This paper presents a generalized closed-form beamforming technique that can achieve the maximum degrees of freedom in compounded multiple-input multiple-output (MIMO) broadcast channels with mixed classes of multiple-antenna users. The contribution is first described within the context of a three-cell network and later extended to the general multicell scenario where we also show how to determine the conditions required to align the interference in a subspace that is orthogonal to the one reserved for the desired signals. This is then followed by an analysis of the impact of antenna correlation for different channel state information acquisition models. The proposed scheme is examined under both conventional and large-scale MIMO systems. It will be shown that the proposed technique enables networks with any combination of user classes to achieve superior performance even under significant antenna correlation, particularly in the case of the large-scale MIMO systems.
- Antenna correlation
- broadcast channel (BC)
- channel state information (CSI)
- degree of freedom (DoF)
- Interference alignment (IA)
- Multiple-input multiple-output (MIMO)