Conventional cooperative spectrum sensing (CSS) schemes in cognitive radio networks (CRNs) require that the secondary users (SUs) report their sensing data separately in the time domain to the fusion center, which yields long reporting delay especially in the case of large number of cooperative SUs. By exploiting the computation over multiple-access channel (CoMAC) method, this paper proposes a novel CoMAC-based CSS scheme that allows the cooperative SUs to encode their local statistics in transmit power and to transmit simultaneously the modulated symbols sequence carrying transmit power information to the fusion center. The fusion center then makes the final decision on the presence of primary users by recovering the overall test statistic of energy detection from the energy of the received signal. Performance metrics of the CoMAC-based CSS scheme, i.e., detection probability and false alarm probability, are further derived based on the central limit theorem. Finally, based on the derived detection and false alarm probabilities, both energy detection threshold and spectrum sensing time are optimized to improve the average throughput of the CRN. Simulations demonstrate the efficiency of this work.

This work was published on Computer Networks,2017,112:84-94. titled  Time-efficient cooperative spectrum sensing via analog computation over multiple-access channel.