Increasing the level of semiconductor devices, improving their characteristics significantly affects the potential stability in a wide range of frequencies. Classical methods and standard measuring equipment are not designed to measure the parameters of potentially unstable transistors. Measuring systems are excited uncontrollably, which increases the measurement error. Therefore, today there is an urgent task of measuring the parameters of both transistors in particular and quadrupoles in general, in the frequency range of potential instability. The clock frequency at which modern computer equipment works is close to the microwave range (ultrahigh frequencies), which makes the problem of measuring and calculating the various functional units of the computer quite relevant. The development of new methods and means of measuring the parameters of potentially unstable quadrupoles in the microwave range is an important scientific field that can significantly improve the accuracy of their measurement on standard equipment. The conducted researches and the developed technique allow to increase accuracy of measurement of parameters of four-poles and to provide stability of experimental installation at measurement of parameters of any four-pole. The proposed methods are more efficient and solve the problem of measuring the S-parameters of potentially unstable quadrupoles in the microwave range. The parameters of the quadrupole have the dimensions of the elements of standard W-parameters, their measurements are performed in the microwave range with an error of less than 5%, and to a lesser extent depends on the coefficient of stability of the measuring path than when measuring S-parameters. When measuring the parameters of the quadrupole, there is no need to use the modes of short circuit and XX or make measurements with a coordinated measuring path, which allows to increase the measurement accuracy and ensure stability when measuring potentially unstable quadrupoles with Ks> 0.

This paper considers the principles of digital signal processing, the general provisions of digital filtering, existing methods of noise filtering in electrical signals, the median filters of one-dimensional signals are studied in detail. To solve this problem, the classification of digital signal processing tools is presented. Since the most effective for filtering noise in electrical signals are digital filters, they are given the most attention. The main purpose of signal filtering is the need to extract the information contained in them. This information, which is usually present in the amplitude of the signal (absolute or relative), in frequency or spectral composition, in phase or in the relative time dependences of several signals. The classification of existing digital filters is carried out. For further development, a median filter was selected, which belongs to the class of heuristics and is one of the most effective in filtering signals from impulse noise and white noise. Highlighting the advantages and disadvantages, a review of existing software that implements the median filter. It is established that the urgent task is to increase the processing speed and reduce resource costs in the implementation of such filters, developed an algorithm for fast median filtering, conducted an experimental test of software-implemented median filters with different apertures at different levels of fluctuation noise. This program meets all the requirements of modern norms and standards, allows its practical use to solve real problems of signal processing. In order to increase the speed of information processing, a median filtering algorithm based on difference matrices using the threshold saturation function has been developed. Developed software that implements the proposed algorithm. Schemes of the main program, reading of values of a signal from a file, filtering, sorting of data on amplitude, a choice of a window of elements, a choice of the registered values are presented. The conditions of data registration and ADC parameters to ensure efficient operation of filters are also defined