets of \(\Omega\). Let \({\mathcal P}\) be a \(\delta\)-ring of subsets of \(\Sigma\), such that \(E\cap A\in{\mathcal P}\) for any \(E\in{\mathcal P}\) and \(A\in\Sigma\). By \({\mathcal E}\) we denote the linear space of all simple functions with supports from \({\mathcal P}\). By \({\mathcal M}\) we will denote the space of all measurable functions, i.e. all functions \(f:\Omega\to{\mathcal R}\) such that there exists a sequence \(\{g_n\} \in{\mathcal E}\), \(|g_n|\leq|f|\) and \(g_n (\infty)\to f(\omega)\) for all \(\omega\in\Omega\). A functional \(\rho:{\mathcal E}\times \Sigma\to (0,\infty)\) is called a functional modular if: \((\text{P}_1)\) \(\rho(0,E)=0\) for any \(E\in\Sigma\), \((\text{P}_2)\) \(\rho(f,E) \leq\rho (g,E)\) whenever \(|f(\omega)|\leq|g(\omega)|\) for any \(\omega\in\Omega\), \(f,g\in {\mathcal E}\) and \(E\in \Sigma\), \((\text{P}_3)\) \(\rho (f,\cdot)\): \(\Sigma\to (0,\infty)\) is a \(\sigma\)-subadditive measure for every \(f\in {\mathcal E}\), \((\text{P}_4)\) \(\rho(\alpha, A)\to 0\) as \(\alpha\) decreases to 0 for every \(A\in\varphi\), where \(\rho(\alpha, A)=\rho (\alpha 1_A,A)\), \((\text{P}_5)\) if there exists \(\alpha>0\) such that \(\rho(\alpha, A)=0\), then \(\rho(\beta, A)=0\) for every \(\beta>0\), \((\text{P}_6)\) for any \(\alpha>0\), \(\rho(\alpha, \cdot)\) is order continuous on \({\mathcal P}\), that is \(\rho(\alpha, A_n)\to 0\) if \(\{A_n\} \in{\mathcal P}\) and decreases to \(\emptyset\). A set \(E\) is said to be \(\rho\)-nul if \(\rho(\alpha, E)=0\) for every \(\alpha>0\). For \(\rho (f,\Omega)\) we write \(\rho(f)\). The modular \(\rho\) defines a corresponding modular space, i.e. the vector \(L\rho\) given by \(L\rho= \{f\in {\mathcal M};\rho (\lambda f)\to 0\) as \(\lambda\to 0\}\). A function modular is said to satisfy the \(\Delta_2\)-condition if \(\sup\{\rho (2f_n,D_k): n\geq 1\}\to 0\) as \(k\to\infty\) whenever \(\{f_n\}_{n\geq 1}\subset {\mathcal M}\), \(D_k\in\Sigma\) decreases to \(\emptyset\) and \(\sup\{\rho (f_n,D_k): n\geq 1\}\to 0\) as \(k\to\infty\). The sequence \(\{f_n\}\subset L\rho\) is said to be \(\rho\)-a.e. convergent to \(f\in L\rho\) if the set \(\{\omega\in \Omega:f_n (\omega) \nrightarrow f(\omega)\}\) is \(\rho\)-nul. A subset \(C\) of \(L\rho\) is called \(\rho\)-a.e. sequentially compact if every sequence in \(C\) has a \(\rho\)-a.e. convergent subsence in \(C\). A subset \(C\) of \(L\rho\) is called \(\rho\)-bounded if \(\delta_\rho (C)=\sup \{\rho(f-g:f,g \in C\} <\infty\). Let \(B\) be a bounded subset of \(L\rho\). The \(\rho\)-ball of center \(f\in L_\rho\) and radius \(r>0\) is \(B(f,r)= \{g\in L_\rho\), \(\rho(g-f)\leq r\}\). The admissible hull of \(B\) is the intersection of all \(\rho\)-ball containing \(B\) and is denoted by \(\text{ad}(B)\). \(B\) is said admissible if \(\text{ad}(B)=B\). The normal structure coefficient \(\widetilde N(L\rho)\) of \(L\rho\) is \(\widetilde N(L\rho)= \sup\{R(B)/ \delta(B)\), \(B\) is admissible, \(\rho\)-bounded and \(\rho\)-a.e. sequentially compact\}, where \(r(f,B)= \sup\{\rho (f-g)\), \(g\in B)\}\); \(\delta(B)= \sup\{r(f,B), f\in B\}\) and \(R(B)=\inf \{r(f, B),f\in B\}.\) The main result of this paper is the following theorem: Let \(\rho\) be a convex function modular satisfying the \(\Delta_2\) condition and \(B\) an admissible \(\rho\)-a.e. sequentially compact and \(\rho\)-bounded subset of \(L\rho\). Suppose that \(\widetilde N(L\rho)<1\) and \(T:B\to B\) be a \(k\)-uniformly Lipschitzian mapping satisfying \(k<(\widetilde N(L\rho))^{-1/2}\). Then, \(T\) has a fixed point. In the last section of the paper the authors show a class of modular spaces where \(\widetilde N(L\rho)<1\) and so, the above theorem can be successfully applied.