A Distributed Radar Multi-Jamming Recognition Method Based on Transformer Network

With the advancement of deception jamming and multi-radar cooperative anti-jamming technologies, multi-radar systems designed to counter multiple jammers have become a research hotspot, with jamming recognition being a critical step. In recent years, deep learning has been successfully applied to radar jamming recognition, where the jamming signal is transformed into a time-frequency map and subsequently treated as an image recognition problem. However, the performance of jamming recognition algorithms deteriorates significantly when multiple jamming signals overlap in the time-frequency domain. To address this issue, this paper proposes a distributed radar multi-jamming recognition method based on a Transformer network. This method employs adaptive beamforming to separate jamming echoes in the spatial domain, and then converts the jamming echoes from each radar node into time-frequency spectrograms using the short-time Fourier transform. These spectrograms are then input into the Transformer network to detect the type of jamming and estimate jamming parameters based on their relative positions in the time-frequency spectrograms. Finally, the multi-node processing results are fused to output the jamming type and the corresponding fused jamming parameters. Simulation experiments verify the effectiveness of the proposed algorithm.