V. U. Ihnatkin


An algorithm which allows to estimate the capacity spectrum for a discrete sample of values of N for a time interval (- T / 2, + T / 2) using a filter with the same narrow spectral band as in a rectangular time window, but with the level of lateral petals, smaller by 4,3 db has been considered. Examples of several «energy» filters have been given. This shows the response of the «energy» filter in comparison with the filter based on traditional time windows. There are possibilities to control both the width of the filter band and the shape of its peak. Also, the dimension of the variable variation increases considerably.

In the case of spectral analysis, the Fourier transform is said to be performed with two different time windows in time. The result is based on the factors of the real and unreal parts of the first and second transformation. It does not require any optimal qualities from each of the time windows separately, only the final result is optimized. This is effective if one of the time windows resembles the Kaiser-Bessel window. In this case, the convolution after the Fourier transform becomes labor-intensive one, requires a lot of computational operations, and it is better to use the time window directly to the signal being analyzed before the Fourier transform.

For such time windows, the construction of «energy» filter increases the analysis time by about twice. But the speed of computing is not always a decisive factor, and the combined use of two different windows instead of one expands the ability to analyze.

The results of the work can be used for angular filtration of the input signal strength for various antennas, in particular, to suppress noise disturbance from the excitement of the sea surface.

There are great possibilities in optimization of the receiving system with a horizontal working direction of reception. The task of optimization in this case is solved taking into account the working range for both the average and for maximum interference.


Fourier transform; time window; filter; spectral analysis


Бендат Дж. Прикладной анализ случайных данных / Бендат Дж., Пирсол А. – М.: Мир, 1989. – 540с.

DOI: https://doi.org/10.31319/2519-2884.32.2018.173


  • There are currently no refbacks.

Copyright (c) 2018 V. U. Ihnatkin

ISSN (print) 2519-2884

ISSN (online) 2617-8389