University of Hertfordshire

Spectrum sensing is one of the most challenging issues in cognitive radio networks. It provides protection to primary users (PUs) from interference and also creates opportunities of spectrum access for secondary users (SUs). It should be performed efficiently to reduce number of false alarms and missed detection. At the same time, spectrum sensing should be energy efficient to ensure the longevity of cognitive radio devices. This work presents a novel scheme which investigates the usage of history for energy efficient spectrum sensing in infrastructure cognitive radio networks. The presented scheme employs an iteratively developed history processing database. It is shown that usage of history helps predicting PU activity and results into reduced spectrum scanning by SUs thereby improving the sensing related energy consumption.