Abstract

In this paper, a cluster-based spectrum sensing strategy is proposed during the low traffic periods of primary user (PU) with the aim of minimizing the average number of sensing bits for secondary users (SUs) while keeping the missing probability below a predefined threshold. In the clustering phase, we introduce a new distance-and-energy-relevant-clustering-algorithm (DERCA) which is based on the distances between the PU and SUs acquired by the trigonometry theory and the remaining energies of SUs. In the next spectrum sensing phase, without need for any coordination among the cluster-member SUs, we increase the reliability of the spectrum sensing strategy by using judging credibility in the energy detection followed by feature detection. Performance evaluations reveal that the proposed strategy can decrease sensing bits consumption greatly and achieve lower missing detection probability at the expense of a little sensing delay performance loss.

Paper Details