摘要
深度学习是当前人工神经网络领域的研究热点,广泛应用于字符识别、图像识别和语音识别等应用中。雷达通信目标识别是通信对抗的前提和关键。文中分析了模板匹配法、DS证据理论等传统通信目标识别方法的在特征提取、模型表达方面的不足,对深度学习神经网络在通信目标识别中的应用进行了初步探讨,并提出了一种基于深度学习的通信目标识别框架。该框架和思路同样适用于雷达对抗目标识别等问题,可为深度学习在雷达目标识别领域的应用提供支撑。
Deep learning is a hot research area in the field of artificial neural network,and has been widely utilized in applications such as optic character recognition,graphic identification and speech recognition.Radar communication object recognition is the premise and crucial part of communication countermeasures.This paper analyses the drawbacks of traditional communication object recognition methods such as template-matching DS evidence theories and explores the potential applications of deep-learning neural network in this regard.A framework of deep learning based communication object recognition is proposed in this paper which is also suitable for that in the field of radar countermeasures.
作者
程嘉远
CHENG Jiayuan(Electronic Information Engineering Institute of Changchun University of Scientific ＆ Technology,Changchun 130022,China)
出处
《现代雷达》
CSCD
北大核心
2018年第8期55-59,共5页
Modern Radar
关键词
雷达
通信
目标识别
BP神经网
RBF神经网络
深度置信网络
卷积神经网络
radar
communication
object recognition
BP neural network
RBF neural network
deep belief network （DBN）
convolutional neural network （CNNI
作者简介
通信作者：程嘉远,Email：13305133716@189．cn,男，1997年生，本科生。研究方向为通信和信息系统。
