Empirical estimation for discrete-time semi-Markov processes with applications in reliability

We consider a discrete-time semi-Markov process, with a finite state space. Taking a censored history, we obtain empirical estimators for the semi-Markov kernel, semi-Markov transition function, reliability and availability. We study the strong consistency and the asymptotic normality for each estimator.