Reliability of Spectrum Sensing Under Noise and Interference Uncertainty

Spectrum sensing is a fundamental problem in cognitive radio. As a result, it has been reborn as a very active research area in recent years despite its long history. Although various sensing methods have been proposed, their reliability at very low signal-to-noise ratio (SNR) and noise/interference varying environment remains to be investigated. In this paper, the noise power and interference uncertainty models are discussed first. Subsequently the reliability of single sensor is studied and its performance is quantized. Then cooperative sensing schemes are reviewed and their capability to combat noise power and interference uncertainty is analyzed. It is mathematically proved that cooperative sensing can alleviate the adversary impact of the uncertainty, but cannot eliminate it completely.