Adaptive threshold based energy detection for cognitive radios

Spectrum sensing is one of the most crucial tasks a cognitive radio has to carry out in order to decide whether a licensed spectrum can be used without causing interference to the licensed or the Primary Users (PU). Many techniques exist in the literature for satisfactory detection performance. However, in many cases, an additive white Gaussian noise is assumed, which is an oversimplified assumption and in many cases not suitable for modeling true practical scenarios. In such cases, usually noise uncertainty is neglected. In this work, noise uncertainty is considered and a new detection technique based on adaptive threshold is proposed. Simulation results reveal that the proposed techniques show better Receiver Operator, Characteristics compared to the existing techniques.