인지 무선 통신에서 전이 학습을 이용한 딥러닝 기반 변조 신호 센싱 기법

초록

A significant amount of labeled training data is required to implement a superior deep learning-based signal detector for cognitive radio network. In general, however, training data is often not sufficiently guaranteed depending on modulation. In this paper, we propose the deep learning-based modulation signal detection scheme via transfer learning to effectively detect a signal of the primary user with less training data in a cognitive radio network. The proposed scheme also takes into account a convolutional neural network model using the similarity between modulation signals through dynamic time warping.

키워드