Composite Radar Modulation Identification by Transfer Learning
Fang Li, Zixian Yang, Bo Huang, Yang Chen · 2021
Radar Signal Modulation Identification plays an important role in electronic countermeasures (ECM) in order to involve radar parameter estimation and radar state prediction. Most of the traditional radar signal modulation sensing techniques are used for known signal and cannot distinguish the unknown composite modulation signal. In this chapter, a new radar signal modulation sensing algorithm using Transfer Learning with the framework of a deep convolutional neural network (CNN) is presented. Using this method, it is possible to distinguish the unknown composite modulation signal with limited training data. Furthermore, a good classification accuracy on testing dataset base on Transfer Learning CNN model is achieved with this method. Several similarity cases are also studied to verify the adaptability of the proposed method.