RU2756934C1 - Method and apparatus for measuring the spectrum of information acoustic signals with distortion compensation - Google Patents

Abstract

FIELD: communication equipment.

SUBSTANCE invention relates to communication equipment, in particular, to digital methods and apparatuses for measuring the spectrum of information acoustic signals. The group of inventions includes a method containing a transform of the original acoustic signal into a primary digital information signal x_o(t), subjected to a first discrete cosine transform resulting in first spectral DCT coefficients V_pr containing interference in the form of side lobes, a mirror signal x₃(t) is formed from the primary signal, subjected to a second discrete cosine transform forming second spectral DCT coefficients V_mir, also containing interference in the form of transform side lobes, but with reverse phase. The first spectral DCT coefficients are then added to the second spectral DCT coefficients, resulting in compensation of interference in the form of transform side lobes, and an apparatus for implementation thereof.

EFFECT: increase in the accuracy of the digital method for measuring the spectrum of information acoustic signals based on distortion compensation in a discrete cosine transform.

2 cl, 10 dwg

Description

Technology area

The invention relates to communication technology, in particular to digital methods and devices for measuring the spectrum of information acoustic signals.

State of the art

Known digital method for measuring the spectrum (Christoph Rauscher "Fundamentals of Spectral Analysis". M. Rohde & Schwarz, Hotline-Telecom. 2006, p. 20, Fig. 3.6). This method includes low-frequency filtering, and then analog-to-digital conversion, as well as storing code combinations of a digital signal, fast Fourier transform, digital indication.

Known device for a digital spectrum analyzer (Christoph Rauscher "Fundamentals of Spectral Analysis". M. Rohde & Schwarz, Hotline-Telecom. 2006, p. 20, Fig. 3.6) for the implementation of a digital spectrum measurement method, containing: low-pass filter, input which is the input of the device, and the output is connected to the input of the analog-to-digital converter, as well as a random access memory, a fast Fourier transform unit, an indication unit with a display.

The disadvantage of the known method and device is a decrease in the accuracy of measuring the spectrum of information acoustic signals at short time intervals (instantaneous values of the spectrum) due to low resolution, and increased oscillation of estimates of the amplitude of spectral components. Also, in the known method and device for the fast Fourier transform, a window without overlapping is used, which leads to the appearance of discontinuities in the analyzed functions. The resulting window transform side lobes in the spectrum, called leakage, will distort the amplitudes of adjacent spectral components.

Leakage leads not only to the appearance of amplitude errors in the signal spectra, but also masks the low-amplitude components in the information signals and, therefore, prevents their measurement. In the known method and device, there is no binding of the instantaneous values of the spectrum to the segments of the temporal acoustic signal on which this spectrum is measured. In addition, in the known method and device, there is no possibility of measuring the modulus of the spectrum and the phase-frequency characteristic of the measured signal.

у которого действительная составляющая представлена в виде цифрового информационного сигнала x(t), а мнимая составляющая jx₁(t) равна нулю, после чего из цифрового комплексного сигнала осуществляют формирование последовательности сегментов из К кодовых комбинаций этого сигнала в каждом сегменте, которые преобразуются в последовательность сегментов цифрового комплексного сигнала из 2К кодовых комбинаций в каждом сегменте, после чего осуществляют наложение оконной функции Наттолла на каждый сегмент и получают последовательность сегментов цифрового комплексного сигнала из 2К кодовых комбинаций этого сигнала в каждом сегменте, затем осуществляют прямое быстрое преобразование Фурье 2К кодовых комбинаций цифрового комплексного сигнала и формируют 2К пар коэффициентов преобразования, соответствующих представлению каждого сегмента цифрового комплексного сигнала в спектральной области, после чего в каждом сегменте осуществляют в цифровом виде поворот фазы 2К пар коэффициентов преобразования, что соответствует повороту фазы на 90° всех спектральных составляющих во временной области в исходном аналоговом сигнале, а затем осуществляют обратное быстрое преобразование Фурье из 2К пар спектральных коэффициентов в каждом сегменте в 2К кодовых комбинаций в каждом сегменте цифрового комплексного сигнала, после чего осуществляют сложение с 50% перекрытием каждого сегмента цифрового комплексного сигнала с предыдущим ему сегментом и компенсацией неравномерности оконной функции Наттолла, и получают таким образом ортогональный цифровой информационный сигнал x₁(t), сдвинутый на 90° по отношению цифрового информационного сигнала x(t), а также включающий первое дискретно-косинусное преобразование, второе дискретно-косинусное преобразование и цифровую индикацию.The known "Method for measuring the spectrum of information acoustic signals of television and radio broadcasting" (patent RU 2573248 C2, published 01/20/2016, BI No. 2), taken as a prototype. This method includes low-pass filtering, and then analog-to-digital conversion with the formation of a digital information signal, which is further represented as a digital complex signal

in which the real component is represented in the form of a digital information signal x (t), and the imaginary component jx ₁ (t) is equal to zero, after which a sequence of segments is formed from the digital complex signal from K code combinations of this signal in each segment, which are converted into a sequence segments of a digital complex signal from 2K code combinations in each segment, after which the Nuttall window function is superimposed on each segment and a sequence of digital complex signal segments is obtained from 2K code combinations of this signal in each segment, then a direct fast Fourier transform of 2K code combinations of a digital complex signal and form 2K pairs of transformation coefficients corresponding to the representation of each segment of the digital complex signal in the spectral domain, after which, in each segment, the phase of 2K pairs of transformation coefficients is rotated in digital form, which corresponds to a 90 ° phase rotation of all spectral components in the time domain in the original analog signal, and then the inverse fast Fourier transform is performed from 2K pairs of spectral coefficients in each segment in 2K code combinations in each segment of the digital complex signal, after which the addition with 50 % overlap of each segment of the digital complex signal with the previous segment and compensation for the unevenness of the Nuttall window function, and thus an orthogonal digital information signal x ₁ (t) is obtained, shifted by 90 ° with respect to the digital information signal x (t), and also including the first discrete cosine transform, second discrete cosine transform and digital indication.

There is a known device for implementing a "method for measuring the spectrum of information acoustic signals of television and radio broadcasting" (patent No. RU 2573248 C2, published 01/20/2016, BI No. 2, containing a low-frequency filter, the input of which is the input of the device, and the output is connected to the input of an analog-to-digital converter, and also containing a block for doubling the frequency of sampling pulses, a block for segmentation and overlay of the Nuttall window function, a series-connected block for fast Fourier transform, a block for phase rotation of the transformation coefficients, a block for inverse fast Fourier transform, a block for overlapping segments and compensation for unevenness of the Nuttall window, and also containing the first block discrete-cosine transformation, the second block of discrete-cosine transformation, a display unit with a display,

The disadvantage of the known method and device is a decrease in the accuracy of measuring the spectrum of information acoustic signals when using a discrete-cosine transformation due to the appearance in the spectrum of interference in the form of side lobes, which can distort the amplitudes of adjacent spectral components. In addition, sidelobe interference leads not only to the appearance of amplitude errors in the signal spectra, but also masks low-amplitude components in information signals and, therefore, prevents their measurement.

The essence of the invention

The objective of the present invention is to improve the accuracy of the digital method for measuring the spectrum of information acoustic signals based on the compensation of distortions in the discrete cosine transform.

у которого действительная составляющая представлена в виде цифрового информационного сигнала x(t), а мнимая составляющая jx₁(t) равна нулю. После чего из цифрового комплексного сигнала осуществляют формирование последовательности сегментов из К кодовых комбинаций этого сигнала в каждом сегменте, которые преобразуются в последовательность сегментов цифрового комплексного сигнала из 2К кодовых комбинаций в каждом сегменте, после чего осуществляют наложение оконной функции Наттолла на каждый сегмент и получают последовательность сегментов цифрового комплексного сигнала из 2К кодовых комбинаций этого сигнала в каждом сегменте, затем осуществляют прямое быстрое преобразование Фурье 2К кодовых комбинаций цифрового комплексного сигнала и формируют 2К пар коэффициентов преобразования, соответствующих представлению каждого сегмента цифрового комплексного сигнала в спектральной области. После этого в каждом сегменте осуществляют в цифровом виде поворот фазы 2К пар коэффициентов преобразования, что соответствует повороту фазы на 90° всех спектральных составляющих во временной области в исходном аналоговом сигнале, а затем осуществляют обратное быстрое преобразование Фурье из 2К пар спектральных коэффициентов в каждом сегменте в 2К кодовых комбинаций в каждом сегменте цифрового комплексного сигнала, после чего осуществляют сложение с 50% перекрытием каждого сегмента цифрового комплексного сигнала с предыдущим ему сегментом и компенсацией неравномерности оконной функции Наттолла, и получают таким образом ортогональный цифровой информационный сигнал x₁(t), сдвинутый на 90° по отношению цифрового информационного сигнала x(t), а также включающий первое дискретно-косинусное преобразование, второе дискретно-косинусное преобразование и цифровую индикацию.The proposed method for measuring the spectrum of information acoustic signals with distortion compensation, including low-frequency filtering, and then analog-to-digital conversion with the formation of a digital information signal x (t), which is further represented as a digital complex signal

in which the real component is presented in the form of a digital information signal x (t), and the imaginary component jx ₁ (t) is equal to zero. Then, from the digital complex signal, a sequence of segments is formed from K code combinations of this signal in each segment, which are converted into a sequence of digital complex signal segments from 2K code combinations in each segment, after which the Nuttall window function is superimposed on each segment and a sequence of segments is obtained digital complex signal from 2K code combinations of this signal in each segment, then direct fast Fourier transform of 2K code combinations of the digital complex signal is performed and 2K pairs of transform coefficients are generated corresponding to the representation of each segment of the digital complex signal in the spectral domain. After that, in each segment, the phase of 2K pairs of transform coefficients is digitally rotated, which corresponds to a 90 ° phase rotation of all spectral components in the time domain in the original analog signal, and then the inverse fast Fourier transform is performed from 2K pairs of spectral coefficients in each segment to 2K code combinations in each segment of the digital complex signal, after which the addition with 50% overlap of each segment of the digital complex signal with the previous segment and compensation of the unevenness of the Nuttall window function is performed, and thus an orthogonal digital information signal x ₁ (t) is obtained, shifted by 90 ° with respect to the digital information signal x (t), and also including the first discrete-cosine transform, the second discrete-cosine transform and digital indication.

состоящий из основного цифрового информационного сигнала x_o(t) и мнимой части этого основного цифрового информационного сигнала jx_1o(t). После этого из основного цифрового комплексного сигнала формируют зеркальный цифровой комплексный информационный сигнал

из которого, в свою очередь выделяют сигнал зеркальной амплитудной огибающей А_з(t) и сигнал зеркальной фазы ϕ_з(t), кроме того основной цифровой комплексный сигнал

задерживают и из его составляющей в виде основного цифрового информационного сигнала x_o(t) производят первое дискретно-косинусное преобразование и получают первые спектральные ДКП коэффициенты В_о, содержащие помехи в виде боковых лепестков. А также из задержанного основного цифрового комплексного сигнала

выделяют сигнал приращения фазы dϕ за один дискретный отсчет, после чего этот сигнал складывают с сигналом зеркальной фазы ϕ_з(t) и формируют таким образом сигнал с корректированной зеркальной фазой ϕ_кз(t). После этого осуществляют формирование корректированного зеркального цифрового информационного сигнала х_кз(t)=А_з(х)⋅cos ϕ_кз(х), над которым осуществляют второе дискретно-косинусное преобразование и формируют вторые спектральные ДКП коэффициенты корректированного зеркального цифрового информационного сигнала Взк, которые также содержат помехи в виде боковых лепестков преобразования, но имеющих противоположную фазу по отношению помехам, содержащихся в первых спектральных ДКП коэффициентах В_о. Затем первые спектральные ДКП коэффициенты основного цифрового информационного сигнала В_о складывают со вторыми спектральными ДКП коэффициентами корректированного зеркального цифрового информационного сигнала В_зк, вследствие чего происходит компенсация помех в виде боковых лепестков преобразования, после чего над сформированными таким образом спектральными ДКП коэффициентами основного цифрового информационного сигнала с компенсированными помехами далее осуществляют цифровую индикацию.Unlike the prototype, after the analog-to-digital conversion, the digital information signal x (t) is delayed and then connected to the orthogonal digital information signal x ₁ (t) and the main digital complex signal is obtained.

consisting of the main digital information signal x _o (t) and the imaginary part of this main digital information signal jx _1o (t). After that, a mirror digital complex information signal is formed from the main digital complex signal

from which, in turn, the signal of the mirror amplitude envelope A _s (t) and the signal of the mirror phase ϕ _s (t) are extracted, in addition, the main digital complex signal

delay and from its component in the form of the main digital information signal x _o (t) perform the first discrete-cosine transformation and obtain the first spectral DCT coefficients B _o containing interference in the form of side lobes. And also from the delayed main digital complex signal

the signal of the phase increment dϕ is isolated in one discrete sample, after which this signal is added with the signal of the mirror phase ϕ _s (t) and thus a signal with the corrected mirror phase ϕ _kz (t) is formed. After that, the formation of the corrected mirror digital information signal x _kz (t) = A _z (x) ⋅ cos ϕ _kz (x) is carried out, over which the second discrete cosine transformation is carried out and the second spectral DCT coefficients of the corrected mirror digital information signal CC are formed, which also contain interference in the form of transform side lobes, but having the opposite phase with respect to the interference contained in the first spectral DCT coefficients B _o . Then the first spectral DCT coefficients of the main digital information signal B _{o are} added with the second spectral DCT coefficients of the corrected mirror digital information signal B _c , as a result of which interference is compensated in the form of side lobes of the transformation, after which over the spectral DCT coefficients of the main digital information signal with compensated by interference, digital indication is further carried out.

The problem is also solved by the fact that the device for measuring the spectrum of information acoustic signals with distortion compensation, containing a low-frequency filter, the input of which is the input of the device, and the output is connected to the input of the analog-to-digital converter, as well as a block for doubling the sampling pulse frequency, a segmentation block and superposition of the Nuttall window function, series-connected block of fast Fourier transform, block of phase rotation of transformation coefficients, block of inverse fast Fourier transform, block of overlapping segments and compensation of unevenness of the Nuttall window, and also containing the first block of discrete-cosine transform, the second block of discrete-cosine transform, display unit with display. Additionally, a first delay line, a second delay line, a complex signal formation unit, a mirror signal generation unit, a phase increment determination unit, a phase summation unit, a mirror signal recovery unit and an adder are introduced. In this case, the first output of the analog-to-digital converter is connected to the input of the Nuttall window function segmentation and superposition unit and to the input of the first delay line, the output of which is connected to the first input of the complex signal forming unit, and the second output of the analog-to-digital converter is connected to the input of the pulse frequency doubling unit sampling, and the output of the block of segmentation and superposition of the Nuttall window function is connected to the input of the block of fast Fourier transform. And the output of the block overlapping segments and compensation of unevenness of the Nuttall window is connected to the second input of the complex signal formation block, the output of which is connected to the input of the mirror signal formation block and to the input of the second delay line, the output of which is discretely connected to the input of the phase increment determination block and to the input of the first block -cosine conversion, the output of which is connected to the first input of the adder. Moreover, the first output of the mirror signal generating unit is connected to the first input of the phase summation unit, the second input of which is connected to the output of the phase increment determination unit. And the output of the phase summation unit is connected to the first input of the mirror signal recovery unit, the second input of which is connected to the second output of the mirror signal generation unit, and the output of the mirror signal recovery unit is connected to the input of the second discrete-cosine transformation unit, the output of which is connected to the second input of the adder, the output of which is connected to the input of the indication unit with a display.

List of drawings

The proposed method and device are illustrated by the figures, where:

FIG. 1. Block diagram of a device for measuring the spectrum of information acoustic signals with distortion compensation.

FIG. 2. Block diagram of the Nuttall window function segmentation and overlay.

FIG. 3. Timing diagrams of the operation of the block of segmentation and the imposition of the Nuttall window function.

FIG. 4. Block diagram of overlapping segments and compensation of unevenness of the Nuttall window.

FIG. 5. Timing diagrams of the block overlapping segments and compensation of unevenness of the Nuttall window.

FIG. 6 Diagram of the block for the formation of a mirror signal. (SSNOFN - scheme of segmentation and overlay of the Nuttall window function; SBPF - scheme of fast Fourier transform; SFZSK - scheme for generating specular spectral coefficients; SOBPF - scheme of inverse fast Fourier transform; SPSKNON - scheme of overlapping segments and compensation of unevenness of the Nuttall window; SWOF - scheme of envelope extraction and phases.)

FIG. 7 Scheme for the formation of mirror spectral coefficients included in the block for the formation of a mirror signal.

FIG. 8 Envelope and phase extraction circuit included in the mirror signal shaping unit.

FIG. 9 Diagram of the block for determining the phase increment.

FIG. 10 Block diagram of the mirror signal recovery.

Implementation of the invention

A feature of the proposed method for measuring the spectrum of information acoustic signals with distortion compensation, in contrast to the prototype, is the use of a mirror signal to compensate for distortion and interference arising from the discrete cosine transformation (DCT) of the information signal. The use of DCT with distortion compensation makes it possible to increase the accuracy of the digital method for measuring the spectrum of information acoustic signals by compensating the side lobes of the window transformation in the spectrum and thereby increasing the resolution and decreasing the oscillation of the estimate of the amplitude of the spectral components.

The method for measuring the spectrum of information acoustic signals with distortion compensation is implemented as follows. Above the input analog information acoustic signal, low-frequency filtering is carried out, and then analog-to-digital conversion with the formation of a digital information signal x (t). The sampling frequency can be, for example, 48 kHz, and the number of bits in the code combination is 16. This digital information signal is further represented as a digital complex signal x (t) = x (t) + jx ₁ (t), in which the real component is represented as digital information signal x (t), and the imaginary component jx ₁ (t) is equal to zero.

Next, from the digital complex signal, a sequence of segments is formed from K code combinations of this signal in each segment, which are converted into a sequence of digital complex signal segments from 2K code combinations in each segment, after which the Nuttall window function is superimposed on each segment and a sequence of digital segments is obtained. complex signal from 2K code combinations of this signal in each segment. This window function, in contrast to a rectangular window without overlapping, does not lead to the appearance of discontinuities in the analyzed functions and the appearance, as a result, in the spectrum of the side lobes of the window transformation, which noticeably distort the amplitudes of the neighboring spectral components. Using the Nuttall window function with the subsequent 50% overlap of each segment of 2K code combinations of the signal in each segment with the previous segment and compensation for the unevenness of this window function allows increasing the protection ratio, which characterizes the level of interference and distortion in the signal, up to 92 dB. This is essential for the transmission of artistic broadcasting signals. The Nuttall window has the lowest level of sidelobes among the existing window functions.

After imposing the Nuttall window function on each segment, 2K direct fast Fourier transform of 2K code combinations of the digital complex signal is performed and 2K pairs of transform coefficients are generated corresponding to the representation of each segment of the digital complex signal in the spectral domain. This transformation is determined by the well-known formula (N. Akhmed, K.R. Rao Orthogonal transformations in digital signal processing: Translated from English / Edited by I.B. Fomenko. - M .: Svyaz, 1980, - 248 p.) :

where: N is the number of counts, n is the harmonic number, k is the index of the signal count from 0 to N-1.

And after the fast Fourier transform in each segment, the phase of 2K pairs of transformation coefficients is digitally rotated, namely for the coefficients from the first to the Kth: the value of the coefficient cos 2πnk / N is replaced by the value of sin 2πnk / N, and the value of the coefficient jsin 2πnk / N is replaced by the value jcos 27tnk / N with the opposite sign. And for the coefficients with K plus the first to 2K minus the first: the value of the coefficient cos 2πnk / N is replaced by the value of sin 2πnk / N with the opposite sign, and the value of the coefficient jsin 2πnk / N is replaced by the value of jcos 2πnk / N. This corresponds to a 90 ° phase rotation of all time domain spectral components in the original analog signal.

Then an inverse fast Fourier transform is performed from 2K pairs of spectral coefficients in each segment in 2K codewords in each segment of the digital complex signal. For a better signal recovery in the case of using the Nuttall window, each segment of the digital complex signal is added with 50% overlap with the segment preceding it, delayed by a duration equal to half the duration of the segment, and thus a digital signal consisting of K code combinations in each segment is obtained.

Since the Nuttall window is not one of the windows providing a unit transmission coefficient when using 50% overlaps, the accuracy of the reconstructed main digital complex signal is increased by compensating for the unevenness of the Nuttall window function. Such compensation allows to increase the protection ratio, which characterizes the level of interference and distortion in the signal, up to 92 dB.

состоящий из основного цифрового информационного сигнала x_o(t) и мнимой части этого основного цифрового информационного сигнала jx_1o(t). После этого из основного цифрового комплексного сигнала

формируют зеркальный цифровой комплексный информационный сигнал

когда для каждой пары коэффициентов с порядковым номером от нуля до К (порядковый номер пар коэффициентов «i» последовательно увеличивается): "i''-я пара коэффициентов взаимно меняется по значению с "2К-i"-ой парой коэффициентов (например, если на входе имеются значения Х(3)=[-5, j300], а Х(2К-3)=[75, -j12], то на выходе будут установлены значения Х(3)=[75, -j12], а Х(2К-3)=[-5, j300]). Из полученного таким образом зеркального цифрового комплексного информационного сигнала

затем выделяют сигнал зеркальной амплитудной огибающей А_з(t) и сигнал зеркальной фазы ϕ₃(t).Thus, an orthogonal digital information signal x ₁ (t) was generated, shifted by 90 ° with respect to the digital information signal x (t). Further, from this orthogonal digital information signal and the delayed digital information signal, the main digital complex signal is formed

consisting of the main digital information signal x _o (t) and the imaginary part of this main digital information signal jx _1o (t). After that, from the main digital complex signal

form a mirror digital complex information signal

when for each pair of coefficients with a serial number from zero to K (the serial number of pairs of coefficients "i" is sequentially increased): "i" - the i-th pair of coefficients mutually change in value with the "2K-i" -th pair of coefficients (for example, if at the input there are values X (3) = [- 5, j300], and X (2K-3) = [75, -j12], then the output will be set to the values X (3) = [75, -j12], and X (2K-3) = [- 5, j300]). From the thus obtained mirror digital complex information signal

then the signal of the specular amplitude envelope A _s (t) and the signal of the specular phase ϕ ₃ (t) are separated.

In addition, the main digital complex signal x _o (t) is delayed and the first discrete-cosine transformation (DCT) of the data array X (m), m = 0, 1, is performed from its component in the form of the main digital information signal x _{o (t),} ..., N-1, in accordance with the well-known formula (N. Akhmed, K.R. Rao Orthogonal transformations in digital signal processing: Translated from English / Edited by IB Fomenko. - M .: Svyaz, 1980 , - 248 p.):

where L _x (b) is the b-th coefficient of DCT, N is the number of samples;

выделяют сигнал приращения фазы dϕ за один дискретный отсчет. Этот сигнал далее складывают с сигналом зеркальной фазы ϕ_з(t) и формируют таким образом сигнал с корректированной зеркальной фазой ϕ_кз(t). После этого осуществляют формирование корректированного зеркального цифрового информационного сигнала х_кз(t)=А_з(t) cos ϕ_кз(f), над которым осуществляют второе дискретно-косинусное преобразование (ДКП) и формируют вторые В_зк спектральных ДКП коэффициентов корректированного зеркального цифрового информационного сигнала х_кз(t), которые также содержат искажения и помехи в виде боковых лепестков преобразования, но имеющих противоположную фазу по отношению искажениям и помехам, содержащихся в первых В_о спектральных ДКП коэффициентах основного цифрового информационного сигнала x_o(t). Затем первые В_о спектральные ДКП коэффициенты основного цифрового информационного сигнала складывают со вторыми В_зк спектральными ДКП коэффициентами корректированного зеркального цифрового информационного сигнала, вследствие чего происходит компенсация искажений и помех в виде боковых лепестков преобразования. Над сформированными таким образом спектральными ДКП коэффициентами основного цифрового информационного сигнала с компенсированными искажениями и помехами далее осуществляют цифровую индикацию.and form the first B _o spectral DCT coefficients of the main digital information signal x _o (t), which contain distortion and interference in the form of transformation side lobes. And also from the delayed main digital complex signal

the signal of the phase increment dϕ is isolated in one discrete sample. This signal is then added with the signal of the mirror phase ϕ _s (t) and thus form a signal with the corrected mirror phase ϕ _kz (t). After that, the formation of the corrected mirror digital information signal x _kz (t) = A _z (t) cos ϕ _kz (f) is carried out, over which the second discrete-cosine transform (DCT) is carried out and the second B _{cc of the} spectral DCT coefficients of the corrected mirror digital information signal x _kz (t), which also contain distortion and interference in the form of transformation side lobes, but having the opposite phase in relation to distortion and interference contained in the first B _{about the} spectral DCT coefficients of the main digital information signal x _o (t). Then the first B _o spectral DCT coefficients of the main digital information signal are added with the second B _cc spectral DCT coefficients of the corrected mirror digital information signal, as a result of which distortion and interference in the form of transformation side lobes are compensated. Digital indication is then performed over the spectral DCT coefficients of the main digital information signal with compensated distortions and interference formed in this way.

The proposed method provides an increase in the accuracy of the digital method for measuring the spectrum of information acoustic signals by compensating in the spectrum of the side lobes of the window transformation and thereby increasing the resolution and reducing the oscillation of the estimation of the amplitude of the spectral components. The method provides a reduction in the duration of the segments of the information acoustic signal, on which the spectrum is measured by using a discrete cosine transformation and a mirror signal to compensate for distortions and interference arising from this discrete cosine transformation of the information signal.

The method is carried out using a device. The device for measuring the spectrum of information acoustic signals with distortion compensation (Fig. 1) consists of a low-pass filter 1, an analog-to-digital converter 2, a block for doubling the frequency of sampling pulses 3, a Nuttall window function segmentation and overlay block 4, a fast Fourier transform block 5, a block phase rotation of the transformation coefficients 6, the block of the inverse fast Fourier transform 7, the block of overlapping segments and compensation of unevenness of the Nuttall window 8, the first block of the discrete-cosine transform 9, the second block of the discrete-cosine transform 10, the display unit with the display 11, the block for generating the mirror signal 12 , the first delay line 13, the complex signal generating unit 14, the second delay line 15, the phase increment determination unit 16, the phase summing unit 17, the mirror signal recovery unit 18, the adder 19.

The input of the low-pass filter 1 is connected to the input of the device, and the output of the filter is connected to the input of the analog-to-digital converter 2, the first output of which is connected to the input of the Nuttall window function segmentation and superposition unit 4 and to the input of the first delay line 13, the output of which is connected to the first input complex signal forming unit 14, and the second output of the analog-to-digital converter 2 is connected to the input of the sampling pulse frequency doubling unit 3, and the output of the Nuttall window function segmentation and superposition unit 4 through the series-connected fast Fourier transform unit 5, the phase rotation unit of the conversion coefficients 6, block of inverse fast Fourier transform 7, block of overlapping segments and compensation of unevenness of the Nuttall window 8 is connected to the second input of the complex signal formation block 14, the output of which is connected to the input of the mirror signal formation block 12 and to the input of the second delay line 15, the output of which is connected to the input block for determining the phase increment 16 and with the input of the first block of discrete-cosine transformation 9, the output of which is connected to the first input of the adder 19, and the first output of the block for generating the mirror signal 12 is connected to the first input of the phase summation unit 17, the second input of which is connected to the output of the determination unit phase increments 16, and the output of the phase summation unit 17 is connected to the first input of the mirror signal recovery unit 18, the second input of which is connected to the second output of the mirror signal generation unit 12, and the output of the mirror signal recovery unit 18 is connected to the input of the second discrete-cosine transformation unit 10 , the output of which is connected to the second input of the adder 19, the output of which is connected to the input of the display unit with the display 11.

у которого действительная составляющая представлена в виде цифрового информационного сигнала x(t), а мнимая составляющая jx₁(t) равна нулю. Далее цифровой комплексный сигнал

в виде параллельных кодовых комбинаций с первого выхода АЦП 2 подается на вход первой линии задержки (ЛЗ₁) 13, а также на вход блока сегментации и наложения оконной функции Наттолла (БСНОФН) 4. В БСНОФН 4 осуществляется формирование последовательности сегментов цифрового комплексного сигнала из К параллельных кодовых комбинаций этого сигнала в каждом сегменте, которые преобразуются в последовательность сегментов цифрового комплексного сигнала из 2К кодовых комбинаций в каждом сегменте, после чего осуществляют наложение оконной функции Наттолла на каждый сегмент и получают последовательность сегментов цифрового комплексного сигнала из 2К кодовых комбинаций этого сигнала в каждом сегменте.The proposed method is carried out using the proposed device as follows (Fig. 1). The analog information acoustic signal is fed to the input of the device and then goes to the input of the low-pass filter (LPF) 1, with the help of which the spectrum of the acoustic signal is limited in relation to high-frequency components, for example, with a frequency of 20 kHz. Further, the informational acoustic signal from the output of the LPF 1 is fed to the input of the analog-to-digital converter (ADC) 2, where it is converted into a digital information signal x (t), for example, with a sampling frequency of 48 kHz and with the number of bits in the code combination 16. This digital information the signal is further represented as a digital complex signal

in which the real component is presented in the form of a digital information signal x (t), and the imaginary component jx ₁ (t) is equal to zero. Further digital complex signal

in the form of parallel code combinations from the first output of ADC 2 is fed to the input of the first delay line (LZ ₁ ) 13, as well as to the input of the Nuttall window function segmentation and overlay block (BSNOFN) 4. In BSNOFN 4, a sequence of digital complex signal segments is formed from K parallel code combinations of this signal in each segment, which are converted into a sequence of digital complex signal segments from 2K code combinations in each segment, after which the Nuttall window function is superimposed on each segment and a sequence of digital complex signal segments is obtained from 2K code combinations of this signal in each segment.

Then the digital complex signal from the output of BSNOFN 4 is fed to the input of the fast Fourier transform (FFT) 5 unit, where 2K point direct fast Fourier transform of 2K code combinations of the digital complex signal is carried out and 2K pairs of transform coefficients are formed corresponding to the representation of each segment of the digital complex signal in the spectral area. After that, 2K pairs of transformation coefficients from the output of the FBFT 5 are fed to the input of the phase rotation unit of the transformation coefficients (FFTKP) 6, in which the phase of 2K pairs of transformation coefficients is rotated in each segment in digital form, namely for the coefficients from the first to the K-th: the value of the coefficient cos 2πnk / N is replaced by the value of sin 2πnk / N, and the value of the coefficient jsin 2πnk / N is replaced by the value of jcos 2πnk / N with the opposite sign. And for the coefficients with K plus the first to 2K minus the first: the value of the coefficient cos 2πnk / N is replaced by the value of sin 2πnk / N with the opposite sign, and the value of the coefficient jsin 2πnk / N is replaced by the value of jcos 2πnk / N. This corresponds to a 90 ° phase rotation of all time domain spectral components in the original analog signal.

Then, 2K pairs of transform coefficients from the output of the FFTKT 6 are fed to the input of the inverse fast Fourier transform (FFFT) 7, where the transformation is carried out from 2K pairs of spectral coefficients in each segment into 2K code combinations in each segment of the digital information signal. Code combinations from the output of BOBPF 7 are fed further to the input of the block overlapping segments and compensation of unevenness of the Nuttall window (BPSKNON) 8. In BPSKNON 8, in order to better restore the signal in the case of using the Nuttall window, the addition with 50% overlap of each segment of the digital signal with the previous one is carried out. a segment delayed by a duration equal to half the duration of the segment. Thus, we obtain a digital signal consisting of K code combinations in each segment. Since the Nuttall window is not one of the windows providing a unit transmission coefficient when using 50% overlap, the increase in the accuracy of the digital signal is carried out in BPSKNON 8 by compensating for the unevenness of the Nuttall window function. Such compensation allows to increase the protection ratio, which characterizes the level of interference and distortion in the signal, up to 92 dB.

состоящего из основного цифрового информационного сигнала х_о(t) и мнимой части этого основного цифрового информационного сигнала jx_1o(t).Thus, at the output of BPSKNON 8, an orthogonal digital information signal x ₁ (t) was formed, shifted by 90 ° with respect to the digital information signal x (t) from the output of ADC 2. This orthogonal digital information signal x ₁ (t) from the output of BPSKNON 8 is fed to the second input of the complex signal formation unit (BFCS) 14, to the first input of which a delayed digital information signal x (t) is received from the output of the first LZ ₁ 13. The delay time in LZ ₁ 13 is equal to the delay time of the digital signal after passing the series-connected BSSNOFN 4, BBPF 5, BPFKP 6, BOBPF 7, BPSIKNON 8. In BFKS 14, the main digital complex signal is formed

consisting of the main digital information signal x _about (t) and the imaginary part of this main digital information signal jx _1o (t).

с выхода БФКС 14 поступает на вход блока формирования зеркального сигнала (БФЗС) 12, в котором осуществляют формирование зеркального цифрового комплексного информационного сигнала

когда для каждой пары коэффициентов с порядковым номером от нуля до К (порядковый номер пар коэффициентов «i» последовательно увеличивается): "i''-я пара коэффициентов взаимно меняется по значению с "2K-i"-ой парой коэффициентов (например, если на входе имеются значения Х(3)=[-5, j300], аХ(2K-3)=[75, -j12], то на выходе будут установлены значения Х(3)=[75, -j12], аХ(2K-3)=[-5, j300]). Из полученного таким образом зеркального цифрового комплексного информационного сигнала

затем выделяют сигнал зеркальной амплитудной огибающей А_з(х) и сигнал зеркальной фазы ϕ₃(t). При этом на первом выходе БФЗС 12 будет сформирован сигнал зеркальной фазы ϕ_з(t), а на втором его выходе будет сформирован сигнал зеркальной амплитудной огибающей А_з(t).This basic digital complex signal

from the output of the BFKS 14 enters the input of the mirror signal formation unit (BFZS) 12, in which the formation of a mirror digital complex information signal is carried out

when for each pair of coefficients with a serial number from zero to K (the serial number of pairs of coefficients "i" is sequentially increased): "i" - the i-th pair of coefficients mutually change in value with the "2K-i" -th pair of coefficients (for example, if at the input there are values X (3) = [- 5, j300], aX (2K-3) = [75, -j12], then the output will be set to the values X (3) = [75, -j12], aX ( 2K-3) = [- 5, j300]). From the thus obtained mirror digital complex information signal

then the signal of the specular amplitude envelope A _s (x) and the signal of the specular phase ϕ ₃ (t) are separated. In this case, at the first output of the BFZS 12, a signal of the mirror phase ϕ _s (t) will be generated, and at its second output, a signal of the mirror amplitude envelope A _s (t) will be generated.

с выхода БФКС 14 поступает на вход второй линии задержки ЛЗ₂ 15, время задержки которой равна времени задержки цифрового сигнала в БФЗС 12. Затем основной цифровой комплексный сигнал

с выхода ЛЗ₂ 15 поступает в виде основного цифрового информационного сигнала x_o(t) на вход первого блока дискретно-косинусного преобразования (БДКЩ) 9, в котором над этим сигналом осуществляют первое дискретно-косинусное преобразование и формируют первые Во спектральные ДКП коэффициенты основного цифрового информационного сигнала x_o(t), которые содержат искажения и помехи в виде боковых лепестков преобразования. Данные спектральные коэффициенты В_о поступают с выхода БДКП1 9 на первый вход сумматора 19.In addition, the main digital complex signal

from the output of BFKS 14 is fed to the input of the second delay line LZ ₂ 15, the delay time of which is equal to the delay time of the digital signal in the BFZS 12. Then the main digital complex signal

from the output of LZ ₂ 15 is supplied in the form of the main digital information signal x _o (t) to the input of the first block of discrete-cosine transformation (BDCS) 9, in which the first discrete-cosine transformation is performed on this signal and the first In spectral DCT coefficients of the main digital information signal x _o (t), which contain distortion and interference in the form of side lobes of the transformation. These spectral coefficients B _about come from the output of the BDKP1 9 to the first input of the adder 19.

с выхода ЛЗ₂ 15 поступает также на вход блока определения приращения фазы (БОПФ) 16, в котором из основного цифрового комплексного сигнала

выделяют сигнал приращения фазы dϕ(t) за один дискретный отсчет. Этот сигнал приращения фазы dϕ(t) с выхода БОПФ 16 поступает на второй вход блока суммирования фаз (БСФ) 17, на первый вход которого поступает сигнал зеркальной фазы ϕ_з(t) с первого выхода БФЗС 12. Данная операция суммирования фаз в БСФ 17 связана с формулой ДКП, в которой в аргументе косинуса есть множитель (2m+1). При этом, "+1" означает сдвиг на половину дискретного отсчета. Для спектральных же ДКП коэффициентов зеркального цифрового информационного сигнала нужно добавить сдвиг еще на половину дискретного отсчета (итого на полный дискретный отсчет) так как вектор этого зеркального цифрового информационного сигнала вращается навстречу вектору основного цифрового информационного сигнала. После суммирования фаз в БСФ 17, на его выходе образуется сигнал с корректированной зеркальной фазой

Basic digital complex signal

from the output of LZ ₂ 15 is also fed to the input of the phase increment determination unit (BOPF) 16, in which from the main digital complex signal

the signal of the phase increment dϕ (t) is isolated in one discrete sample. This signal of the phase increment dϕ (t) from the output of the BOPF 16 is fed to the second input of the phase summation unit (BPF) 17, the first input of which receives the signal of the mirror phase ϕ _z (t) from the first output of the BFZS 12. This operation of summing the phases in the BPF 17 is related to the DCT formula, in which the argument of the cosine contains the factor (2m + 1). In this case, "+1" means a shift by half of a discrete sample. For the spectral DCT of the coefficients of the mirror digital information signal, it is necessary to add a shift by another half of the discrete sample (total for a full discrete sample), since the vector of this mirror digital information signal rotates towards the vector of the main digital information signal. After summing the phases in the BSF 17, a signal with a corrected mirror phase is formed at its output

After that, the signal with the corrected mirror phase ϕ _kz (t) is fed to the first input of the mirror signal recovery unit (MSS) 18, to the second input of which the signal of the mirror amplitude envelope A _s (t) is fed from the second output of the BFZS 12. In the BVZS 18 from the signal with a corrected mirror phase ϕ _ks (t), a corrected mirror signal of the cosine of the phase cos ϕ _ks (t) is formed, which is multiplied by the signal of the mirror amplitude envelope A _s (t) and a corrected mirror digital information signal x _kz (t ) = A _z (t) * cos ϕ _kz (x). This signal from the output of the BVZS 18 is fed further to the input of the second block of discrete-cosine transformation (BDKT ₂ ) 10, in which a discrete-cosine transformation is performed over this signal and the second V _zk spectral DCT coefficients of the corrected mirror digital information signal x _kz (t) , which also contain distortion and interference in the form of transformation side lobes, but having the opposite phase with respect to the distortion and interference contained in the first B _{about the} spectral DCT coefficients of the main digital information signal x _o (t). These second B _cc spectral DCT coefficients of the corrected mirror digital information signal from the BDKS 10 output are fed to the second input of the adder 19. In the adder 19, the first B ₀ spectral DCT coefficients of the main digital information signal are added with the second B _cc spectral DCT coefficients of the corrected mirror digital information signal, as a result, distortion and interference in the form of conversion side lobes is compensated. These corrected spectral DCT coefficients of the main digital information signal with compensated distortions and interference are further from the output of the adder 19 to the input of the display unit with the display 11, in which the digital indication of the spectrum of the information acoustic signals is carried out.

The proposed method provides an increase in the accuracy of the digital method for measuring the spectrum of information acoustic signals by compensating in the spectrum of the side lobes of the window transformation and thereby increasing the resolution and reducing the oscillation of the estimation of the amplitude of the spectral components. The method provides a reduction in the duration of the segments of the information acoustic signal, on which the spectrum is measured by using a discrete cosine transformation and a mirror signal to compensate for distortion and interference arising from this discrete cosine transformation of the information signal.

A feature of the proposed device for measuring the spectrum of information acoustic signals by compensating for distortions is that BUCHID 3, BSNOFN 4, BPSKNON 8, BFZS 12 and BOPF 16 are non-standard in it, which require additional explanation or disclosure. In this case, the block for doubling the frequency of sampling pulses (BUCHID) 3 can be made in the form of sequentially connected: a meander shaper, a differential circuit, a full-wave rectifier and a short pulse shaper.

An example of the implementation of the block of segmentation and overlay of the Nuttall window function (BSNOFN) 4 is shown in Fig. 2, and timing diagrams of operation are shown in FIG. H. This block contains the first and second buffer memory, multiplication circuit, counter and memory circuit. The input (code) BSNOFN 4 is connected inside the block to the first (code) input of the first buffer memory, the code output of which is connected through the second buffer memory to the first code input of the multiplication circuit, the second (code) input of which is connected to the code output of the memory circuit, and the output is connected to the code output BSNOFN 4. The second input of BSNOFN 4 (not shown in Fig. 1) inside the block is connected to the second input of the first buffer memory and to the input of the counter, the output of which is connected to the third input of the first buffer memory, to the second input of the second buffer memory and to the first input of the memory circuit. The third input BSNOFN 4 (not shown in Fig. 1) inside the block is connected to the third input of the second buffer memory and to the second input of the memory circuit.

у которого действительная составляющая представлена в виде цифрового информационного сигнала x(t), а мнимая составляющая jx₁(t) равна нулю. Этот сигнал в виде параллельных кодовых комбинаций подается на первый (кодовый) вход первой буферной памяти (Фиг. 2). Одновременно на второй вход БСНОФН 4 со второго выхода АЦП 2 (фиг. 1) поступают импульсы частоты дискретизации (цепь на фиг. 1 не показана), которые внутри БСНОФН 4 подаются на вход счетчика и второй вход первой буферной памяти (Фиг. 2). На третий вход БСНОФН 4 с выхода блока удвоения частоты импульсов дискретизации 3 (Фиг. 1) поступают импульсы с удвоенной частотой дискретизации (цепь на Фиг. 1 не показана), которые внутри БСНОФН 4 подаются на третий вход второй буферной памяти и второй вход схемы памяти (Фиг. 2). При этом счетчик в БСНОФН 4 предназначен для подсчета количества кодовых комбинаций, равных половине длительности сегмента (полусегмента), на который затем накладывается оконная функция Натолла. Например, из цифрового сигнала, имеющего частоту дискретизации 48 кГц нужно сформировать последовательность сегментов, каждый из которых должен содержать К=960 дискретных отсчетов (кодовых комбинаций). Такой сегмент, в свою очередь, состоит из двух полусегментов, каждый из которых должен содержать К/2=480 дискретных отсчетов (кодовых комбинаций). При этом каждый дискретный отсчет представляет из себя, в нашем примере, 16-разрядную кодовую комбинацию. Тогда на длительности каждого полусегмента будет умещаться 480 шестнадцатиразрядных кодовых комбинаций. Именно после данного количества импульсов частоты дискретизации на выходе счетчика появляется короткий импульс, свидетельствующий об окончании данного полусегмента и начале следующего (Фиг. 3а,б). Импульсы с выхода счетчика подаются на третий вход первой буферной памяти, на второй вход второй буферной памяти и на первый вход схемы памяти.Block of segmentation and overlay window function Nuttall 4 (Fig. 2) works as follows. Initially, the first and second buffer memories and the counter are zeroed. The memory circuit is also in its initial state when at its code output there is a code combination corresponding to the Natall window gain for the first of the K parallel code combinations (discrete samples) of the digital information signal in the segment. The code input BSNOFN 4 from the first (code) output of ADC 2 (Fig. 1) receives a digital information signal, which is represented as a digital complex signal

in which the real component is presented in the form of a digital information signal x (t), and the imaginary component jx ₁ (t) is equal to zero. This signal in the form of parallel code combinations is fed to the first (code) input of the first buffer memory (Fig. 2). Simultaneously to the second input BSNOFN 4 from the second output of ADC 2 (Fig. 1) pulses of the sampling frequency (the circuit is not shown in Fig. 1), which inside the BSNOFN 4 are fed to the input of the counter and the second input of the first buffer memory (Fig. 2). At the third input of BSNOFN 4 from the output of the block of doubling the frequency of sampling pulses 3 (Fig. 1), pulses with a doubled sampling frequency are received (the circuit is not shown in Fig. 1), which inside the BSNOFN 4 are fed to the third input of the second buffer memory and the second input of the memory circuit (Fig. 2). In this case, the counter in BSNOFN 4 is designed to count the number of code combinations equal to half the duration of a segment (half-segment), which is then superimposed on the Nutall window function. For example, from a digital signal with a sampling rate of 48 kHz, it is necessary to form a sequence of segments, each of which should contain K = 960 discrete samples (code combinations). Such a segment, in turn, consists of two half-segments, each of which must contain K / 2 = 480 discrete samples (code combinations). Moreover, each discrete sample is, in our example, a 16-bit code combination. Then 480 sixteen-bit code combinations will fit in the duration of each half-segment. It is after a given number of pulses of the sampling rate that a short pulse appears at the output of the counter, indicating the end of this half-segment and the beginning of the next one (Fig. 3a, b). The pulses from the counter output are fed to the third input of the first buffer memory, to the second input of the second buffer memory and to the first input of the memory circuit.

The first buffer memory in BSNOFN 4 contains K / 2 = 480 code combinations (half-segment), and the second buffer memory consists of two halves and contains K = 960 code combinations (two half-segments of 480 code combinations). As parallel code combinations arrive at 1 code input of the first buffer memory, they are written into it under the action of pulses with a sampling frequency. These codewords appear at the code output of the first buffer memory and are applied to the code input of the second buffer memory, but are not written to it. At the same time, K = 960 zero code combinations are read from the second buffer memory under the action of pulses with a doubled sampling rate. These zero 16-bit codewords are sequentially fed to the first code input of the multiplication circuit. At this time, 16-bit code combinations are fed to the second code input of this circuit, corresponding to the Natall window transmission coefficients. After multiplying the code combinations fed to the 1 and 2 code inputs of the multiplication circuit, its output will also be zero 16-bit code combinations.

Thus, during the period when the first buffer memory is filled with code combinations corresponding to the first half-segment ( ₁ p.s. in Fig. 3a), the first segment is formed at the output of the multiplication circuit (0 1 -0 ₀ segm. In Fig. 3c ) from zero code combinations. After filling 480 sixteen-bit code combinations of the first buffer memory, the first short pulse appears at the counter output (Fig.3b) under the action of the leading edge of which these codewords from the first buffer memory are written into the first half of the second buffer memory (1 p.s. in Fig. 3a ). Under the action of the same short pulse, 480 zero codewords from the first half of the second buffer memory are shifted and written into the second half of this buffer memory (0 p.s. in Fig. 3a). Thus, the first segment is formed from the zero and the first half-segments (1 segment in Fig. 3a). Under the influence of the decay of the same short pulse, the first buffer memory and the memory circuit are reset to their initial state. In this case, a code combination appears at the code output of the memory circuit, corresponding to the Natall window transmission coefficient for the first of K = 960 code combinations in the first segment (1 segment in Fig. 3a). It should be noted that the Natall window gains (and their corresponding codewords) for the first half of the segment (for example, 0 pp in 1 segment in Fig. 3a) are increasing, and for the second half of the segment (for example, 1 pp. in 1 segment in Fig. 3a) are decreasing.

Parallel code combinations that continue to arrive at 1 code input of the first buffer memory are written into this memory under the action of pulses with a sampling frequency. At the same time, under the action of pulses with a doubled sampling rate at the third input of the second buffer memory and the second input of the memory circuit, 16-bit code combinations from their code outputs are fed to the first and second code inputs of the multiplication circuit, respectively. Zero codewords (from the second half of the second buffer memory) of the zero half-segment of the first segment (1 segment in Fig. 3a) are multiplied first, therefore only zero 16-bit codewords appear at the code output of the multiplication circuit. Next, the information code combinations (from the first half of the second buffer memory) of the first half-segment of the first segment (1 segment in Fig. 3a) begin to be multiplied, therefore multiplied 16-bit code combinations appear at the code output of the multiplication circuit, corresponding to the original code combinations, but with superimposed on them by the Natall window transmission coefficients.

That. during the period when the first buffer memory is filled with code combinations corresponding to the second half-segment (1 _{p.s. in Fig. 3a) at the output of the multiplication circuit, the second segment is formed (1 1} -0 ₂ segments. in Fig. 3c), consisting of from the second time used the zero half-segment and the first time used the first half-segment (in which the Natall window gains are decreasing). After filling the next 480 code combinations of the first buffer memory, a second short pulse appears at the counter output (Fig. 3b) under the action of the leading edge of which these code combinations are written into the first half of the second buffer memory. Under the action of the same short pulse, 480 previously recorded code combinations from the first half of the second buffer memory are shifted and written into the second half of this buffer memory. Thus, the second segment is formed from the first and second half-segments (2 segments in Fig. 3a). Under the influence of the decay of the same short pulse, the first buffer memory and the memory circuit are reset to their initial state. In this case, a code combination appears at the code output of the memory circuit, corresponding to the Natall window transmission coefficient for the first of K = 960 code combinations in the second segment (2 segments in Fig. 3a).

Under the action of pulses at the third input of the second buffer memory and the second input of the memory circuit, 16-bit code combinations from their code outputs are fed to the first and second code inputs of the multiplication circuit, respectively. The code combinations (from the second half of the second buffer memory) of the first half-segment of the second segment (2 segments in Fig. 3a) are multiplied first. These multiplied codewords appear at the output of the multiplication circuit. Next, the code combinations (from the first half of the second buffer memory) of the second half-segment of the second segment (2 segments in Fig. 3a) are multiplied. These multiplied codewords also appear at the output of the multiplication circuit. That. at the output of the multiplication circuit, the third segment is formed (2 ₁ -1 ₂ segm. the Natall window gains are decreasing).

While 16-bit codewords are read from the second buffer memory, codewords corresponding to the third half-segment are written into the first buffer memory (3 p.s. in Fig. 3a). After filling with the next 480 code combinations of the first buffer memory, a third short pulse appears at the counter output (Fig. 3b) under the action of the leading edge of which these code combinations are written into the first half of the second buffer memory. Under the action of the same short pulse, 480 previously recorded code combinations from the first half of the second buffer memory are shifted and written into the second half of this buffer memory. Thus, the third segment is formed from the second and third half-segments (3 segments in Fig. 3a). Under the influence of the decay of the same short pulse, the first buffer memory and the memory circuit are reset to their initial state. In this case, a code combination appears at the code output of the memory circuit, corresponding to the Natall window transmission coefficient for the first of K = 960 code combinations in the third segment (3 segments in Fig. 3a).

Under the action of pulses at the third input of the second buffer memory and the second input of the memory circuit, 16-bit code combinations from their code outputs are fed to the first and second code inputs of the multiplication circuit, respectively. The code combinations (from the second half of the second buffer memory) of the second half-segment of the third segment (3 segments in Fig. 3a) are multiplied first. These multiplied codewords appear at the output of the multiplication circuit. Next, the code combinations (from the first half of the second buffer memory) of the third half-segment of the third segment (3 segments in Fig. 3a) are multiplied. These multiplied codewords also appear at the output of the multiplication circuit. That. at the output of the multiplication circuit, the fourth segment (3 ₁ -2 ₂ segments in Fig. 3c) is formed, consisting of the second half-segment used a second time (in which the Natall window gains are increasing) and the third half-segment used for the first time (in which the Natall window gains are decreasing).

Further, the work of BSNOFN 4 proceeds in a similar way. Thus, the sequence of digital signal segments formed in BSNOFN 4 from K parallel code combinations of this signal in each segment (or from K / 2 code combinations in a half-segment) is converted (due to the doubled sampling rate) into a sequence of digital signal segments from 2K parallel code combinations of this signal in each segment with Nuttall windowing superimposed on them.

An example of the implementation of the block overlapping segments and compensation of unevenness of the Nuttall window (BPSKNON) 8 is shown in Fig. 4, and timing diagrams of operation are shown in FIG. 5. This block contains: the first, second, third and fourth buffer memories (BP), adder, memory circuit (SP), multiplication circuit (MS), counter, trigger, shaper, delay element (EZ). The first (code) input of the first buffer memory (BP ₁ ) is connected to the (code) input of BPSKNON 8, and its code output is connected to the first (code) input of the second buffer memory (BP ₂ ) and to the first (code) input of the third buffer memory ( BP ₃ ). The second input of the power supply _{unit 1 is} connected to the output of the delay element EZ, and the third input of the power supply _{unit 1 is} connected to the second input of BPSKNON 8 (not shown in Fig. 1), to which the input of the counter is also connected, the output of which is connected to the input of the trigger, the input of the EZ and to the second the input of the power supply _{unit 2} , the code output of which is connected to the first (code) input of the power supply unit ₄ . The third input BPSKNON 8 (not shown in Fig. 1) is connected to the second input of the memory circuit (SP), the second input BP ₃ and the second input BP ₄ . The trigger output is connected to the input of the shaper, the output of which is connected to the first input of the joint venture, to the third input of the power supply unit ₃ and to the third input of the power supply _{unit 4} . Code outputs BP ₃ and BP _{4 are} connected, respectively, with the first and second code inputs of the adder, the code output of which is connected to the first code input of the multiplication circuit (MS), the second code input of which is connected to the code output of the SP, and the code output of the CS is connected to the code output BPSKNON 8.

BPSKNON 8 (Fig. 4) works as follows. In the initial state, BSh, BSh, BP ₃ , BP ₁ , the counter, as well as the trigger are reset. The SP is also in the initial state when at its code output there is a codeword corresponding to the gain to compensate for the unevenness of the Nuttall window function for the first of the K codewords in the first slot. Parallel code combinations of an orthogonal digital signal from the code output of BOBFT 7 (Fig. 1) are fed to the (code) input of BPSKNON 8 (Fig. 4) and further to the first (code) input of the BP _1. At the same time, pulses with a double sampling rate from the output of BUCHID 3 (the circuit is not shown in Fig. 1) are fed to the second input BPSKNON 8 (not shown in Fig. 1), which are then fed to the third input of the BP ₁ (Fig. 4). Under the influence of these pulses, the code combinations arriving at the input of BP ₁ are written into it and appear at the code output of BP ₁ . These code combinations are applied to the first (code) inputs BP ₂ and BP ₃ , but are not recorded in them.

At the same time, the counter starts counting pulses at twice the sampling rate. This counter is designed to count the number of code combinations equal to half the duration of a segment (half segment). For example, from a digital signal having a doubled sampling rate, it is necessary to form a sequence of half-segments, each of which should contain K / 2 = 480 discrete samples (code combinations). In this case, each discrete sample is a 16-bit code combination. Then 480 sixteen-bit code combinations will fit in the duration of each half-segment. It is after a given number of pulses with a doubled sampling rate that a short pulse appears at the output of the counter, indicating the end of this half-segment and the beginning of the next one (Fig. 5a, b). BP ₁ , BP ₂ , BP ₃ , BP ₄ in our example, each contain 480 sixteen-bit code combinations (i.e., each in a half-segment), Code combinations from the code outputs of the adder, SU and SP are also 16-bit. BPSKNON 8 is intended for the formation of segments of an orthogonal digital signal from K code combinations in each segment and addition with 50% overlap of each segment with the previous segment. In order to avoid discontinuities in the sequence of the digital signal formed after overlapping the segments, it is necessary that the code combinations are written to the BP ₁ with a double sampling rate, and the code combinations from the BP ₃ and BP ₁ are read with the sampling frequency.

Simultaneously with the writing of the code combinations to the BP ₁ , from the BP ₃ and the BP ₄ , zero code combinations are read under the action of pulses at their second inputs. These zero 16-bit codewords are fed to the first and second code inputs of the adder, the output of which will also be zero 16-bit codewords, which are fed to the first code input of the SD. At this time, 16-bit code combinations are supplied to the second code input of this circuit from the code output of the SP, which correspond to the transmission coefficients to compensate for the unevenness of the Nuttall window function. After multiplying the code combinations fed to 1 and 2 code inputs of the control system, its code output will also have zero 16-bit code combinations. That. during the filling of BP _{1 with} code combinations corresponding to the first half-segment (0 ₀ p.s. in Fig. 5a) at the code output of the control system, a half-segment (0 _n in Fig. 5d) is formed from zero code combinations. After filling 480 with sixteen-bit zero code combinations BP ₁ , corresponding to 0 ₀ - half-segment (Fig. 5a), the first short pulse appears at the output of the counter (Fig. 5b) from which a trigger is triggered, and a short pulse appears at the output of the shaper.

Under the action of the leading edge of the pulse from the output of the shaper arriving at the third inputs of BP ₃ and BP ₄ , zero code combinations corresponding to the 0 ₀ half-segment from the output of BP _{1 are} written into BP ₃ , and in BP ₄ , zero code combinations are also written, which were present at BP ₂ . Thus, from 0 and 0 ₀ half-segments (Fig. 5a) the first segment is formed (1 segment in Fig. 5a - below). At the same time, under the action of the same short pulse from the output of the shaper, the SP is set to its initial state, when a code combination appears at its code output, corresponding to the transmission coefficient to compensate for the unevenness of the Nuttall window function for the first codeword in the segment. After that, under the influence of the decay of the pulse from the output of the counter arriving at the second input of the power supply unit ₂ , the code combinations from the code output of the power supply _{unit 1} corresponding to the 0 ₀ half-segment are written into the power supply unit ₂ and appear at its code output. In addition, under the action of the same short pulse delayed in the EZ, the BP ₁ is reset to zero and starts writing the code combinations corresponding to the next 0 ₁ half-segment (Fig. 5a).

Under the influence of pulses with a sampling rate at the second inputs of the BP ₃ and BP ₄ , 16-bit zero code combinations from their code outputs are fed to the first and second code inputs of the adder, respectively. Further, zero codewords from the code output of the adder (0 ₀ p. S. +0 p. S. In Fig. 5a) are fed to the first code input of the CS, the second code input of which receives the code combinations from the output of the SP. That. at the output of the control system, the first segment of the orthogonal digital signal is formed (0 ₀ + 0 segm. in Fig. 5c). While reading of 16-bit code combinations is carried out from BP ₃ and BP ₄ slowed by 2 times (compared to the write speed in BP ₁ ), code combinations corresponding to 0 ₁ half-segment are _{written to BP 1.} After filling 480 with zero code combinations of the power supply unit ₁ , a second short pulse appears at the output of the counter (Fig. 5b) under the action of which the trigger is triggered and at its output a "logical 0"("log.0") appears, from which no signal, and hence the recording in BP ₃ and BP _{4 of} parallel code combinations from BP ₁ and BP ₂ does not occur. At this time, reading, addition and multiplication of zero code combinations corresponding to 0 ₀ and 0 half-segments _{continues from BP 3} and BP _{4 , and 0 0} -0 segment is formed (Fig. 5c). Under the influence of the decay of the pulse from the counter output, the code combinations from the code output of the BP ₁ corresponding to the 0 ₁ half-segment are written into the BP ₂ and appear at its code output. In addition, under the action of a short pulse delayed in the EZ, the BP ₁ is reset to zero and starts writing the code combinations corresponding to the next 0 ₂ half-segment (Fig. 5a).

After filling with zero code combinations BP ₁ (0 ₂ p.s. in Fig. 5a), a third short pulse appears at the output of the counter (Fig. 5b) under the action of which the trigger is triggered and a "logical 1"("log. 1 "), From which a second short pulse appears at the output of the shaper (Fig. 5c). Under the action of the leading edge of the pulse from the output of the shaper, zero code combinations corresponding to the 0 ₂ half-segment, from the output of BP _{1, are} written to BP ₃ , and in BP ₄ , zero code combinations are also written corresponding to 0 ₁ and which were present in BP ₂ . Thus, from 0 ₂ and 0 ₁ half-segments, the second segment of the orthogonal digital signal is formed (2 segments in Fig. 5a - at the bottom). At the same time, under the action of the same short pulse from the output of the shaper, the SP is set to its initial state, when a code combination appears at its code output, corresponding to the transmission coefficient to compensate for the unevenness of the Nuttall window function for the first codeword in the segment. After that, under the influence of the decay of the pulse from the counter output, the code combinations from the code output of the BP ₁ corresponding to the 0 ₂ half-segment are written into the BP ₂ and appear at its code output. In addition, under the action of a short pulse delayed in the EZ, BP ₁ is reset to zero and starts writing code combinations corresponding to the next h-half-segment (Fig. 5a).

Under the action of sampling pulses at the second inputs of BP ₃ and BP ₄ , 16-bit zero code combinations from their code outputs are fed to the first and second code inputs of the adder, respectively. Further, zero codewords from the code output of the adder (0 ₂ p.s. +0 ₁ p.s. in Fig. 5a) are fed to the first code input of the CS, to the second code input of which the code combinations are received from the output of the DT. At the code output of the CU, zero 16-bit code combinations appear. That. at the output of the control unit, the second segment of the orthogonal digital signal is formed (0 _{2 + 0 1} segm. in Fig. 5d). While from PD ₃ and PD ₄ reads 16-bit codewords (0 ₂ n. S. And 0 ₁ n. S. FIG. 5a), in the BS ₁ is recorded codewords corresponding to h half portion (1 ₁ n. S. in Fig.5a). After filling with code combinations BP ₁ (1 ₁ p. S in Fig. 5a), a fourth short pulse appears at the output of the counter (Fig. 5b) under the action of which the trigger is triggered and at its output “log. 0 ", from which no signal arises at the output of the shaper, which means that no code combinations from BP ₁ and BP ₂ are _{written to BP 3} and BP _4. At this time, reading, addition and multiplication of zero code combinations corresponding to 0 ₂ and 0 ₁ half-segments _{continues from BP 3} and BP _{4 , and 0 2} -0 ₁ segment is formed (Fig. 5d).

Under the influence of the decay of the pulse from the counter output, the code combinations from the code output BP ₁ , corresponding to 1 ₁ half-segment are written into the BP ₂ and appear at its code output. In addition, under the action of a short pulse delayed in the EZ, the BP ₁ is reset to zero and starts writing the code combinations corresponding to the next 1 ₂ half-segment (Fig. 5a). After filling with code combinations BP ₁ (1 ₂ p. S. And in Fig. 5a), a fifth short pulse appears at the output of the counter (Fig. 5b), under the action of which the trigger is triggered and at its output “log. 1 ", from which a third short pulse appears at the output of the shaper (Fig. 5c). Under the action of this impulse, code combinations from the code outputs BP ₁ and BP ₂ . are recorded, respectively, in BP ₃ and BP ₄ . Thus, from 1 ₂ and 1 ₁ half-segments, the third segment of the orthogonal digital signal is formed (3 segments in Fig. 5a - below). At the same time, under the action of the same short pulse, the memory block is reset to its initial state, when a code combination appears at its code output, corresponding to the transmission coefficient to compensate for the unevenness of the Nuttall window function for the first codeword in the third segment.

After that, under the influence of the decay of the pulse from the counter output, the code combinations from the code output of the BP ₁ corresponding to the 1 ₂ half-segment are written into the BP ₂ and appear at its code output. In addition, under the action of a short pulse delayed in the EZ, the BP ₁ is reset to zero and starts writing the code combinations corresponding to the next 2 ₁ half-segment (Fig. 5a). Under the action of pulses with a sampling rate at the second inputs of the BP ₃ and BP ₄ , 16-bit information code combinations from their code outputs are fed to the first and second code inputs of the adder, respectively. During the summation, the code combinations included in the 1 ₂ half-segment (in which the transmission coefficients of the Natall window are increasing) are added with the same code combinations included in the 1 ₁ half-segment (in which the transmission coefficients of the Natall window are decreasing), therefore, at the output of the adder, the transmission coefficients Nutall windows flatten out (close to 1), although some unevenness remains.

Further, after summation, the code combinations from the code output of the adder (1 _{2 p.} +1 ₁ p. S. In Fig. 5a) are fed to the first code input of the CS, the second code input of which receives the code combinations from the output of the SP. After multiplying the code combinations, the Nuttall window function non-uniformity is compensated. If we compare (1 ₁ -0 ₂ ) segment and (2 ₁ -1 ₂ ) segment (at the top of Fig. 5a) at the input of BPSKNON 8 with 3 segments (3 segments in Fig. 5a or 1 ₂ +1 ₁ segments in Fig. 5d) at the output of the adder, it can be seen that there is an addition with 50% overlap of the segment with the previous segment. 16-bit code combinations with compensated non-uniformity of the Nuttall window function are fed to the code output of the CU. That. at the output of the control unit, the third segment of the orthogonal digital signal is formed (1 ₂ +1 ₁ segment in Fig. 5 d). While from PD ₃ and PD ₄ reads 16-bit codewords (1 ₂ n. S., 1 ₁ n. S. FIG. 5a), in the BS ₁ is recorded codewords corresponding 2 _one half portion (2 ₁ n. C . in Fig. 5a). After filling with code combinations (2 ₁ p. C in Fig. 5a) of the PSU ₁ , the sixth short pulse appears at the output of the counter (Fig. 5b) under the action of which the trigger is triggered and "log.0" appears at its output, from which at the output of the shaper there is no signal, and therefore no _{parallel code combinations are written to BP 3} and BP ₄ from the code outputs of BP ₁ and BP ₂ .

At this time, reading, addition and multiplication of code combinations corresponding to 1 ₂ and 1 ₁ half-segments _{continues from BP 3} and BP _{4 , and 1 2} -1 ₁ segment is formed (Fig. 5 d). Under the influence of the decay of the pulse from the counter output, the code combinations from the code output of the BP ₁ , corresponding to the 2 ₁ half-segment, are written to the BP ₂ and appear at its code output. In addition, under the action of a short pulse delayed in the EZ, the BP ₁ is reset to zero and starts recording the code combinations corresponding to the next 2 ₂ half-segment (Fig. 5a). After filling PD ₁ codewords corresponding -polusegmentu February ₂ (2 _2n. S. FIG. 5a) at the output of the seventh counter appear shorter pulse (FIG. 5b), which is triggered by the action of a trigger and its output appears on the "log. 1 ", from which a fourth short pulse appears at the output of the shaper (Fig. 5c). Under the action of this pulse, the code combinations from the code outputs BP ₁ and BP _{2 are} written into BP ₃ and BP ₄ . Thus, from 2 ₂ and 2 ₁ half-segments, the fourth segment is formed (4 segments in Fig. 5a below). At the same time, under the action of the same short pulse, the memory block is reset to its initial state, when a code combination appears at its code output, corresponding to the transmission coefficient to compensate for the unevenness of the Nuttall window function for the first codeword in the fourth segment.

After that, under the influence of the decay of the pulse from the counter output, the code combinations from the code output BP ₁ corresponding to the 2 ₂ half-segment are written into the BP ₂ and appear at its code output. In addition, under the action of a short pulse delayed in the EZ, the BP ₁ is reset to zero and starts writing the code combinations corresponding to the next 3 ₁ half-segment (Fig. 5a). Under the action of sampling pulses at the second inputs of BP ₃ and BP ₄ , 16-bit information code combinations from their code outputs are fed to the first and second code inputs of the adder, respectively. When summing, the code combinations included in the 2 ₂ half-segment are added (in which the Natall window transmission coefficients are increasing) with the same code combinations included in the 2 ₁ half-segment (in which the Natall window transmission coefficients are decreasing), therefore, at the output of the adder, the transmission coefficients Nutall windows flatten out (close to 1), although some unevenness remains.

Further, after summation, the code combinations from the code output of the adder (2 _{2 p.} +2 ₁ p. S. In Fig. 5a) are fed to the first code input of the CS, the second code input of which receives the code combinations from the output of the SP. After multiplying the code combinations, the Nuttall window function non-uniformity is compensated. If we compare (2 ₁ -1 ₂ ) segment and (3 ₁ -2 ₂ ) segment (above Fig. 5a) at the input of BPSKNON 8 with 4 segments (4 segments in Fig. 5a below or 2 ₂ +2 ₁ segments on Fig. 5 d) at the output of the adder, it can be seen that there is an addition with 50% overlap of the segment with the previous segment. 16-bit code combinations with compensated non-uniformity of the Nuttall window function are fed to the code output of the CU. That. at the output of the control unit, the fourth segment of the orthogonal digital signal is formed (2 ₂ +2 ₁ segment in Fig. 5 d).

Further, the work of BPSKNON 8 proceeds in a similar way. Thus, at the output of BPSKNON 8, an orthogonal digital information signal x ₁ (t) was generated, shifted by 90 ° with respect to the digital information signal x (t).

An example of an implementation of a mirror signal forming unit (BFZS) 12 is shown in FIG. 6. This block contains serially connected: a Nuttall window function segmentation and overlay scheme (SSNOFN), a fast Fourier transform (FFT) scheme, a mirror spectral coefficients formation scheme (FFTS), an inverse fast Fourier transform (FFFT) scheme, a segment overlap and compensation scheme Nuttall window unevenness (SPSCNON), envelope and phase extraction scheme (SWOF). In this case, SSNOFN, SBPF, SOBPF and SPSKNON are similar, respectively, BSNOFN 4, BBPF 5, BOBPF 7, BPSKNON 8. Input (code) BFZS 12 is connected to the output of BFKS 14 (Fig. 1), and inside BFZS 12 (Fig. 6 ) its input is connected to the SSNOF input. Moreover, the first output of the SWOF is the first output of the BFZS 12, and the second output of the SWOF is the second output of the BFZS 12.

с выхода БФКС 14 (фиг. 1) в виде параллельных кодовых комбинаций поступает на вход БФЗС 12, а внутри БФЗС 12 (Фиг. 6) этот сигнал поступает на вход ССНОФН. В ССНОФН осуществляется формирование последовательности сегментов основного цифрового комплексного сигнала из К параллельных кодовых комбинаций этого сигнала в каждом сегменте, которые преобразуются в последовательность сегментов основного цифрового комплексного сигнала из 2К кодовых комбинаций в каждом сегменте, после чего осуществляют наложение оконной функции Наттолла на каждый сегмент и получают последовательность сегментов основного цифрового комплексного сигнала из 2К кодовых комбинаций этого сигнала в каждом сегменте. Затем основной цифровой комплексный сигнал с выхода ССНОФН поступает на вход СБПФ, где осуществляется 2К точечное прямое быстрое преобразование Фурье 2К кодовых комбинаций основного цифрового комплексного сигнала и формируется 2К пар спектральных коэффициентов преобразования, соответствующих представлению каждого сегмента основного цифрового комплексного сигнала в спектральной области. После этого 2К пар спектральных коэффициентов преобразования с выхода СБПФ подаются на вход СФЗСК, в котором осуществляется в каждом сегменте в цифровом виде формирование 2К пар зеркальных спектральных коэффициентов преобразования, когда для каждой пары коэффициентов с порядковым номером от нуля до К (порядковый номер пар коэффициентов «i» последовательно увеличивается): "i''-я пара коэффициентов взаимно меняется по значению с "2К -i"-ой парой коэффициентов (например, если на входе имеются значения Х(3)=[-5, j300], а Х(2K-3)=[75, -j12], то на выходе будут установлены значения Х(3)=[75, -j12], аХ(2K-3)=[-5,j300]). Это соответствует формированию зеркальных спектральных составляющих во временной области в исходном аналоговом сигнале.BFZS 12 works as follows. Basic digital complex signal

from the output of the BFKS 14 (Fig. 1) in the form of parallel code combinations is fed to the input of the BFZS 12, and inside the BFZS 12 (Fig. 6) this signal is fed to the SSNOFN input. In SSNOFN, a sequence of segments of the main digital complex signal is formed from K parallel code combinations of this signal in each segment, which are converted into a sequence of segments of the main digital complex signal from 2K code combinations in each segment, after which the Nuttall window function is superimposed on each segment and a sequence of segments of the main digital complex signal of 2K code combinations of this signal in each segment. Then the main digital complex signal from the SSNOFN output is fed to the FFT input, where 2K point direct fast Fourier transform of 2K code combinations of the main digital complex signal is carried out and 2K pairs of spectral transform coefficients are formed, corresponding to the representation of each segment of the main digital complex signal in the spectral domain. After that, 2K pairs of spectral transform coefficients from the output of the FFT are fed to the input of the FDSC, in which 2K pairs of mirror spectral transform coefficients are formed in each segment in digital form, when for each pair of coefficients with a serial number from zero to K (the serial number of pairs of coefficients " i "sequentially increases):" i "- the i-th pair of coefficients mutually changes in value with the" 2K-i "-th pair of coefficients (for example, if the input has values X (3) = [- 5, j300], and X (2K-3) = [75, -j12], then the output will be set to the values X (3) = [75, -j12], aX (2K-3) = [- 5, j300]). This corresponds to the formation of mirror spectral components in the time domain in the original analog signal.

Then, 2K pairs of mirror spectral transform coefficients from the output of the SFZSK are fed to the SOBPF input, where the conversion is carried out from 2K pairs of mirror spectral coefficients in each segment to 2K code combinations in each segment of the mirror digital complex information signal. Code combinations from the SOBPF output are fed further to the SPSKNON input. In SPSCNON, in order to better restore the mirror signal in the case of using the Nuttall window, the addition with 50% overlap of each segment of the digital mirror signal with its previous segment, delayed by a duration equal to half the duration of the segment. Thus, we obtain a mirror digital complex information signal consisting of K code combinations in each segment. Since the Nuttall window is not one of the windows providing a unit transmission coefficient when using 50% overlap, the increase in the accuracy of the mirror digital complex signal is carried out in SPSCNON by compensating for the unevenness of the Nuttall window function. Such compensation allows to increase the protection ratio, which characterizes the level of interference and distortion in the signal, up to 92 dB.

Этот сигнал с выхода СПСКНОН поступает на вход СВОФ, в которой осуществляют выделение сигнала зеркальной фазы ϕ_з(t) и сигнала зеркальной амплитудной огибающей А_з(t), которые поступают, соответственно, с первого и второго выходов СВОФ на, соответственно, первый и второй выходы БФЗС 12.Thus, at the output of SPSCNON (Fig. 6), a mirror digital complex information signal was formed

This signal from the SPSCNON output is fed to the SWOF input, in which the signal of the mirror phase ϕ _s (t) and the signal of the mirror amplitude envelope A _s (t) are extracted, which are received, respectively, from the first and second outputs of the SWOF to the first and second outputs of BFZS 12.

An example of the implementation of the scheme for the formation of specular spectral coefficients (MFZSK) included in the BFZS 12 is shown in FIG. 7. This circuit contains the first buffer memory (BP ₁ ), the second buffer memory (BP ₂ ) and a counter. The first (code) input of the SFZSK is connected to the output of the SBPF (Fig. 6), and inside the SFZSK this input is connected to the first (code) input of the BP ₁ (Fig. 7), the code output of which is connected to the first (code) input of the BP ₂ , the code the output of which is the code output of the SFZSK. In this case, the second input of the SFZSK is connected to the input of the counter and to the second inputs of BP ₁ and BP ₂ , the third inputs of which are connected to the output of the counter.

SFZSK works as follows. In the initial state, PSU ₁ and PSU ₂ and the counter are reset to zero. At the first (code) input of the SFCS, 2K pairs of spectral transform coefficients in each segment are received from the output of the SFT (Fig. 6). Inside the SFZSC (Fig. 7), these 2K pairs of spectral transform coefficients in each segment are fed to the first (code) input of the BP ₁ . Pulses with doubled sampling frequency (the circuit is not shown in Fig. 1 and Fig. 6) are received from the output of the block of doubling the frequency of sampling pulses 3 (Fig. 1) to the second input of the SFZSK from the output of the block of doubling the frequency of sampling pulses (the circuit is not shown in Fig. 1 and Fig. 6), which inside the SFZSK (Fig. 7) are fed to the input of the counter and to the second inputs of BP ₁ and BP ₂ . Under the action of these pulses with a doubled sampling rate of 2K pairs of spectral transform coefficients included in the first segment are written into the BP ₁ and appear at its code output. These 2K pairs of spectral transform coefficients of the first segment are applied to the first (code) input of the PSU ₂ , but are not written into it. Under the action of the same pulses with a doubled sampling rate, zero values of 2K pairs of spectral transform coefficients from BP ₂ begin to be read from the end of the segment and these zero values appear at the code output of BP ₂ . At the same time, pulses with a doubled sampling frequency are fed to the input of the counter, which starts counting them and, with the appearance of a sampling pulse corresponding to the end of the segment, a short pulse appears at the output of the counter. Under the action of the leading edge of this pulse arriving at the third input of BP ₂ , 2K pairs of spectral transform coefficients included in the first segment and present at the code output of BP ₁ and the code input of BP ₂ are recorded in BP ₂ . Under the influence of decrease of the same short pulse supplied to the third input of the BP ₁ installs the PD ₁ to the initial state.

Further, the next 2K pairs of spectral transform coefficients included in the second segment are recorded in the BP ₁ under the action of pulses at its second input. These spectral transform coefficients appear at the code output of the power supply unit ₁ and are applied to the first (code) input of the power supply unit ₂ , but are not written into it. At the same time, under the influence of the same pulses with a doubled sampling rate, arriving at the second input of the PSU ₂ , the values of 2K pairs of spectral transform coefficients included in the first segment are read from the PSU from the end of this segment to its beginning, as a result of which the "i" -th pair coefficients mutually change in value with the "2K-i" -th pair of coefficients (for example, if the input contains values X (3) = [- 5, j300], and X (2K-3) = [75, -j12], then the output will be set to the values X (3) = [75, -j12], and X (2K-3) = [- 5, j300]) and these mirror values of the spectral transform coefficients appear at the code output of BP ₂ . At the same time, pulses with a doubled sampling rate are counted in the counter and, with the appearance of a sampling pulse corresponding to the end of the second segment, a short pulse appears at the output of the counter. Under the action of the leading edge of this pulse arriving at the third input of BP ₂ , 2K pairs of spectral transform coefficients included in the second segment and present at the code output of BP ₁ and the code input of BP ₂ are recorded in BP ₂ . Under the influence of decrease of the same short pulse supplied to the third input of the BP ₁ installs the PD ₁ to the initial state.

Further, the work of the SFZSK proceeds in a similar way. Thus, in the SFZSK, from 2K pairs of spectral transform coefficients included in each segment, 2K pairs of specular spectral transform coefficients are formed in each segment, which corresponds to the formation of specular spectral components in the time domain in the original analog signal.

An example of the implementation of the envelope and phase extraction scheme (SWOF) included in the BFZS 12 is shown in FIG. 8. This circuit contains a first squaring circuit (CBK ₁ ), a second squaring circuit (SVK ₂ ), an adder, a square root extraction circuit (SIKK), a division circuit (SD) and an argtan circuit. The first code input SVOF is connected to the first output of SPSCNON (Fig. 6), and inside the SVOF (Fig. 8) it is connected to the code input CBK ₁ and the second code input CD. The second code input of the SVOF is connected to the second output of the SPSKNON (Fig. 6), and inside the SVOF (Fig. 8) it is connected to the code input of the SVO ₂ and the first code input of the SD. In this case, the code outputs CBK ₁ and SVK _{2 are} connected, respectively, to the first and second code inputs of the adder, the code output of which is connected to the code input of the SIKK, the code output of which is the second output of the SWOF. Moreover, the code output of the SD is connected to the code input of the argtan circuit, the code output of which is the first output of the SVOF.

в виде действительной зеркальной составляющей х_з(t) и мнимой зеркальной составляющей х_1з(t) поступают, соответственно, с первого и второго выходов СПСКНОН на, соответственно, первый и второй входы СВОФ (Фиг. 6). А внутри СВОФ (фиг. 8) сигнал действительной зеркальной составляющей х_з(t) поступает на кодовый вход CBK₁ и на второй кодовый вход СД, а сигнал мнимой зеркальной составляющей х_1з(t) поступает на кодовый вход СВК₂ и на первый кодовый вход СД. При этом в СД осуществляют деление х_1з(t) на х_з(t) после чего результат деления с кодового выхода СД поступает на кодовый вход схемы argtan. На выходе схемы argtan, согласно известной формулы, оказывается сформированным сигнал зеркальной фазы ϕ_з(t)=argtg [х_1з(t)/х_з(t)], который поступает на первый выход СВОФ (первый выход БФЗС 12 на фиг. 6).SVOF works as follows. Specular digital complex information signal

in the form of a real mirror component x _s (t) and an imaginary mirror component x _1s (t) are supplied, respectively, from the first and second outputs of the SPSCNON to the first and second inputs of the SVOF, respectively (Fig. 6). And inside the SVOF (Fig. 8), the signal of the real mirror component x _s (t) is fed to the code input CBK ₁ and to the second code input of the SD, and the signal of the imaginary mirror component x _1s (t) is fed to the code input of the SVK ₂ and to the first code SD input. In the DM is performed division _1h x (t) to x _s (t) and then dividing the result with the coded SD output is supplied to a code input circuit argtan. At the output of the argtan circuit, according to the well-known formula, a signal of the mirror phase ϕ _s (t) = argtg [x _1s (t) / x _s (t)] is generated, which is fed to the first output of the SVOF (the first output of the BFZS 12 in Fig. 6 ).

In this case, over the signal of the real mirror component x _s (t) arriving at the code input CBK ₁ , in this circuit, squaring x ² _s (t) is performed. Similarly, over the signal of the imaginary mirror component x _1s (t) arriving at the code input of the SVK ₂ , square x ² _1s (t) is performed. After that, the signals from the code outputs CBK ₁ and SVK ₂ are fed, respectively, to the first and second code inputs of the adder, in which these signals are summed, as a result of which x ² ₃ (t) + x ² _1z (t) are obtained at its code output ... Then this summed signal from the code output of the adder is fed to the code input of the SIKK, in which the operation of extracting the square root is carried out and as a result, according to the well-known formula, a signal of the mirror amplitude envelope is obtained at its output A _s (t) = [x ² _s (t) + x ² _1z (t)] ^1/2 . This signal of the mirror amplitude envelope A _s (t) is fed to the second output of the SVOF (the second output of the BFZS 12 in Fig. 6).

An example of the implementation of the block for determining the phase increment (BPOF) 16 is shown in FIG. 9. This block contains a delay line for one discrete count (LZODO), a first phase separation circuit (SVF ₁ ), a second phase separation circuit (SVF2) and a phase difference separation circuit (SVRF). The input (code) of the BOPF 16 is connected to the output LZ2 15 (Fig. 1), and inside the BOPF 16 (Fig. 9) its input is connected to the input of the LZODO and the input of the SVF ₁ , the output of which is connected to the first input of the SVRF, the second input of which is connected to the output of the SVFg, the input of which is connected to the output of the LZODO, and the output of the SVRF is the output of the BOPF 16.

с выхода ЛЗ₂ (фиг. 1) в виде параллельных кодовых комбинаций поступает на вход БОПФ 16, а внутри БОПФ 16 (Фиг. 9) этот сигнал поступает на вход ЛЗОДО и вход СВФ₁ В СВФ1 из основного цифрового комплексного сигнала

выделяют фазу этого сигнала, согласно известной формулы ϕ_o(t)=arg tg [x_1o(t)/x_o(t)]. Реализация СВФ1 является такой же как в СВОФ, входящую в БФЗС 12 и показанную на Фиг. 8. Сигнал ϕ_o(t) с выхода СВФ₁ поступает на первый вход СВРФ. Этот же основной цифровой комплексный сигнал

подвергается задержки на один дискретный отсчет в ЛЗОДО

и с выхода этой схемы поступает на вход СВФ2, в которой из этого задержанного сигнала также выделяют фазу ϕ_оз(t)=arg tg [х_1оз(t)/х_оз(t)]. Реализация СВФ₂ является такой же как в СВОФ, входящую в БФЗС 12 и показанную на Фиг. 8. Данный сигнал ϕ_оз(t) с выхода СВФ₂ поступает на второй вход СВРФ. В СВРФ производят выделение разности фаз (приращения фазы) dϕ(t)=ϕ_o(t)-ϕ_оз(t). Этот сигнал приращения фазы dϕ(t) с выхода СВРФ поступает на выход БОПФ 16.BOPF 16 works as follows. Basic digital complex signal

from the output of the LZ ₂ (Fig. 1) in the form of parallel code combinations is fed to the input of the BOPF 16, and inside the BOPF 16 (Fig. 9) this signal is fed to the input of the LZODO and the input of the SVF ₁ In the SVF1 from the main digital complex signal

select the phase of this signal, according to the well-known formula ϕ _o (t) = arg tan [x _1o (t) / x _o (t)]. The implementation of the SVF1 is the same as in the SVOF included in the BFZS 12 and shown in FIG. 8. The signal ϕ _o (t) from the output of the SVF _{1 is} fed to the first input of the SVRF. The same basic digital complex signal

Delayed by one discrete sample in LZODO

and from the output of this circuit is fed to the input of SVF2, in which the phase ϕ _oz (t) = arg tan [x _1oz (t) / x _oz (t)] is also separated from this delayed signal. The implementation of the SVF ₂ is the same as in the SVOF included in the BFZS 12 and shown in FIG. 8. This signal ϕ _oz (t) from the output of SVF _{2 is} fed to the second input of the SVRF. In SVRF produce separation of the phase difference (phase increment) dϕ (t) = ϕ _o (t) -ϕ _oz (t). This signal of the phase increment dϕ (t) from the output of the SVRF is fed to the output of the BOPF 16.

An example of an implementation of a mirror signal recovery unit (MSS) 18 is shown in FIG. 10. This block contains a circuit for generating a phase cosine signal (SFSCF) and a multiplication circuit (SD). The first input of the BVZS 18 is connected to the output of the BVZS 17, and the second input of the BVZS 18 is connected to the second output of the BFZS 12 (Fig. 1). And inside the BVZS 18 (Fig. 10), its first input is connected to the input of the BVZS, the output of which is connected to the first input of the control system, the second input of which is connected to the second input of the BVZS 18, and the output of the control system is connected to the output of the BVZS 18.

Этот сигнал внутри БВЗС 18 (фиг. 10) поступает на вход СФСКФ, в которой осуществляют формирование корректированного зеркального сигнала косинуса фазы cos ϕ_зк(t)=cos [ϕ_з(t)+dϕ(t)]. После этого данный сигнал с выхода СФСКФ поступает на первый вход СУ, на второй вход которой поступает сигнал зеркальной амплитудной огибающей A₃(t) со второго входа БВЗС 18 (со второго выхода БФЗС 12 на фиг. 1). В СУ после перемножения этих двух сигналов получают корректированный зеркальный цифровой информационный сигнал

Этот сигнал с выхода СУ поступает на выход БВЗС 18.BVZS 18 works as follows. A signal with a corrected mirror phase is fed to the first input of the BVZS 18 from the output of the BSP 17 (Fig. 1)

This signal inside the BVZS 18 (Fig. 10) is fed to the input of the SFSKF, in which the formation of the corrected mirror signal of the phase cosine cos ϕ _zk (t) = cos [ϕ _z (t) + dϕ (t)] is carried out. After that, this signal from the output of the SFSKF is fed to the first input of the SU, the second input of which receives the signal of the mirror amplitude envelope A ₃ (t) from the second input of the BVZS 18 (from the second output of the BFZS 12 in Fig. 1). In the SU, after multiplying these two signals, a corrected mirror digital information signal is obtained

This signal from the SU output is fed to the BVZS 18 output.

The use of the proposed method and device for its implementation makes it possible to: increase the accuracy of the digital method for measuring the spectrum of information acoustic signals by compensating in the spectrum of the side lobes of the window transformation and thereby increasing the resolution and decreasing the oscillation of the estimation of the amplitude of the spectral components. The method provides a reduction in the duration of the segments of the information acoustic signal, on which the spectrum is measured by using a discrete cosine transformation and a mirror signal to compensate for distortion and interference arising from this discrete cosine transformation of the information signal.

The proposed method and device make it possible to better control and regulate the spectral characteristics of signals and television and radio broadcasting equipment, which makes it possible to generate acoustic signals with a higher level of quality. This, in turn, contributes to an increase in the rating of TV and radio broadcasting stations and programs. The fact is that the spectral characteristics of information acoustic signals largely determine the emotional saturation of broadcasting programs, which is closely related to the popularity of these programs and broadcasting stations. This not only contributes to an increase in the number of listeners of these broadcasting stations and an increase in the influence on these listeners, but, as a result, brings higher incomes to these stations. Thus, the use of the proposed method and device makes it possible to improve the quality of sound signals and programs, better control and regulate broadcasting equipment, increase the popularity ratings of broadcasting stations, and also increase the economic efficiency of these stations.

Claims (2)

1. A method for measuring the spectrum of information acoustic signals with distortion compensation, including low-frequency filtering, and then analog-to-digital conversion with the formation of a digital information signal x (t), which is further represented as a digital complex signal

in which the real component is represented in the form of a digital information signal x (t), and the imaginary component jx ₁ (t) is equal to zero, after which a sequence of segments is formed from the digital complex signal from K code combinations of this signal in each segment, which are converted into a sequence segments of a digital complex signal from 2K code combinations in each segment, after which the Nuttall window function is superimposed on each segment and a sequence of digital complex signal segments is obtained from 2K code combinations of this signal in each segment, then a direct fast Fourier transform of 2K code combinations of a digital complex signal and form 2K pairs of transformation coefficients corresponding to the representation of each segment of the digital complex signal in the spectral domain, after which, in each segment, the phase of 2K pairs of transformation coefficients is rotated in digital form, which corresponds to a 90 ° phase rotation of all spectral components in the time domain in the original analog signal, and then the inverse fast Fourier transform is performed from 2K pairs of spectral coefficients in each segment in 2K code combinations in each segment of the digital complex signal, after which the addition with 50 % overlap of each segment of the digital complex signal with the previous segment and compensation for the unevenness of the Nuttall window function, and thus an orthogonal digital information signal x ₁ (t) is obtained, shifted by 90 ° with respect to the digital information signal x (t), and also including the first discrete-cosine transform, the second discrete-cosine transform and digital indication, characterized in that after the analog-to-digital conversion, the digital information signal x (t) is delayed and then connected to the orthogonal digital information signal x ₁ (t), and the main digital complex whitefish cash

consisting of the main digital information signal x _o (t) and the imaginary part of this main digital information signal jx _1o (t), then from the main digital complex signal

form a mirror digital complex information signal

from which, in turn, the signal of the mirror amplitude envelope A _s (t) and the signal of the mirror phase are separated, in addition, the main digital complex signal

delay and from its ϕ _z (t) component in the form of the main digital information signal x _o (t) perform the first discrete-cosine transformation and obtain the first B ₀ spectral DCT coefficients containing interference in the form of side lobes, as well as from the delayed main digital complex signal

the signal of the phase increment dϕ is selected in one discrete sample, after which this signal is added with the signal of the mirror phase ϕ _s (t) and thus a signal with the corrected mirror phase ϕ _kz (t) = ϕ _s (t) + dϕ is generated, after which formation of a corrected mirror digital information signal

over which the second discrete-cosine transformation is carried out and the second spectral DCT coefficients B _{sk of the} corrected mirror digital information signal x _kz (t) are formed, which also contain interference in the form of transformation side lobes, but having an opposite phase with respect to the interference contained in the first spectral DCT coefficients В _o , then the first spectral DCT coefficients of the main digital information signal В _{o are} added with the second spectral DCT coefficients of the corrected mirror digital information signal В _s , as a result of which interference is compensated in the form of transformation side lobes, after which over the spectral DCT coefficients formed in this way of the main digital information signal with compensated interference, then digital indication is carried out. 2. A device for implementing a method for measuring the spectrum of information acoustic signals with distortion compensation, containing a low-pass filter, the input of which is the input of the device, and the output is connected to the input of an analog-to-digital converter, as well as a block for doubling the frequency of sampling pulses, a block for segmentation and overlay of a window function Nuttall, series-connected block of fast Fourier transform, block of phase rotation of conversion coefficients, block of inverse fast Fourier transform, block of overlapping segments and compensation of unevenness of the Nuttall window, and also containing the first block of discrete-cosine transform, second block of discrete-cosine transform, display unit with display, characterized in that the first delay line, the second delay line, the complex signal generation unit, the mirror signal generation unit, the phase increment determination unit, the phase summation unit, the mirror image recovery unit are additionally introduced. o signal and adder, while the first output of the analog-to-digital converter is connected to the input of the Nuttall window function segmentation and superposition unit and to the input of the first delay line, the output of which is connected to the first input of the complex signal generating unit, and the second output of the analog-to-digital converter is connected to the input of the block for doubling the frequency of sampling pulses, and the output of the block of segmentation and superposition of the Nuttall window function is connected to the input of the fast Fourier transform block, and the output of the block of overlapping segments and compensation of unevenness of the Nuttall window is connected to the second input of the complex signal shaping block, the output of which is connected to the input of the shaping block of the mirror signal and to the input of the second delay line, the output of which is connected to the input of the phase increment determination unit and to the input of the first discrete-cosine transformation unit, the output of which is connected to the first input of the adder, and the first output of the mirror signal generation unit a is connected to the first input of the phase summation unit, the second input of which is connected to the output of the phase increment determination unit, and the output of the phase summation unit is connected to the first input of the mirror signal recovery unit, the second input of which is connected to the second output of the mirror signal formation unit, and the output of the recovery unit the mirror signal is connected to the input of the second block of discrete-cosine transformation, the output of which is connected to the second input of the adder, the output of which is connected to the input of the display unit with the display.

Images