Compressive spectrum sensing augmented by geo-location database

In cognitive radio (CR), white space devices (WSDs) need to have the knowledge of spectrum occupancy in TV white space (TVWS) before dynamic access. There are two common schemes proposed to achieve this: 1) geo-location database and 2) spectrum sensing. In geo-location database, calculating digital terrestrial television (DTT) location probability and maximum permitted power in each channel in an efficient way becomes important as the database is supposed to give a quick response once a request comes. Spectrum sensing is a scheme which can provide a more reliable and real-time results for spectrum occupancy. However, the high sampling rate is a big challenge in spectrum sensing for power limited WSDs. In this paper, we proposed to combine the location probability based geo-location database with compressive sensing (CS) based spectrum sensing to achieve sub-Nyquist sampling rates for WSDs. The history data from geo-location database is utilized to support the signal recovery for the spectrum sensing. In addition, a new method to calculate DTT location probability efficiently is proposed. Theoretical analysis of the proposed algorithm are tested in TVWS and it shows that performance of the proposed algorithm outperforms the traditional algorithm.