S. Yu. Medvedev

Effect of the non-gaussianity on the measurement error for the filtered 1/F noise intensity

To study the nature of the 1/f noise phenomenon in conductors, we seek a tool for testing different hypotheses of 1/f noise origin. The method analyzing the noise intensity at the output of a bandpass filter is discussed for the case of non-Gaussian processes. Data on measurement error are presented for the 1/f noise intensity in GaAs films and the Gaussian white noise emulated by a computer. A numerical model of 1/f noise as the superposition of telegraph random processes has been created. This method requires further improvement to check the noise for stationarity. Some ideas of how to do that are proposed.

Application of a modified gradient least squares algorithm to an adaptive, actively quenched, sound field system

A modified least squares algorithm, preventing the overflow of the discharge grid of weight coefficients of an adaptive transverse filter and guaranteeing stable system operation, is suggested for the tuning of an adaptive system of an actively quenched sound field. Experimental results are provided for an adaptive filter with a modified algorithm in a system of several harmonic components of an actively quenched sound field.

Correlation between the Intensities of Spectral Components of 1/f Noise

We analyze the estimate of the correlation factor between noise intensities at the outputs of two bandpass filters as a parameter that allows us to obtain additional information on the nature of the 1/f noise. The confidence interval is calculated for the experimental evaluation of the correlation factor between readings of noise intensities. Data on noise in GaAs films and also for the numerical model of the 1/f noise as a superposition of telegraphic random processes are presented and discussed.