Abstract: Clustering is an important technique in cognitive radio networks.There are two reporting channels in the cluster:the channel between the secondary users and the cluster head,and the channel between the cluster head and the fusion center.This research is about the performance limitation when these two reporting channels are not ideal.Under the phenomenon of “BEP Wall”,this paper puts forward the index of “SNR Wall”.First of all,closed-form expressions for the global probabilities of detection and false alarm are derived for non identical SNR on the listening channels and non ideal reporting channels.Then the existence of SNR wall is shown.Moreover,the effect of SNR wall of the system is analyzed under different channel distributions and different modulations.It is shown that the simulated SNR wall values are very close to that of the theoretical one.The conclusions is of great value for the design of clustering algorithm.

Key words: cognitive radio, cluster-based spectrum sensing, reporting channels, bit error probability, SNR Wall