A novel threshold selection for hypotheses tests in cognitive radios

Spectrum sensing is a central part in a cognitive radio and therefore a thorough design of a spectrum sensor is of highest priority. Here, a novel implementation of a spectrum sensing technique, using a generalized version of the Neyman Pearson test, is considered. This novel implementation overcomes the usual prerequisite of the known noise power. This benefit is achieved by normalizing the signal and accurately calculated the threshold for that particular hypothesis which reflects the absence of the useful signal, usually termed hypothesis H2. It is shown by simulation results that the performance of the novel detector is promising.