Spatial characterisation of spectrum occupancy in urban environment for cognitive radio

Cognitive Radio (CR) has been proposed to solve the so-called spectrum scarcity problem. CR uses opportunistically the licensed bands when they are temporarily unused. The realization of spectrum models and the statistical characterization of spectrum occupancy is an important point to assist the study and the development of various features required to CR. In this context we conducted a measurement campaign in the DCS and UMTS cellular bands in an urban environment. We collected spectral occupation data in order to analyze and characterize the behavior of spatial correlation of the spectrum occupancy. We analyze the spatial correlation by calculating some metrics as a function of the distance. Finally we propose a dierent approach to the calculation of the metrics, that is to characterize spatial correlation as a function of the SNR-dierence at the receiver instead of the distance