The Complexity as a Criterion for Blind Spectrum Sensing in Cognitive Radio

Spectrum sensing is one of the most important issues in cognitive radio technology. It is because the unlicensed secondary user, to avoid interference, after detecting licensed primary users, should leave the spectrum or move to another idle spectrum resource. Considering the fact that received signal in terms of presence or absence of the primary user has different complexity. In this paper, a new algorithm for detection of primary users is presented. Our proposed method is based on Sample entropy which is criterion for time series complexity. This kind of measurement of spectrum due to lack of need to preliminary information through the channel, primary user and noise variance, is called blind spectrum sensing. In simulation section, it is shown that for sufficiently large data lengths, this method in comparison with similar methods like approximate entropy, spectral entropy and Shannon entropy has a better performance.