Performance analysis of low average reporting bits cognitive radio schemes in bandwidth constraint control channels

Collaborative spectrum sensing is regarded as a key technology for tackling the challenges of the practical implementation of cognitive radio (CR). However, most of the proposed solutions require infinite bits or several bits exchanging per each CR to make the final decision to indicate the absence or presence of a primary user. It is well known that if N CRs employ energy detector with the same threshold and then send individual decision bits to a fusion centre, the OR-rule (1-out-of-N rule) will be optimal under Neyman-Pearson criterion. To decrease the average number of exchanging bits, we propose novel schemes while they perform as well as the OR-rule, our analytical and simulation results show that the proposed schemes has two important advantages over previously proposed schemes: (1) the average number of exchanging bits for N CRs always is 1/N bits per CR in perfect control channels; however, it is much higher than 1/N for the other proposed methods, e.g. it is 1 bit for the OR-rule and (2) the proposed method is very robust against the errors of imperfect control channels, specially, in low signal-to-noise ratio (SNR) regimes. Furthermore, the average number of reporting bits do not change considerably under the imperfect control channels.