Programmable Radiometer-noise Model with Random Drift Contribution

V. N. Chernenkov
DOI: https://doi.org/10.1134/s199034132360045x
2024-05-10
Astrophysical Bulletin
Abstract:A method for radiometer-noise generation and its software implementation in the form of a Python library module is proposed. The method can be used to simulate radiometer noise with variable component properties: Gaussian white noise and flicker noise (random drift) with a power spectrum linearly increasing with decreasing frequency. The method is based on the transformation of spectral properties of Gaussian noise via sequential filtering by a differentiator, nonlinear threshold switch, and subsequent integration. By changing the integrator accumulation time constant and the threshold level relative to noise intensity one can change the resulting form of the spectrum and the corresponding statistical parameters of the output noise over a wide range: from stationary white noise to noise dominated by a -type spectrum component. Empirical polynomial approximation formulas for the dependence of the spectral index of the resulting noise records on the nonlinear threshold switch level in the 0.0– interval are derived based on numerical simulations using a software implementation of the model. Examples of noise-record implementations, their spectra, and autocorrelation functions are demonstrated. The generated noise is shown to be similar to the noise observed on radio telescopes of the Institute of Applied Astronomy of the Russian Academy of Sciences.
astronomy & astrophysics
What problem does this paper attempt to address?