MACHINE LEARNING BETWEEN RADIO LOADING AND USER EXPERIENCE

A resource upgrade predictor can be operable to receive, from a first network node device, traffic information. Based on the traffic information, the resource upgrade predictor can obtain network utilization data related to other network node devices having a similar interference characteristic (e.g., signal-to-noise ratio) to the first network node device. The resource upgrade predictor can use this network utilization data to determine a demand (e.g., demand level, demand point) at which at least a defined value related to a transmission link capacity associated with transmissions between the first network node device and the user equipment, is attained (e.g., a percentage of physical resource block loading). The resource upgrade predictor can also obtain projected demand data associated with a geographic area serviced by the first network node device, and determine, based upon the demand and the projected demand data, a time at which a network resource upgrade related to the first network node device, is to be performed.