Spatial Statistics of Spectrum Usage: From Measurements to Spectrum Models

Several measurement studies have found a large amount of underutilized radio spectrum. More flexible regulation employing dynamic spectrum access (DSA) has been proposed as solution to this problem. The analysis of several aspects of DSA systems, e.g., cooperative sensing, requires good spatial models of spectrum usage. However, only very focused models such as propagation or shadowing correlation models exist. In this paper we apply techniques developed by the spatial statistics community to the modelling of spectrum. In more detail, we use random fields and the semivariogram to describe the spatial correlation of spectrum usage. We extract parameters from extensive real-life measurements for multiple wireless technologies. These parameter sets enable other researchers to use the model for different tasks ranging from theoretical to simulation-based studies.