Title:Kirk's Fixed Point Theorem in Complete Random Normed modules

Abstract:Recently, stimulated by financial applications and $L^0$--convex optimization, Guo, this http URL introduced the notion of $L^0$--convex compactness for an $L^0$--convex subset of a Hausdorff topological module over the topological algebra $L^0(\mathcal{F},K)$, where $K$ is the scalar field of real or complex numbers and $L^0(\mathcal{F},K)$ the algebra of equivalence classes of $K$--valued measurable functions defined on a $\sigma$--finite measure space $(\Omega,\mathcal{F},\mu)$, endowed with the topology of convergence locally in measure. A complete random normed module $($briefly, $RN$ module$)$, as a random generalization of a Banach space, is just such a kind of topological module, this paper further introduces the notion of random normal structure and gives various kinds of determination theorems for a closed $L^0$--convex subset to have random normal structure or $L^0$--convex compactness, in particular we prove a characterization theorem for a closed $L^0$--convex subset to have $L^0$--convex compactness, which can be regarded as a generalization of the famous James characterization theorem for a closed convex subset of a Banach space to be weakly compact. Based on these preparations, we generalize the classical Kirk's fixed point theorem from a Banach space to a complete $RN$ module as follows: Let $(E,\|\cdot\|)$ be a complete $RN$ module and $V \subset E$ a nonempty $L^0$--convexly compact closed $L^0$--convex subset with random normal structure, then every nonexpansive self--mapping $f$ from $V$ to $V$ has a fixed point in $V$. The generalized Kirk's fixed point theorem is also of fundamental importance in random functional analysis, for example, we can derive from it a very general random fixed point theorem, which unifies and improves several known random fixed point theorems for random nonexpansive mappings.