A new sensing-throughput tradeoff scheme in cooperative multiband cognitive radio network

Abstract

A conventional cognitive radio network (CRN) uses the spectrum of the licensed primary network (PN) on the premise of detecting the absence of the PN by the spectrum sensing of the sensor node (SN). In this paper, a cooperative multiband CRN is considered, wherein the SNs are allowed to use some time of the transmission slot to relay PN data by cooperative communication, while using the remaining time of the transmission slot to forward its own data, over multiple sub-bands during each frame, if the presence of PN is detected by cooperative spectrum sensing of the SNs in the sensing slot. A new sensing-throughput tradeoff scheme is formulated as a multi-variable optimization problem, which maximizes the average aggregate throughput of the CRN over all the sub-bands by jointly optimizing spectrum sensing time and sub-band transmission power, subject to the constraints on the average aggregate throughput of the PN, the maximal aggregate power of each SN, and the false alarm and detection probabilities of each sub-band. The bi-level optimization method is adopted to obtain the optimal solution by dividing the multi-variable optimization problem into two convex single-variable sub-optimization problems. The simulations show that there exists the optimal sensing time and sub-band transmission power that maximize the average aggregate throughput of the CRN and, compared with the conventional scheme, the throughput obtained by the proposed scheme is outstanding. © 2014 John Wiley & Sons, Ltd.