A New Weighting Procedure for Timescale Algorithms Based on the Sigmoid Function

SummaryThe aim of this paper is to present a new weighting procedure for timescale calculation. We utilize the sigmoid function to transform statistical quantities of clocks, such as Allan variance and Time deviation, into a fixed range of (0, 1). This transformation smooths the fluctuations in weights as the statistical quantities vary. Results from a small ensemble of clocks demonstrate the effectiveness of the sigmoid function in improving the frequency stability of the timescale.