Empirical spatio-temporal characterization of radio environment properties

During the last decade, the increasing popularity of wireless communications and the exponential growth in the number of wireless devices has raised the question of how the radio environment is changing. Recent developments in the areas of small-cell networks, Internet of Things devices, machine-to-machine communications, and dynamic spectrum access schemes require a better understanding of radio noise characteristics and of spectrum usage in general. The main goal of this thesis is to provide an up-to-date characterization of the radio environment below 6 GHz. Our measurements and analysis are particularly tailored to provide input to radio network researchers, industry, and regulators. The major focus of this thesis is an empirical analysis of spectrum usage based on carefully designed spectrum measurements across Europe. Our measurement campaigns are unique in capturing spectrum use characteristics in depth. This thesis has two major results. First, our concurrent measurements across Europe allow us to study the time dynamics of spectrum use, revealing common characteristics between the different sites at identical times. Second, we study the spatial variations of spectrum use in Paris and London. We use spatial statistics to capture the overall variance and to describe the dynamics present during the measurements. The different measurement approaches we adopted provide complementary views of the statistical characteristics of spectrum use in these high population density study areas. A second theme of the thesis is a noise level study in the contemporary radio environment. Experimental knowledge of radio noise characteristics has not been updated recently; in fact, most high quality analyses are based on measurements conducted decades ago. The way networks are designed and developed has, of course, changed considerably since then. Our measurements capture radio noise by-products at diverse locations stemming from devices that act as intentional, unintentional or incidental radiators. We provide detailed characterization of frequency and time characteristics of the captured noise by-products. We also discuss the issue of the spatial correlation of noise components which can vary from tens to hundreds of meters. The characteristics of the noise by-products also reveal strong deviation from the AWGN noise characteristics typically assumed in literature. We suggest that future receiver designs should assume the constant presence of noise sources with very diverse time characteristics