It is well-known that interference alignment (IA) can achieve substantial theoretical gains in multiple-input multiple-output (MIMO) networks. However, the conventional works usually assume the full channel state information (CSI) is available at the transmitter side, which would create a overwhelming CSI feedback overhead for practical frequency-division duplexing (FDD) wireless systems. To implement IA in practice, it is highly desirable to reduce the amount of required CSI feedback overhead. In this paper, we consider IA in MIMO interference multicast networks under partial CSI feedback, and we attempt to minimize the CSI feedback cost subject to IA feasibility constraints with a given degree of freedom (DoF) requirements. First, we propose a CSI feedback profile to embrace two important CSI feedback reduction strategies and we use the metric of CSI feedback dimension to quantify the associated CSI feedback cost.

We then formulate the IA conditions under partial CSI feedback in MIMO interference multicast networks and we derive new IA feasibility conditions under the proposed partial CSI feedback framework. Based on these results, we consider the CSI feedback dimension minimization subject to the IA feasibility constraints with a given DoF requirements in MIMO interference multicast networks, which is formulated as a combinatorial optimization problem. Based on the specific problem structure, we derive an asymptotically optimal solution for a category of network topologies and we further obtain closed-form tradeoff results between DoFs and the CSI feedback cost for MIMO interference multicast networks.