MEASUREMENT OF THE INFORMATION PARAMETER OF THE SIGNAL AT A HIGH LEVEL OF ADDITIVE AND MULTIPLICATIVE NOISE
V. M. Artyushenko1, V. I. Volovach2,3
1Leonov University of Technology, Korolev, Russia
2Volga Region State University of Service, Togliatty, Russia
3Voronezh State University, Voronezh, Russia
Keywords: multiplicative (modulating) noise, additive noise, measurement accuracy, high noise level, probability of correct measurement, probability of spurious counts, parameter discreteness interval
Abstract
The influence of simultaneously acting additive and fluctuating multiplicative noise with a high level on the accuracy of measuring the information parameter of the signal is considered and analyzed. Expressions defining conditional probabilities of the correct measurement of the information parameter and a spurious count are obtained. The channel probability densities for an N-channel receiver are determined; a special case of a two-channel receiver is considered. The probability of correctly measuring the parameter at a high level of multiplicative noise is approximately equal to the probability of detecting a signal in the channel where the undistorted part of the signal is present. The value of the threshold voltage of the N-channel receiver has been determined, which allows us to obtain the specified probabilistic characteristics of the measurement. Dependences characterizing the probability of correct measurement on the level of the undistorted part of the signal are obtained for the case of a bell-shaped power distribution of fluctuations in the distorted signal through the channels of the receiving device with different spectrum widths of the noise modulation function and different numbers of channels of the receiving device. It is proved that, in the case of a large number of channels or with a power-law distribution of fluctuations of the noise component of the signal distorted by multiplicative noise symmetrical with respect to the k-th channel, the average value of measurement errors (spurious counts) is zero, i.e., there are no systematic measurement errors.
|
