{"title":"基于层次神经网络的自动调制识别","authors":"C. Louis, P. Sehier","doi":"10.1109/MILCOM.1994.473878","DOIUrl":null,"url":null,"abstract":"Introduces a methodology for building neural networks based on a hierarchical approach, and a priori knowledge incorporation to speed up the learning phase. Superiority over a single, large, fully connected neural network classifier is demonstrated in the area of the automatic modulation recognition. This approach reduces the complexity of the system in order to improve generalization reduced sensitivity to initial conditions also allows the automation of the learning phase. Experimental results illustrate the superiority of the hierarchical approach. For 10 modulation types, the hierarchical neural network classifier is compared with the conventional backpropagation learning, the K-nearest-neighbour classifier and the well-known binary decision trees. Recognition rates are as high as 90% with a signal-to-noise ratio (SNR) ranging from 0 to 50 dB.<<ETX>>","PeriodicalId":337873,"journal":{"name":"Proceedings of MILCOM '94","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"1994-10-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"65","resultStr":"{\"title\":\"Automatic modulation recognition with a hierarchical neural network\",\"authors\":\"C. Louis, P. Sehier\",\"doi\":\"10.1109/MILCOM.1994.473878\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Introduces a methodology for building neural networks based on a hierarchical approach, and a priori knowledge incorporation to speed up the learning phase. Superiority over a single, large, fully connected neural network classifier is demonstrated in the area of the automatic modulation recognition. This approach reduces the complexity of the system in order to improve generalization reduced sensitivity to initial conditions also allows the automation of the learning phase. Experimental results illustrate the superiority of the hierarchical approach. For 10 modulation types, the hierarchical neural network classifier is compared with the conventional backpropagation learning, the K-nearest-neighbour classifier and the well-known binary decision trees. Recognition rates are as high as 90% with a signal-to-noise ratio (SNR) ranging from 0 to 50 dB.<<ETX>>\",\"PeriodicalId\":337873,\"journal\":{\"name\":\"Proceedings of MILCOM '94\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":0.0000,\"publicationDate\":\"1994-10-02\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"65\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Proceedings of MILCOM '94\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/MILCOM.1994.473878\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of MILCOM '94","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/MILCOM.1994.473878","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Automatic modulation recognition with a hierarchical neural network
Introduces a methodology for building neural networks based on a hierarchical approach, and a priori knowledge incorporation to speed up the learning phase. Superiority over a single, large, fully connected neural network classifier is demonstrated in the area of the automatic modulation recognition. This approach reduces the complexity of the system in order to improve generalization reduced sensitivity to initial conditions also allows the automation of the learning phase. Experimental results illustrate the superiority of the hierarchical approach. For 10 modulation types, the hierarchical neural network classifier is compared with the conventional backpropagation learning, the K-nearest-neighbour classifier and the well-known binary decision trees. Recognition rates are as high as 90% with a signal-to-noise ratio (SNR) ranging from 0 to 50 dB.<>