Robust Spectrum Sensing and User Identification for PCP-OFDM Signal Using Noise Insensitive Threshold

To ensure proper operations of primary users and a fair spectrum sharing among the secondary users, spectrum sensing and user identification at cognitive radio (CR) have to be achieved with high accuracy under low signal- to-noise ratio (SNR). Motivated by the unavailability of perfect synchronization and prior noise knowledge at the CR sensing device, a robust sensing and user identification technique using noise insensitive threshold under low SNR is proposed in this paper for signals from precoded cyclic prefix Orthogonal Frequency Division Multiplexing (PCP-OFDM) system, which is proposed recently to provide a fast-adapting CR communication platform using a PCP enabled signaling link for dynamic system interaction. The proposed spectrum sensing and user identification are achieved with time domain cyclic correlations based on the inherent PCP feature in the received signals. Different users are distinguished in the sensing process with their unique PCPs. The impact of unknown noise statistics on the sensing threshold determination is removed with the proposed correlation ratio metric. The effect of unknown timing offset is also mitigated with the cyclic correlations. Numerical results are provided to verify the analysis and evaluate the performance of the proposed algorithm.