Conditional Diffusion Model for Anti-Interference in Train Satellite Navigation Receiver Signals

In this paper, a Conditional Diffusion Model (CDM) based signal anti-jamming algorithm is proposed to solve the problem that traditional algorithms such as Discrete Wavelet Transform (DWT) and Empirical Mode Decomposition (EMD) are not good at dealing with interference signals in train satellite navigation receivers. We use one-dimensional Quadrature Phase Shift Keying (QPSK) time domain signals as input conditions, and then use CDM to predict the noise distribution from the fully noisy signal and reconstruct the QPSK signal. Experimental analysis shows that CDM has significant advantages over traditional methods in denoising QPSK signals, and shows higher accuracy under high-intensity noise interference, and the Signal-to-Noise Ratio (SNR) can be increased from -25dB to 30dB. This provides precise navigation and positioning support for train operation.