Font Size: a A A

Research On Modulation Recognition Of Communication Signals Based On Deep Learning

Posted on:2022-05-26Degree:MasterType:Thesis
Country:ChinaCandidate:Z W ZhangFull Text:PDF
GTID:2518306485950759Subject:Electronics and Communications Engineering
Abstract/Summary:PDF Full Text Request
Communication signal modulation recognition refers to the technology of researching and analyzing the signal under the condition that many relevant parameters are unknown,so as to determine the modulation mode of the communication signal.With the development of radio communication technology,modulation recognition is widely used in military and civilian fields such as electronic countermeasures and spectrum detection.In the increasingly complex electromagnetic environment,improving modulation recognition performance at low SNR is a research hotspot in this field.At present,deep learning algorithms have made many breakthroughs in signal processing technology by virtue of their excellent characterization and classification capabilities.Based on the deep learning algorithm to extract and process signal features,this paper constructs a CG(CNN-GRU)joint network modulation recognition algorithm.The main research contents are as follows:Aiming at the problems of high real-time requirements and small batches of data sets in actual application of modulation recognition,this paper adopts the GRU network that processes small-scale data and has a faster speed as the basic model,optimizes the design of the GRU,and simplifies the data set.And enhanced processing further improves the efficiency of modulation signal recognition.Aiming at the problem that the extracted features of the signal will be severely distorted with the decrease of the SNR,which leads to the reduction of the recognition effect.From the perspective of optimizing the data set,this paper incorporates the estimated characteristics of the SNR into the data set to complete the recognition of the modulation signal.The results show that incorporating the SNR estimation into the feature improves the recognition performance of the modulated signal.Aiming at the problem of low recognition rate of traditional modulation recognition algorithm at low SNR,this paper starts from the perspective of optimizing the classifier,and according to the recognition characteristics of CNN and GRU network,proposes a modulation recognition algorithm of CG joint network,and combines it with CNN The algorithm,GRU algorithm,decision tree algorithm,SVM algorithm and BP network algorithm are compared.The results show that the algorithm proposed in this paper effectively improves the recognition performance when the SNR is low.
Keywords/Search Tags:modulation recognition, deep learning, SNR estimation, feature extraction
PDF Full Text Request
Related items