Contrastive representation learning for Radar Signals Intra-Pulse Modulation Classification Using data augmentation

Automatic classification of radar waveforms is a fundamental technique in electronic countermeasures (ECM). Recent supervised deep learning-based methods have achieved great success in such classification tasks. However, for the non-cooperative systems, it is difficult to obtain the labels. In this paper, we propose a representation learning technique to automatically extract useful features from the raw IQ (Inphase / Quadrature) inputs. Firstly, a pretext model is trained in a self-supervised way with the help of several data augmentation techniques to extract the class-dependent features. After that, the pseudo-supervised contrastive training is involved to further promote the separation between the extracted class-dependent features. Simulation results show that the model trained by the proposed method can be useful for various downstream classification tasks including unsupervised clustering, semi-supervised learning, zero-shot and few-shots transfer classification.