Sky–Ground Wave Time Delay Estimation in eLoran Based on Two‐Stage MS‐CNN

ABSTRACT

The working accuracy of the eLoran system is susceptible to the interference sky–ground wave multipath aliasing. Therefore, the accurate estimation of the sky–ground wave relative delay plays an important role in suppressing skywave interference. To address the performance degradation of existing methods under low‐SNR or small‐delay‐interval conditions, this paper constructs a delay estimation method that combines IFFT preprocessing with two‐stage MS‐CNN. The IFFT preprocessing is used to enhance the time‐domain features related to delay. The two‐stage MS‐CNN achieves accurate estimation of the groundwave, skywave and their relative delay through multi‐scale convolution, channel attention and local residual correction. Simulation results show that the proposed algorithm can accurately estimate the delay over an SNR range of −5 to 20 dB, and the estimation error can be reduced to as low as 0.021 under high‐SNR conditions. The results of multiple comparative experiments show that the proposed method maintains high delay estimation accuracy under different SNR conditions, and exhibits better stability especially in low‐SNR and small‐delay‐interval scenarios, further verifying its applicability and robustness under complex conditions.