Investigation of the Dynamic Filters' Characteristics for the Analysis of Random Signals During Data Transmission

The paper analyzes methods for optimizing the parameters of the distribution function of the dynamic filters” characteristics. Such methods are used in the spectral analysis of random signals during data transmission. A feature of the work is the proposed functional, which is the starting point for solving the problem of optimizing the parameters of the distribution function of the characteristics of narrow-band dynamic filters for the spectral analysis of ergodic random signals using the correlation-filter method. Three problems of optimizing the parameters of the distribution function of the characteristics of narrow-band dynamic filters are considered and methods for their solution are proposed. It is shown that the first two optimization problems have only an asymptotic solution, which significantly limits their practical application. A simplified problem and a method for optimizing the parameters of the distribution function of the characteristics of narrow-band dynamic filters are proposed. The results of solving the optimization problem obtained by the Levenberg-Marquardt method using a software package for mathematical calculations are presented. Promising methods for solving such a problem are neural network, genetic and other intelligent algorithms for solving optimization problems. The results obtained make it possible to set optimal characteristics when designing narrow-band dynamic filters for random signal spectrum analyzers. The results of the study are proposed to be used in communication systems for filtering interference, radar to improve the quality of identifying airborne objects against the background of interference, and in measuring technology when building spectrum analyzers.