中文摘要：

针对噪声不确定度严重影响能量检测算法频谱感知性能的问题，提出了一种基于功率谱分段对消的频谱感知算法．该算法利用被监测信号周期图估计功率谱谱线互不相关的性质，以频带内一些谱线强度和与剩余谱线强度和的比值作为检验统计量，可实时鲁棒地感知监测频段的占用情况．理论分析和仿真结果表明，该算法可以在较宽的信噪比范围内获得较低的虚警概率和较高的检测概率，能有效克服噪声不确定度对检测性能的影响，判决门限不随次级用户周围环境噪声电平的变化而改变，适用于复杂电磁环境下的频谱监测和认知无线电系统．

英文摘要：

Since noise uncertainty seriously degrades the spectrum sensing performance of energy detection algorithms, a novel spectrum sensing algorithm based on the power spectral density segment cancellation （PSDSC） is proposed. This spectrum sensing algorithm, which can yield real-time and robust performance, makes use of the un-correlation of different power spectrum lines and takes the ratio of some PSD （power spectral density） lines to the residual PSD lines as the detection statistics. Theoretical analysis and simulation results show that the PSDSC algorithm can overcome the noise uncertainty problem effectively and that the decision threshold does not vary with the ambient noise level of secondary users. The PSDSC algorithm could offer a high probability of detection （Pa） at a low probability of the false alarm （P 1a） for a wide range of signal to noise ratios （SNR） and could be applied to spectrum monitoring and cognitive radio systems in the complex electromagnetic environment.