A statistical weighting average approach for cognitive radio networks

In this paper, a statistical weighting average (SWA) approach is proposed for cooperative spectrum sensing (CSS) system in Cognitive Radio Networks (CRNs). Two SWA algorithms, including ordered weighted average (OWA) and weighted ordered weighted average (WOWA), are initially designed in CRNs. In a weighting process, the final weighting factor (WF) for a distributed sensor is jointly determined by its previous sensing contributions and its currently gathered primary user (PU) samples. The WF is initially calculated by the sensing contributions, which are numerically calibrated by probability of detection. The gathered PU samples are used to further adjust the corresponding WF. Simulation results show that the proposed SWA approach can outperform the conventional weighting schemes.