Optimal Hybrid Spectrum Sensing Under Control Channel Usage Constraint

Cooperative spectrum sensing significantly improves the detection reliability of a cognitive radio network. In cooperative spectrum sensing, cognitive radio nodes send their hard decisions/detections to the fusion center via a control channel. Traditional control channels are bandwidth constrained, and a large number of hard decisions to the fusion center may saturate the control channel, thereby degrading the performance of the network. In this paper, we consider a cooperative spectrum sensing scheme, where each cognitive radio node performs a fixed sample sensing test and sends a hard decision to the fusion center via a common control channel. The fusion center collects the hard decisions from all cognitive radio nodes to form an observation vector, and performs a sequential probability ratio test to make the final decision. We term this sensing scheme as hybrid spectrum sensing. An optimization problem is formulated for the hybrid sensing strategy in order to maximize the cognitive radio network's throughput while considering interference (interference on primary user by cognitive radio network) and control channel's bit rate constraint. The decision thresholds for the cognitive radio nodes and the fusion center are considered as optimization parameters. Though the optimization problem is nonconvex, we provide an efficient algorithm for obtaining the global optimal solution. Extensive simulation results are provided to demonstrate the efficacy of our proposed approach.