Embodiments of the disclosure relate to a massive multiple-input multiple-output (M-MIMO) wireless distribution system (WDS) and related methods for optimizing the M-MIMO WDS. In one aspect, the M-MIMO WDS includes a plurality of remote units each deployed at a location and includes one or more antennas to serve a remote coverage area. At least one remote unit can have a different number of the antennas from at least one other remote unit in the M-MIMO WDS. In another aspect, a selected system configuration including the location and number of the antennas associated with each of the remote units can be determined using an iterative algorithm that maximizes a selected system performance indicator of the M-MIMO WDS. As such, it may be possible to optimize the selected system performance indicator at reduced complexity and costs, thus helping to enhance user experiences in the M-MIMO WDS.