基于残差网络的自动调制识别  

Automatic modulation classification based on residual network

作　　者：郭坚 漆轩[1,2] GUO Jian;QI Xuan(College of Telecommunication and Information Engineering,Chongqing University of Posts and Telecommunications,Chongqing 400065,China;Key Laboratory of Optical Communication and Networks,Chongqing University of Posts and Telecommunications,Chongqing 400065,China)

机构地区：[1]重庆邮电大学通信与信息工程学院,重庆400065 [2]重庆邮电大学光通信与网络重点实验室,重庆400065

出　　处：《计算机工程与设计》2019年第9期2406-2410,共5页Computer Engineering and Design

基　　金：重庆市教委科学技术研究基金项目(KJ1704095)

摘　　要：针对现有自动调制识别算法识别种类少、整体识别率不高和需要预处理这些缺点,提出基于深度学习模型的自动调制识别算法,设计残差网络实现11种在高斯噪声下的数字调制信号的识别。仿真结果表明,采用自适应学习率的残差网络能提供更好的识别效果,在信噪比大于0 dB时,识别率达到95%,与其它深度学习算法相比具有更低的计算复杂度,验证了该算法的有效性。Aiming at the shortages of existing automatic modulation classification algorithms such as few recognizable types,low overall recognition rate and the need of preprocessing,an automatic modulation classification algorithm based on deep learning model was proposed.A residual network was designed to realize the recognition of 11 kinds of digital modulation signals under Gaussian noise.The simulation results show that the residual network with adaptive learning rate can provide better recognition effects,and when the SNR is greater than 0 dB,the recognition rate reaches 95%,which verifies the effectiveness of the algorithm.

关 键 词：自动调制识别 深度学习 残差网络 数字调制 自适应学习率 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]