基于特征根比分布的协作频谱感知算法研究

Cooperative spectrum sensing algorithms based on eigenvalues ratio distribution

介绍两种基于大维随机矩阵理论(Random Matrix Theory,RMT)的频谱感知方法,综合对比讨论并验证其感知性能。这些方法以采样协方差矩阵的最大最小特征根之比作为判决统计量与相应的判决门限相比较来确定频谱是否空闲。通过与能量检测(Energy Detection,ED)算法仿真对比,它们克服了能量检测中需要预知噪声信息的缺点,并具有不受噪声不确定度影响的优点,有较高检测性能和优越性。

Two spectrum detection methods based on the large Random Matrix Theory（RMT） are introduced and comprehensive discussion and tests of their detection performance are presented.These algorithms use the ratio of the minimum and maximum eigenvalues of sample covariance matrix as test statistic,which is compared with the corresponding decision threshold to infer the presence or absence of primary signal.Simulations based on QPSK signals derived from practical scene are presented.These algorithms can overcome the noise uncertainty difficulty and don＇t require the knowledge of noise.Results verify that these algorithms outperform Energy Detection（ED） algorithm.