Diffusion SigFormer for Interference Time-Series Signal Recognition

The various interferences in the actual environment make electromagnetic signal recognition challenging, and this topic has an extremely important application value. In this paper, a novel interference signal recognition transformer was proposed, named Diffusion SigFormer, to explore the mechanism of signal interference and improve the ability to identify interference signals. Firstly, we explored the interference law of electromagnetic signals and designed a signal interference mechanism for simulating real scenes. Secondly, the Diffusion Signal Denoising Module was proposed to denoise the input interference signal. We used different types of noise and various interference rates to simulate noise in real electromagnetic environments. Thirdly, SigFormer was designed to extract and classify the denoised signal. For the characteristics of electromagnetic signals, SigFormer leveraged 1-D Patch Embedding and combined Transformer with convolution. Finally, we conducted experimental verification on RML2016.10a, RML2016.10b and BT datasets. The experimental results have shown that the proposed method has excellent anti-interference ability.