Spatio-Temporal-Frequency Graph Attention Convolutional Network for Aircraft Recognition Based on Heterogeneous Radar Network

This article proposes a knowledge- and data-driven graph neural network-based collaboration learning model for reliable aircraft recognition in heterogeneous radar network. The recognizability analysis shows that the semantic feature of an is motion patterns driven by kinetic characteristics, grammatical features contained cross-section (RCS) signals present spatial–temporal-frequency (STF) diversity decided both electromagnetic radiation shape pattern aircraft. Then, spatio-temporal-frequency attention convolutional network (STFGACN) developed to distill from RCS received Extensive experiment results verify STFGACN outperforms baseline methods terms detection accuracy, ablation experiments are carried out further show expansion information dimension can gain considerable benefits perform robustly low signal-to-noise ratio region.