One of the most critical issues in femtocell network deployment is interference management, especially for femtocells sharing the spectrum occupied by conventional cellular networks. In this work we propose and analyze an optimal power allocation strategy based on modeling the interferer’s activity as a two-state Markov chain. In the single femto-user access, we show how to maximize the expected value of femto-user rate, averaged over the interference statistical model. Then, we extend the approach to the multiuser case, adopting a game-theoretic formulation to devise decentralized access strategies, particularly suitable in view of potential massive deployment of femto access points.