Application of Fourier transform in processing inter-satellite measurements

The article presents a method for noise filtering in inter-satellite measurements based on Fourier transform and adaptive colored noise processing. The relevance of the study stems from the need to enhance the accuracy of satellite navigation systems amid increasing interference and operational complexity. The primary goal is to reduce random and systematic errors in measurement residuals caused by discrepancies between calculated and actual signal delays between satellites. The proposed method combines spectral analysis using Fourier transform to decompose signals into frequency components with adaptive algorithms that account for the spectral density characteristics of colored noise. This approach effectively isolates low-frequency useful signals, suppresses high-frequency noise, and minimizes signal distortion. The results demonstrate a noise level reduction of 0.5 m while preserving data structural integrity, confirmed by amplitude and phase spectrum analysis, as well as correlograms. The method proved effective in environments dominated by pink noise associated with long-term trends and hardware delays. The developed technique improves delay estimation accuracy in geodetic tasks, aviation navigation, and monitoring applications. Future research directions include integrating machine learning algorithms for real-time dynamic adjustment of filtering parameters and optimizing computational resources.