In some configurations, the expected KPI includes one of an averaged number of multiple-input multiple-output (MIMO) layers used by the transceiver, an expected signal-to-interference-plus-noise ratio (SINR), an expected reference signal received power (RSRP), or an expected channel quality indicator (CQI). Obtaining the expected KPI may include querying a data source stored on the memory hardware. Here, the data source includes a mapping of: the current location to corresponding CCs that the transceiver previously connected to at the current location while operating in a non-carrier aggregation mode; and the corresponding CCs to corresponding expected KPI values. In another example, obtaining the expected KPI may include querying a data source stored on a distributed storage system in communication with the UE device. In this example, the data source includes a mapping of: the current location to corresponding CCs that a pool of historical UEs previously connected to at the current location; and the corresponding CCs to corresponding averaged KPI values. In some implementations, each historical UE includes substantially a same number of available transceiver resources as the UE device.

Allocating the available transceiver resources to the available CCs may include ordering the available CCs from the available CC associated with the highest expected KPI to the available CC associated with the lowest expected KPI, and allocating more transceiver resources to the available CC associated with the highest expected KPI than to the available CC associated with the lowest expected KPI.