首页 | 本学科首页   官方微博 | 高级检索  
     

基于卷积神经网络的无线电信号搜索
引用本文:何元利,龚晓峰.基于卷积神经网络的无线电信号搜索[J].兵工自动化,2017,36(10).
作者姓名:何元利  龚晓峰
作者单位:四川大学电气信息学院,成都,610065;四川大学电气信息学院,成都,610065
摘    要:针对军事侦察或干扰查找环节解决方案较少、已有方案算法速度慢、准确度低的问题,提出一种采用卷积神经网络的无线电信号快速查找方法.该方法将卷积神经网络应用于无线电监测领域,通过数据预处理、参数训练、信号匹配3个步骤实现信号的快速搜索.实验结果表明:该方法查找速度快,精度达到98%,并且有效减少对信号复杂参数的依赖,在无线电信号快速搜索方面的良好性能.

关 键 词:卷积神经网络  深度学习  无线电  频谱  军事侦察  干扰查找
收稿时间:2017/6/19 0:00:00
修稿时间:2017/7/22 0:00:00

Radio Signal Searching Based on Convolution Neural Network
He Yuanli,Gong Xiaofeng.Radio Signal Searching Based on Convolution Neural Network[J].Ordnance Industry Automation,2017,36(10).
Authors:He Yuanli  Gong Xiaofeng
Abstract:Aiming at the problem that a few of solutions about military reconnaissance or interference detect, low calculation speed of current algorithm and low correctness, put forward a radio signal quick find method by using convolution neural network. The method uses convolution neural network in radio detecting field, and realizes signal quick searching by data pretreatment, parameter training, and signal matching. The test results show that the method is fast, at the same time the search accuracy is about 98%, and effectively reduce the dependence on the complex parameters of the signal. Test results also indicate the great performance in terms of quick search of radio signals.
Keywords:convolution neural network  deep learning  radio  frequency spectrum  military reconnaissance  interference detect
本文献已被 万方数据 等数据库收录!
点击此处可从《兵工自动化》浏览原始摘要信息
点击此处可从《兵工自动化》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号