根据高阶循环统计量(HOCS)具有的严格平稳特性,提出了一种基于HOCS的频谱感知方法.该方法利用渐近最优估计理论,构造了进行频谱感知的二元假设检验模型,推导了相应的检验统计量,得出了检测概率的闭式表达式.通过对感知算法所需实数乘法运算次数的比较,说明了该算法具有与能量检测(ED)几乎相同数量级的计算复杂度.仿真结果表明,相比传统的ED法和2阶循环统计量检测法(SCFD),该方法具有良好的检测性能和抗同频干扰性能.
A spectrum sensing method based on the high order cyclic statistic(HOCS) in cognitive radio was presented according to the strict cyclostationarity of the HOCS. The binary hypothesis test model and the test statistic are constructed and the explicit formula of the detection probability is derived by exploiting the asymptotic optimal estimation theory. The computational complexity of the proposed method has nearly the same order as that of energy detection(ED) by comparing their real multiplication operation times. The simulation results show that the new scheme has superior performances of detection and resistance to same frequency interference compared with the conventional spectrum sensing methods.