The Presence of Weakly Correlated Noise over Fading Channels on Signal Detection

Cognitive radio is a technology which allows the secondary usage of spectrum allocated to primary users. The cognitive radio networks are networks that have cognitive and reconfigurable properties and the capability to detect the unoccupied spectrum holes and change frequency for end to end communication. In CRNs in order to avoid interference to the primary users efficient and reliable spectrum sensing is necessary. In the case where noise samples are correlated impairments due to independent noise samples must be considered. To address this issue, Locally Optimum (LO) detection technique is utilized. The performance of LO detector depends mainly on false alarm probability and detection probability. The probabilities of false alarm and detection at different SNRs are analyzed. The simulation results demonstrate the superiority of LO detector over known energy detection method. Furthermore, the scenario where estimated correlation is different from real correlation is considered for detailed study in this paper.