In order to apply emerging technologies (e.g. dynamic spectrum sharing) to address the wireless spectrum shortage problem, there is a critical need to understand global RF spectrum usage trends. To accomplish this, a three-pronged approach is being pursued: 1) deployment of geographically dispersed, temporally coordinated RF spectrum observatories in multiple U.S. locations, and through international collaboration, in Finland. The spectrum observatories use a common platform generating a single RF spectrum measurement dataset. 2) Development of empirically validated, statistical models of spectrum utilization for different wireless application types based on this dataset. 3) Use of "big data" analytical techniques to mine the dataset to discover temporal and spectral correlations not obvious using traditional approaches. As the models and relationships are refined, they will enable temporal and spectral occupancy predictions to support spectrum sharing for various circumstances and wireless applications.
The generation of a high-resolution, multi-location, multi-national spectrum usage dataset using a common, consistent measurement and storage approach is unique and allows direct, unambiguous comparisons of spectrum usage across geographies and demographics. The statistical models of spectrum utilization and the identified similarities and differences between regions and wireless services are unique and inform dynamic spectrum sharing research and related regulatory action with "real-world" data. Importantly, this is the first time that "big data" analytic approaches are being systematically applied to RF utilization data providing new insights motivating novel dynamic spectrum sharing approaches and improved spectrum efficiency.
