基于采样协方差矩阵的频谱感知算法仿真分析  

Simulation and analysis of spectrum sensing algorithms based on sample covariance matrix

作　　者：宋云飞[1] 卢光跃[1] 

机构地区：[1]西安邮电学院通信与信息工程学院,陕西西安710121

出　　处：《西安邮电学院学报》2011年第5期12-16,共5页Journal of Xi'an Institute of Posts and Telecommunications

基　　金：陕西省自然科学基金资助项目(2010JQ80241);陕西省教育厅基金资助项目(2010JK836;11JK0925)

摘　　要：为了克服噪声不确定度的影响,实现真正意义上的盲检测,该文介绍了3种基于协方差矩阵的感知算法,并通过基于MATLAB的仿真实验,综合对比并验证3种算法的感知性能。仿真结果表明,与基于最大最小特征值之比算法(Maximum-Minimum Eigenvalue,MME)和基于协方差矩阵绝对值算法(Covariance Abosute Value,CAV)相比,基于协方差矩阵Cholesky分解的感知算法(Covariance Cholesky Factorization,CCF)避免了对采样大小和采样维数的渐进假设,从而可以精确地设置判决门限,具有更高的检测性能。To overcome the obstacles of noise uncertainty and realize blind detection in the true sense, three algorithms based on sample covariance matrix are studied. By simulation experiments with MATLAB, the performance of these algorithms is verified and contrasted. The simulations results show that, the algorithm based on Cholesky factorization can work well without any asymptotic assumptions on the sampling size and sampling dimension, it can make the decision threshold settled accurately and its detection performance is better than the others.

关 键 词：频谱感知 盲检测 采样协方差 CHOLESKY分解 

分 类 号：TN911.23[电子电信—通信与信息系统]