Effect of noise uncertainty during spectrum sensing for cognitive radio ad hoc networks

Cognitive radio network is a technique where underutilised licensed spectrum allows opportunistic access to unlicensed secondary users. Different spectrum sensing algorithms are used to sense the presence or absence of primary licensed users in noisy and fading channel. This paper provides a comparative analysis of different spectrum sensing algorithms for cognitive radio network based on dynamic threshold and noise uncertainty. Spectrum sensing in cognitive radio network based on fixed threshold are sensitive to noise, therefore these are not efficient enough. The use of dynamic threshold improves performance of the spectrum detection without much computational complexity. Different computer simulations are provided for the dynamic threshold with and without noise uncertainty for spectrum sensing of cognitive radio network. The performance of the spectrum sensing techniques is evaluated by receiver operating characteristics (ROC) curves over additive white Gaussian noise (AWGN) and different fading channels (Rayleigh and Nakagami-m).