This paper jointly studies interference management and self-organization schemes in a hybrid self-organized small cell (hybrid-SO small cell) network, which is a combination of the centralized and distributed small cell networks. From the view of downlink transmission, we formulate a resource optimization based interference management (ROIM) problem. In particular, to accommodate to the hybrid architecture, the problem is divided into two subproblems. The first is subchannel allocation, which is investigated with overall network interference alleviation under the assumption of imperfect channel state information (CSI).

The second is power assignment, which aims at constraining the local interference while guaranteeing the network utility. To solve the ROIM problem, we propose a twostep self-optimization based interference management (TSSOIM) scheme. At the first step, operation and management (O&M) unit adopts aggressive discrete stochastic approximation (ADSA) algorithm to find the proper channel set for users. At the second step, small cells employ game theory based Kalai-Smorodinsky bargaining solution (KSBS) to assign power to the channels. Simulation results are presented to show the effectiveness of the proposed TSSOIM scheme.