中文摘要：

提出了基于最优用户选择的协作频谱感知方案,通过引入Gerschgorin圆盘理论建立用户选择机制,筛选出信道条件最优的若干个认知用户,由其对应的感知数据空间生成全局判决统计量以实现最优协作感知,并在此基础上提出迭代门限算法,进一步优化感知性能.理论分析和仿真结果表明,该方案无需知晓授权用户信号、信道、噪声功率等先验信息,对噪声功率不确定性具有较强的鲁棒性,且当接收机采样次数和参与协作的用户数都受限时,感知性能仍然很稳定,可作为协作频谱感知的现实可实现方案.

英文摘要：

Gerschgorin disk theorem based optimal user selection algorithm is proposed for cooperative spectrum sensing. The fusion center receives the sensing data from the secondary users and establishes the optimal user selection mechanism by employing Gerschgorin disk theorem. The sensing data of the se- lected secondary users will subsequently be fused as the global test statistic. Aided by the optimal user selection strategy, the cooperative sensing performance can be ameliorated. Furthermore, the detection performance can also be improved through the proposed iterative thresholding algorithm. Theoretical analysis and simulation results show that the algorithm depends on no a priori knowledge of the primary user signal and the noise power. The proposed method is hence robust against the noise power uncertainty, and can maintain consistent performance when slightly fewer sampling data and users are involved in cooperative sensing.