Distribution of Success Runs for Markov Dependent Bernoulli Trials

For a sequence of Markov dependent Bernoulli trails, asymptotic normality of the number of success runs with length $r \geq 1$ is established, when the number of trials tends to infinity.