Zero-sum ergodic semi-Markov games with weakly continuous transition probabilities (Q1028603)

The author extends previous results related to zero-sum Markov games with Feller transition probabilities onto zero-sum ergodic semi-Markov games of the same type. The line of reasoning is based on the application of the Banach fixed point theorem in an appropriate space of lower semi continuous functions. The proof of the existence of a solution to the optimality equation is preceded by some regularization procedure. The author considers two types of payoff criteria i.e. the ratio average and the time average and it is proved that both have the same value and one player possesses an \(\varepsilon\) -- optimal strategy and the other -- an optimal stationary strategy. Moreover two examples of stochastic games with weakly continuous transition probabilities are discussed.