1/f noise modeling

Author: Giorgi Gogibedashvili
Co-authors: Giorgi Gogibedashvili
Keywords: Flicker noise, modeling
Annotation:

Annotation This work shows classification methods for signals gotten from physical experiments. These methods are dependent on signal spectral analysis. Specifically: for modeled signals power spectral density is calculated. If spectrums are characterized by frequency to the power of -1 then this noise is called flicker noise. In random processes theory signal is classified by ‘’colors’’. This work discusses both ‘’white’’ and other ‘’colored’’ noise modeling. Examples of flicker noise in real-time physical processes is also discussed here. For describing signal’s chaotic nature Recurrence plot method is used. Recurrence plots are shown for ‘’colored’’ noise. The causing reason for flicker noises to occur in different physical problems varies, but they are all characterized by the same non-linear dynamical properties. For better understanding the properties of given signals modern non-linear dynamics statistical processing, power spectral density and recurrent plot methods are used in this work. These methods for signal processing in dynamical systems is used for flicker noise modeling. Lorenz dynamical system is discussed which describes convectional currents’ chaotic states. Chaotic signal’s characteristics derived from this system is part of ‘’colorful’’ noise. Rikitake and Dufing models’ signals are also shown. This work also examined particle dynamics for co-ordinate and velocity signal characteristics in periodic potential well, and it’s shown that the signal is a flicker noise.