一种低运算复杂度的优化DAG频谱感知算法  

An Optimized DAG Spectrum Sensing Algorithm with Low Computational Complexity

作　　者：高鹏[1] 刘芸江[1] 高维廷[1] 李曼[2] 陈娟 GAO Peng;LIU Yunjiang;GAO Weiting;LI Man;CHEN Juan(Institute of Information and Navigation,Air Force Engineering University,Xi'an 710077,China;Xi'an Aeronautical University,Xi'an 710077,China;No.91917 Unit,Beijing 102400,China)

机构地区：[1]空军工程大学信息与导航学院,西安710077 [2]西安航空学院,西安710077 [3]91917部队,北京102400

出　　处：《弹箭与制导学报》2018年第1期109-113,共5页Journal of Projectiles,Rockets,Missiles and Guidance

基　　金：国家自然科学基金(61571364);第60批中国博士后科学基金(2016M603044)资助

摘　　要：针对经典频谱感知算法在低信噪比、低采样下检测效果不佳,特征值分解运算复杂度高的问题,文中基于最大最小特征值之差算法(DMM)提出了一种低运算复杂度的优化算法(DAG)。算法利用特征值算数均值与几何均值之差构建检测统计量,以"累积法"迭代计算最大特征值获取动态检测门限。仿真结果表明,该算法在低信噪比、相对低采样以及多协作认知用户数下,较经典特征值类算法降低了运算复杂度、提升了检测概率。In order to solve the problems that for the classical spectrum sensing algorithm,the detection effect was poor and the complexity of eigenvalue decomposition operation was high at low signal to noise ratio and low sampling,based on the difference between the maximum and minimum eigenvalue algorithm (DMM),an optimization algorithm with low computational complexity was proposed in this paper.The algorithm uses the difference between the mean value and the geometric mean value of the eigenvalue to construct the detection statistics,and applies cumulative method to iteratively calculate the maximum eigenvalue to obtain dynamic detection threshold.The simulation result shows that the algorithm reduces the computational complexity and improves the detection probability compared with the classical eigenvalue algorithm in the situation of low SNR,relatively low sampling and multi cooperative cognitive users.

关 键 词：特征值 频谱感知 算数均值 几何均值 运算复杂度 最大最小特征值差 

分 类 号：TN925[电子电信—通信与信息系统]