摘要
针对低信噪比下无线认知网络的主用户频谱检测问题,提出了一种基于循环谱和随机森林的检测方法。该方法首先将信号变换到循环谱域,计算信号非零循环频率最大时的信号循环谱,并以循环谱的均值和方差构成特征向量。然后,基于随机森林构建无线认知网络主用户信号频谱检测模型,并分别构建正负样本信号进行模型训练。最后利用训练完成的模型检测无线认知网络主用户信号。仿真实验结果表明,该方法能够有效实现主用户信号频谱检测,且检测性能优于目前常用算法。
Aiming at the problem of spectrum detection of primary users in wireless cognitive networks with low SNR,a detection method based on cyclic spectrum and random forest is proposed.The method first transforms the signal into the cyclic spectrum domain,calculates the signal cycle spectrum when the non-zero cycle frequency of the signal is maximum,and forms the feature vector by the mean and variance of the cyclic spectrum.Then,based on the random forest,the main user signal spectrum detection model of wireless cognitive network is constructed,and the positive and negative sample signals are constructed separately for model training.Finally,the training completed model is used to detect the main user signal of the wireless cognitive network.The simulation results show that the proposed method can effectively realize the spectrum detection of the main user signal,and the detection performance is better than the commonly used algorithms.
作者
杨波
YANG Bo(Sichuan University of Arts and Science,Sichuan Dazhou 635000,China)
出处
《机械设计与制造》
北大核心
2021年第2期282-285,共4页
Machinery Design & Manufacture
基金
2018年国家级实验教学示范中心课题(SFZX2018B15)。
关键词
无线认知网络
频谱检测
循环谱
随机森林
Wireless Cognitive Network
Spectrum Detection
Cyclic Spectrum
Random Forest
