Performance of a TV white space database with different terrain resolutions and propagation models

Cognitive Radio has now become a realistic option for the solution of the spectrum scarcity problem in wireless communication. TV channels (the primary user) can be protected from secondary-user interference by accurate prediction of TV White Spaces (TVWS) by using appropriate propagation modelling. In this paper we address two related aspects of channel occupancy prediction for cognitive radio. Firstly we investigate the best combination of empirical propagation model and spatial resolution of terrain data for predicting TVWS by examining the performance of three propagation models (Extended-Hata, Davidson-Hata and Egli) in the TV band 470 to 790 MHz along with terrain data resolutions of 1000, 100 and 30 m, when compared with a comprehensive set of propagation measurements taken in randomly-selected locations around Hull, UK. Secondly we describe how such models can be integrated into a database-driven tool for cognitive radio channel selection within the TVWS environment