Automatic modulation classification using different neural network and PCA combinations

作者：

Highlights：

• A complete intelligent digital modulation recognition system is proposed.

• We study the influence of compressing data on the performance of the classifier.

• The training data sets can be reduced strikingly via signal parameter optimizer.

• The method was shown to be highly performance and scalability for software radios.

摘要

•A complete intelligent digital modulation recognition system is proposed.•We study the influence of compressing data on the performance of the classifier.•The training data sets can be reduced strikingly via signal parameter optimizer.•The method was shown to be highly performance and scalability for software radios.

论文关键词：Modulation classification,Neural networks,Statistical features,RBF-PCA combination,DVB-S2 APSK

论文评审过程：Received 22 May 2019, Revised 27 February 2021, Accepted 21 March 2021, Available online 30 March 2021, Version of Record 21 April 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.114931