A Model for Spectrum Valuation

The effective use of radio frequency (RF) spectrum has become more important as industry and government continue to increase their demand for bandwidth to enable new and improved wireless communications services. However practical supply of the resource is finite. Determining the market value of spectrum is a key step for stakeholders and decision makers to understand the impacts of RF spectrum investments, tradeoffs, and alternatives.
¹
To meet this need, The MITRE Corporation (MITRE) developed a methodology and valuation model to assess the potential market value of various RF spectrum bands. The intent of the model is to allow non-industry experts to generate a range of possible valuations rather than a single point estimate and to inform decision makers about where the largest uncertainties exist that could influence the valuations. The spectrum valuation tool uses a novel approach to generate spectrum valuation. It integrates multiple regression models across the various factors that influence value, including transaction factors, network infrastructure factors, geographic factors, policy/economic factors, and consumer factors. This ensemble technique compensates for uncertainty between individual regression algorithms by integrating the results of a “group of experts” together to build a range of outcomes. This paper describes the methodology used to develop the spectrum valuation tool, the data included in the regression models, and the model's prediction capability with respect to the recently held Federal Communication Commission's 28 Gigahertz (GHz) auction.