摘要
针对于认知无线电中传统感知算法受信噪比(SNR)影响、过度依赖主用户先验知识和感知时间长等问题,提出基于支持向量机(SVM)的频谱感知算法,通过将信号能量值、SNR值与类别标签作为训练数据,对其进行SVM学习后,得出此CR环境下的分类模型。仿真结果显示在低SNR环境下,机器学习频谱感知算法检测概率比能量检测的提高了近40%,分类错误率仅为1.4%,因此具有更优良的感知性能。
the spectrum sensing algorithm based on support vector machine(SVM) is proposed to solve the problem that the traditional sensing algorithm in cognitive radio is affected by the signal-to-noise ratio(SNR), over-dependence on the prior knowledge of the main user and the long sensing time. the classification model in this cr environment is obtained by using the signal energy value, snr value and category label as training data. simulation results show that in low snr environment, the detection probability of machine learning spectrum sensing algorithm is nearly 40% higher than that of energy detection, and the classification error rate is only 1.4%, so it has better sensing performance.
作者
胡浩
屈少晶
Hu Hao;Qu Shaojing(China Mobile Group Yunnan Company Limited,Ministry of planning and construction,Kunming 650000,China;Kunming University,Economy and Management School,Kunming 650000,China)
出处
《长江信息通信》
2021年第4期59-62,共4页
Changjiang Information & Communications
作者简介
胡浩(1979-),男,云南曲靖人,昆明理工大学硕士,高级工程师,主要从事通信工程建设管理、供应链管理、网络运营管理等方面研究;屈少晶(1981-),女,湖北宜昌人,云南大学硕士,讲师,主要从事企业管理、财务管理等方面研究。
