Research Article
Open-Set Recognition Algorithm of Signal Modulation Based on Siamese Neural Network
@INPROCEEDINGS{10.1007/978-3-030-99200-2_3, author={Pengcheng Liu and Yan Zhang and Mingjun Ma and Zunwen He and Wancheng Zhang}, title={Open-Set Recognition Algorithm of Signal Modulation Based on Siamese Neural Network}, proceedings={Communications and Networking. 16th EAI International Conference, ChinaCom 2021, Virtual Event, November 21-22, 2021, Proceedings}, proceedings_a={CHINACOM}, year={2022}, month={4}, keywords={CNN Modulation recognition Open-set recognition Siamese network}, doi={10.1007/978-3-030-99200-2_3} }- Pengcheng Liu
Yan Zhang
Mingjun Ma
Zunwen He
Wancheng Zhang
Year: 2022
Open-Set Recognition Algorithm of Signal Modulation Based on Siamese Neural Network
CHINACOM
Springer
DOI: 10.1007/978-3-030-99200-2_3
Abstract
To realize the open-set recognition of various signal modulations, we propose a three-stage recognition algorithm based on the siamese neural network. Firstly, the convolutional neural network (CNN) is employed to extract the features of the original signal to obtain the multi-dimensional feature vector, thereby constructing the feature data set. Then, a double-threshold method basedon average valuesis designed to distinguish signals with known modulation types and unknown ones. Finally, unknown modulation types are recognized by the siamese neural network. We proposed an iterative algorithm based on the siamese neural network to obtain modulation types of the unknown signals. Experimental results show that the proposed algorithm model can be used for open-set recognition and is suitable for applications in complex wireless communication environments.
