Abstract: The enhanced Loran （ELoran） system， an advancement of the Loran-C navigation system， plays a critical role in high-precision positioning and timing through accurate signal period recognition. However， in practical applications， conventional signal period recognition methods are susceptible to errors caused by skywave interference and noise， which degrade positioning accuracy. To address this limitation， this study first employed a spectrum division approach with an adaptive window width to extract delay and amplitude information for both skywave and groundwave components. By adjusting the window width， the proposed method can obtain relatively accurate delay and amplitude information under varying signal-to-noise ratio （SNR） and skywave-groundwave strength conditions. The particle swarm optimization （PSO） algorithm is employed to refine the IFFT time-delay estimation results. By simulating the dynamic behavior of a particle swarm， the PSO algorithm effectively searches for the optimal delay estimation， thereby significantly reducing estimation errors. This approach addresses the issue of significant delay estimation errors in traditional spectrum division methods under low SNR conditions， where susceptibility to noise is pronounced. Simulation results demonstrate that the proposed algorithm accurately estimates groundwave delay across a range of SNR levels， delay differences， and amplitude ratios， with an error less than 0.5 μs， significantly outperforming traditional IFFT and MUSIC algorithms. Finally， a skywave suppression algorithm was implemented to attenuate the amplitude of the skywave， thereby minimizing its interference with the groundwave signal and enhancing the overall performance of the IFFT. Simulation results indicate that the algorithm achieves over 90% accuracy in groundwave time-delay estimation for SNRs above 0 dB. Further analysis reveals that the algorithm not only enables the separation of sky wave and ground wave under the conditions of strong sky wave and low signal-to-noise ratio but also addresses the error limitations inherent in traditional methods. This advancement offers a novel approach for high-precision positioning and decoding of ELoran signals.