CALCULATION AND SIMULATION OF DIGITAL FILTERS

Abstract

         The paper proposes models of digital low-pass filters, the coefficients of which are selected automatically when changing both the sampling frequency of the filter and its cutoff frequency.

    A digital filter, like an analog filter, is designed to select the required frequency or frequency band from the input signal. However, with digital filtering, one has to operate with discrete sequences, for the formation of which the input analog signals are first converted into the digital form. In addition, to apply the filter in analog systems, it is necessary to turn on a smoothing filter (SF) at the output. Thus, in the general case, the digital filter at the input contains an analog-to-digital converter for digitizing the input analog signal. Further, in the digital processor, the already discrete (digital) signal is processed according to algorithms corresponding to a given filter, i.e. at its output, a signal is formed, which, after a digital-to-analog converter, takes the form of a stepwise continuous function. Finally, an anti-aliasing filter is used to obtain a continuous output signal, limiting the frequency spectrum by suppressing high-frequency components.    At computer modeling of various digital radio engineering systems, it is often necessary to investigate these systems at different sampling rates (clock frequencies). If there are digital filters in the system, you also have to vary their clock frequency. However, each time, you should recalculate the filter coefficients for a new sampling rate and continue modeling, which naturally leads to certain inconveniences.    The work investigates the models of digital low-pass filters (LPF), the coefficients of which are calculated automatically using the obtained relations when changing the cutoff frequencies and sampling of the filter.