In this paper, we propose a resource allocation scheme for interference management in heterogeneous networks. We particularly consider downlink interference from a Home eNB (HeNB) to macrocell user equipments (MUEs) in the coverage of the traditional macrocell basestation (MBS). By overhearing uplink feedback information from MUEs and Home User Equipments (HUEs) together with downlink control information (DCI) from the MBS, the HeNB formulates a dynamic programming problem with the objective of maximizing the total reward over an infinite horizon. By exploiting the feedback information as well as the DCI at the HeNB, we show that the solution of the dynamic programming problem follows a greedy policy at the HeNB, which simplifies the solution of the infinite horizon dynamic programming problem. We also examine the effect of uncertainty in the feedback information on the performance of our proposed scheme.