RU2691122C1 - Method and apparatus for companding audio broadcast signals - Google Patents

Abstract

FIELD: electrical communication engineering.SUBSTANCE: invention relates to telecommunication and can be used to transmit audio broadcast signals in analogue and digital communication channels. To achieve the technical result, the method involves companding not the broadband audio broadcast signal but the Hilbert amplitude envelope selected from it. To increase signal-to-noise ratio not only in pause, but on background of signal, this amplitude envelope is divided into low-frequency, medium-frequency and high-frequency components. Further, these three components are treated in the form of compression at low and medium frequencies and expansion at high frequencies. These three components of the Hilbert amplitude envelope after summation are used to reconstruct the audio broadcast signal by multiplying the processed Hilbert amplitude envelope by the phase cosine signal extracted from the audio signal together with the Hilbert amplitude envelope.EFFECT: high quality of transmitting audio broadcast signals.2 cl, 8 dwg

Description

Technical field

The invention relates to telecommunications and can be used to transmit audio broadcast signals in analog and digital communication channels.

The level of technology

There is a method of compaction implemented in the device (AS No. SU 1665518 A1 BI No. 27 dated 07.23.1991), including on the transmitting side the summation of the sound signal, the compression of this signal, low-frequency filtering, summation with a lower frequency-modulated control signal, modulation and transfer of spectra to high frequencies with a single sideband, limiting the dynamic range of the control signal, high-frequency signal compression and high-frequency filtering. And on the receiving side - the expansion of the compressed high-frequency signal, high-frequency filtering and demodulation of this signal, the summation of the compressed audio signal and its expansion, band-pass filtering and demodulation of the frequency-modulating control signal, low-frequency filtering with the release of the restored audio signal.

A disadvantage of the known method and device is that sound signals can be transmitted only in high-frequency communication channels, and the modulation and demodulation processes of signals introduce additional distortion and interference to these signals. In addition, there is an insufficient quality of companding, expressed in distortions of the sound signal, in modulation of the variable transfer coefficient of the high-frequency components of the signal and noise, as well as in reducing the visibility of noise only in the pause.

The closest method of the same assignment to the claimed one is the method implemented in the device (AS No. SU 1030975 A BI No. 27 dated 07.23.1983), which includes on the transmitting side a frequency correction of the analog broadcast audio signal, amplitude limitation of this signal, analog-to-digital conversion of the signal with the formation of a digital broadcast signal transmission and amplitude compression of this digital signal, and on the receiving side the expansion of the digital broadcast signal reception, digital-analog conversion Calling with the formation of the reconstructed analog audio broadcast signal reception, the inverse frequency correction and receiving the output analog audio broadcast signal reception.

A device for digital companding of audio signals (A.S. No. SU 1030975 A BI No. 27 dated 07.23.1983) is known, containing on the transmitting side serially connected source of audio signals, a correction unit, the first limiter, the first analog-to-digital converter, and also the first. the compressor, and on the receiving side - the first expander and the first digital-to-analog converter and the reverse correction unit connected in series.

A feature of the known method and device is that it uses the generation of a compression control signal, which depends both on the frequency composition of the encoded audio signal and on the probability of its subband being divided into which the entire dynamic range of the encoded signal is divided.

A disadvantage of the known method and device is insufficient companding quality, expressed in distortions of the sound signal form, in modulation by the variable transmission coefficient of the high-frequency components of the signal and noise, and also in reducing the noticeable noise only during a pause.

Summary of Invention

The task of the invention is to improve the quality of transmission of audio broadcast signals.

The problem is solved by using on the transmitting side the Hilbert amplitude envelope of the transmission, extracted from the broadcast sound signal, from which the transmission phase cosine signal is also extracted. From the Hilbert amplitude envelope of the transmission, low-frequency, mid-frequency and high-frequency components are distinguished by filtering. Then the low frequency components of the transmission envelope. they compress with a high compression ratio, mid-frequency components compress with a lower compression ratio, and high-frequency components expand. After that, all three components are summed up and receive the processed Hilbert amplitude transmission envelope, which is used to recover after processing the broadcast sound signal by multiplying this processed Hilbert transmission envelope by the cosine signal of the transmission phase. And on the receiving side, the Hilbert amplitude envelope of the reception and the cosine signal of the reception phase are extracted from the sound broadcasting reception signal. Low-frequency, mid-frequency and high-frequency components are distinguished from the Hilbert amplitude envelope of the reception. Then, the low-frequency components of the envelope of the prima are expanded with a large coefficient of expansion, the mid-frequency components are expanded with a lower coefficient of expansion, and the high-frequency components are compressed. After that, all three components are summed up and receive the processed Hilbert amplitude reception envelope, which is used to recover after processing the audio broadcast signal of the reception by multiplying this processed Hilbert reception envelope by the cosine signal of the reception phase.

The proposed method of companding audio broadcast signals, which includes on the transmitting side a frequency correction of an analog audio broadcast signal, amplitude limitation of this signal, analog-digital conversion of a signal with the formation of a digital broadcast signal of transmission in such a manner, the formation of a Gilbert conjugate orthogonal transmission signal and obtaining in this way complex transmission signal, from which the cosine signal of the transmission phase and the Hilbert signal are separated the bulk envelope of the transmission, from which “emit low-frequency, mid-frequency and high-frequency components of the transmission by filtering, and then the low-frequency components of the envelope are compressed with a high compression ratio, the mid-frequency components of the envelope compresses with a lower compression ratio, and the high-frequency components of the envelope expose, after which all three Hilbert components the transmission envelope are summed up and get the processed Hilbert amplitude transmission envelope, which is transmitting the recovered digital broadcast signal from which the output analog audio broadcast signal is generated by digital-analog conversion, and the analog audio broadcast signal receiving amplitude limiting on the receiving side with the formation in this way of a digital broadcast signal of reception, the formation of the orthogonal reception signal associated with it according to Hilbert and Thus, the complex reception signal from which the cosine signal of the reception phase and the signal of the Hilbert amplitude reception envelope are extracted, from which the low-frequency, mid-frequency and high-frequency reception components are extracted by filtering, and then the low-frequency components of the envelope are expanded with a large expansion coefficient, the mid-frequency envelope components are expanded by lower coefficient of expansion, and the high-frequency components of the envelope are compressed, after which all three components of the gil receive berth envelope are summed up and receive the processed Hilbert amplitude reception envelope, which is multiplied by the cosine signal of the reception phase and receive a digital broadcast reception signal, restored after processing, from which a reconstructed analog audio broadcast signal is generated, which is inversely frequency-corrected by digital-analog conversion , and receive the output analog audio broadcast signal reception.

And in the companding device of audio broadcasting signals containing on the transmitting side serially connected source of audio signals, a correction unit, the first limiter, the first analog-to-digital converter, and also the first compressor, and on the receiving side the first expander and the serially connected first digital-analog converter and an inverse correction unit, the first orthogonal signal shaping unit, the first modulation decomposition unit, ne the first low pass filter, the first bandpass filter, the first high pass filter, the second compressor, the second expander, the first adder, the first modulation signal recovery unit and the second digital-to-analog converter, while the output of the first analog-digital converter is connected to the input of the first orthogonal shaping unit the signal, the first and second outputs of which are connected, respectively, with the first and second inputs of the first block of the modulation decomposition of the signal, the first output of which is connected to the first input p The first unit of the modulation signal recovery, and its second output is connected to the input of the first low-pass filter, the input of the first band-pass filter and the input of the first high-pass filter, and the output of the first low-pass filter is connected to the input of the first compressor, the output of which is connected to the first input of the first adder, the output of the first bandpass filter is connected to the input of the second compressor, the output of which is connected to the second input of the first adder, and the output of the first high-pass filter is connected to the input of the second eq pander, the output of which is connected to the third input of the first adder, the output of which is connected to the second input of the first modulation signal recovery unit, the output of which is connected to the input of the second D / A converter, the output of which is the output of the transmitting side of the device, and the second limiter is additionally inputted at the receiving side , the second analog-to-digital converter, the second block forming the orthogonal signal, the second block of the modulation decomposition of the signal, the second low-pass filter, in The second bandpass filter, the second high-pass filter, the third expander, the third compressor, the second adder and the second modulation signal recovery unit, the input of the second limiter is the input side of the device, and its output is connected to the input of the second analog-to-digital converter, the output of which is connected with the input of the second block forming the orthogonal signal, the first and second outputs of which are connected, respectively, with the first and second inputs of the second block of the modulation decomposition of the signal, the first the output of which is connected to the first input of the second block j of the modulation signal recovery, and its second output is connected to the input of the second low-pass filter, the input of the second band-pass filter and the input of the second high-pass filter, and the output of the second low-pass filter is connected to the input of the first expander, the output of which connected to the first input of the second adder, the output of the second bandpass filter is connected to the input of the third expander, the output of which is connected to the second input of the second adder, and the output of the second filter is high These frequencies are connected to the input of the third compressor, the output of which is connected to the third input of the second adder, the output of which is connected to the second input of the second modulation signal recovery unit, the output of which is connected to the input of the first D / A converter, the output of which is connected to the input of the inverse correction unit, output which is the output side of the device.

Thanks to this solution of the problem, the proposed method and companding device of audio broadcast signals, unlike the prototype, allows to avoid distortion of the sound signal shape (retains the waveform), reduce modulation of the high frequency components of the signal and noise by a variable transfer coefficient, and also reduce the visibility of noise not only during a pause but also in signal. As a result, it is possible to improve the quality of transmission of sound broadcasting signals.

List of figures

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

FIG. 1 Block diagram of the device companding sound broadcasting signals;

FIG. 2 Block diagram of the formation of an orthogonal signal;

FIG. 3 Block diagram of the modulation signal decomposition;

FIG. 4 Block diagram of the modulation signal recovery;

FIG. 5 Segmentation and overlay scheme of Nuttall's window function included in the orthogonal signal generation unit;

FIG. 6 Timing diagrams of the segmentation scheme and the imposition of Nuttall's cloth function, which is part of the orthogonal signal generation unit;

FIG. 7 Scheme of overlapping segments and compensating for the unevenness of the Nuttall window function included in the orthogonal signal generating unit;

FIG. 8 Timing diagrams of the operation of the segment overlapping scheme and compensation for the unevenness of the Nuttall window function included in the orthogonal signal generation unit.

The implementation of the invention

A feature of the proposed method of compiling sound broadcasting signals, unlike the prototype, is the lack of a compression control signal, which depends both on the frequency composition of the encoded audio signal and on the probability of its sub-band falling into which the dynamic range of the encoded signal is broken. The difference from the prototype is also in the absence of distortion of the sound signal, in the absence of modulation of the high frequency components of the signal and noise by the variable transfer coefficient, as well as in reducing the noticeable noise not only in the pause, but also against the background of the signal itself.

The basis of the proposed method is companding not the wideband audio broadcasting signal, but the Hilbert amplitude envelope extracted from the negro. To increase the signal-to-noise ratio not only in the pause, but also against the background of the signal, this amplitude I envelope is divided into low-frequency, mid-frequency and high-frequency components. Further, these three components are processed in the form of compression in the region of low and medium frequencies and expander in the region of high frequencies. These three components of the Hilbert amplitude envelope after summation are used to reconstruct the audio broadcast signal by multiplying the processed Hilbert amplitude transmission envelope by the phase cosine signal extracted from the audio signal along with the Hilbert amplitude envelope. And on the receiving side, the Hilbert amplitude reception envelope, extracted from the broadcast sound reception signal, is subdivided into low-frequency, mid-frequency and high-frequency components. Further, these three components are processed in the form of expansion in the region of low and medium frequencies and compression in the region of high frequencies. These three components of the Hilbert amplitude reception envelope after summation are used to reconstruct the audio broadcast signal by multiplying the received Hilbert amplitude reception envelope by the phase cosine signal extracted from the reception audio sound signal together with the Hilbert amplitude envelope.

The companding method of sound broadcasting is implemented as follows. On the transmitting side, the original analog audio broadcast signal from the source of audio signals is subjected to frequency correction (Fig. 1) with a slight increase in the high-frequency components of the signal, which makes it possible to slightly increase the amplitude of the most low-level components of this signal. After that, an amplitude limit is applied to the signal before its analog-to-digital conversion. Amplitude limiting allows you to match the amplitude of the audio broadcast signal with the dynamic range of the analog-to-digital converter. Analog-to-digital conversion allows you to generate a digital broadcast signal transmission from the analog audio broadcast signal transmission. Next, from the digital broadcast signal transmission carry out the formation of the orthogonal transmission signal associated with it according to Hilbert, according to (Broadcasting and Electroacoustics. Edited by Kovalgina Yu.A. Radio and Telecommunications, 1999, p. 75):

where S ₁ (t) is the Hilbert conjugate signal from the original digital broadcast signal S (t).

The Hilbert conjugate signal is exactly the same as the original signal, but having a phase rotation of all its spectral components through 90 °.

Next, from the thus obtained complex signal consisting of S (t) and S ₁ (t), a pair of parametric signals is extracted that contains the Hilbert amplitude envelope signal A (t) of the transmission and the cosine signal of the cos ϕ (t) phase, according to (Broadcasting and Electroacoustics. Edited by Kovalgina, Yu.A.M. Radio and Communication, 1999, p. 75):

Thus, to highlight the Hilbert amplitude envelope A (t), according to [2], it is necessary to square each of the signals S (t) and S ₁ (t), then fold and get:

Further, according to [2], the received signal [4] is required to undergo the operation of extracting the square root (Fig. 3).

The cosine phase signal is obtained by dividing the original digital broadcast signal S (t) by the Hilbert amplitude envelope A (t), according to [3].

The Hilbert transform allows us to represent the signal as a product of two functions — the envelope A (t) and cosine of the cos ϕ (t) phase:

Next, from the thus obtained signal of the Hilbert amplitude envelope of the transmission, the low-frequency, mid-frequency and high-frequency components of the transmission are filtered out. Then, the low-frequency components of the envelope are compressed with a high compression ratio, the mid-frequency components are compressed with a lower compression ratio, and the high-frequency components are expanded. Such separate processing avoids modulation with a variable compression ratio, determined mainly by powerful low-frequency components and less powerful mid-frequency components, with respect to low-level high-frequency components and noise in the Hilbert amplitude envelope. Expanding the high-frequency components of the envelope allows you to further increase the amplitude of the relatively high-level components of this signal and reduce the amplitude of the low-level components of this signal and noise. If the frequency correction allowed a slight rise in both the low-level and relatively high-level components of the high-frequency components, which improved the quality of the analog-digital conversion, the expansion makes it possible to separate the high-level components perceived by the ear from the low-level components of the high-frequency part of the envelope that are not perceived by the ear and the noise.

After that, all three processed components of the envelope are summed up and the processed Hilbert amplitude envelope of the transmission A _o (t) is obtained. This processed envelope is then multiplied by the cosine phase cos ϕ (t) signal and the reconstructed digital broadcast signal S _b (t) is obtained.

The reconstructed digital broadcast signal is then subjected to digital-to-analog conversion and an output analog audio broadcast signal is generated, which can be transmitted to the receiving side either in analog form or digital (based on additional analog-digital conversion).

And on the receiving side, the amplitude limit of the analog sound broadcasting signal is performed, which allows the amplitude of the analog sound signal of the reception to be matched to the dynamic range of the analog-to-digital converter. After this, an analog-to-digital conversion is carried out, which makes it possible to form a digital broadcasting reception signal from the analogue audio broadcasting reception signal. Further, from the digital broadcast signal of reception, the formation of the orthogonal reception signal associated with it according to Hilbert is carried out, just as it was done on the transmitting side.

The Hilbert conjugate orthogonal reception signal is exactly the same as the analogue sound reception signal, but having a phase rotation of all its spectral components through 90 °. Next, a pair of parametric signals is obtained from the complex reception signal obtained in this way, which contains the signal of the Hilbert amplitude reception envelope A (t) and the cosine signal of the reception phase cos ϕ (t), according to formulas [2] and [3].

Then, from the received signal of the Hilbert amplitude reception envelope, the low-frequency, mid-frequency, and high-frequency components of the reception are filtered out. After that, the low-frequency components of the reception envelope are expanded with a large coefficient of expansion, the mid-frequency components are expanded with a lower coefficient of expansion, a. high-frequency components are compressed. Such separate processing, which is of the opposite nature with respect to the processing of low-frequency, mid-frequency and high-frequency components of the Hilbert amplitude envelope on the transmitting side, allows one to accurately reconstruct these three components of the Hilbert amplitude envelope at the reception.

After that, all three components of the reception envelope are summed up and the processed Hilbert amplitude reception envelope A _o (t) is obtained. This processed amplitude reception envelope is then multiplied by the cosine signal of the prim phase and a reconstructed digital broadcast reception signal S _in (t) is obtained.

Next, a digital-to-analog conversion is performed over the reconstructed digital broadcast reception signal and a reconstructed analog audio reception signal is generated, which is inverse frequency corrected with a slight lowering of the high-frequency signal components and thus an output analog audio broadcast signal is received, corresponding to the original analog audio signal broadcast broadcast signal.

The method is carried out using a device that contains on the transmitting side (Fig. 1) a series-connected audio signal source (FMC) 1, a correction unit (CC) 2, a first limiter 3, a first analog-to-digital converter (ADC ₁ ) 4, and the first compressor 5, and on the receiving side - the first expander 6 and the first digital-to-analog converter (DAC ₁ ) 7 and the inverse correction unit (SCC) 8 in series 8. The first orthogonal signal generating unit (BFOS ₁ ) is additionally inserted in the device on the transmitting side 9, the first block of the modulation decomposition of the signal (BMRS ₁ ) 10, the first low-pass filter (LPF ₁ ) 11, the first band-pass filter (PF ₁ ) 12, the first high-pass filter (HPF ₁ ) 13, the second compressor 14, the second expander 15, the first adder 16, the first unit of the modulation signal recovery (BBCS ₁ ) 17 and the second digital-to-analog converter (DAC ₂ ) 18. In this case, the output of the first ADC ₁ 4 is connected to the input of the first BFOS ₁ 9, the first and second outputs of which are connected respectively to the first and second inputs of the first BMRS ₁ 10. ₁ BMRS first output 10 is connected to a first input m BVMS first ₁ 17 and second ₁ BMRS output 10 is connected to the input of the first LPF ₁ 11 input of the first PD ₁ 12 and _one input of the first HPF 13. Moreover, the output of the first LPF 11 is connected to _one input of the first compressor 5, the output of which is connected to a first input the first adder 16. The output of the first PF ₁ 12 is connected to the input of the second compressor 14, the output of which is connected to the second input of the first adder 16. The output of the first high-pass filter ₁ 13 is connected to the input of the second expander 15, the output of which is connected to the third input of the first adder 16. In its the queue, the output of the first adder 16 is connected with the second inputs of the first BMBC ₁ 17, the output of which is connected to the input of the second D / A converter ₂ 18, the output of which is the output of the transmitting side of the device.

And on the receiving side, the second limiter 19, the second analog-to-digital converter (ADC ₂ ) 20, the second block forming the orthogonal signal (BFOS ₂ ) 21, the second block of the modulation signal decomposition (BMPS ₂ ) 22, the second filter are additionally introduced (Fig. 1) low frequencies (LPF ₂ ) 23, the second band-pass filter (PF ₂ ) 24, the second high pass filter (HPF2) 25, the third expander 26, the third compressor 27, the second adder 28 and the second unit of modulation signal recovery (BBCS ₂ ) 29. When This input of the second limiter 19 is the input side of the receiver and its output is connected to the input of the second ADC ₂ 20, the output of which is connected to the input of the second BFOS ₂ 21. In turn, the first and second outputs of the second BFOS ₂ 21 are connected, respectively, to the first and second inputs of the second BPCS ₂ 22, first the output of which is connected to the first input of the second BBCS ₂ 29, and the second output of the second BMRS ₂ 22 is connected to the input of the second LPF ₂ 23, the input of the second IF ₂ 24 and the input of the second HPF ₂ 25. At the same time, the output of the second LPF ₂ 23 is connected to the input of the first expander 6, the output of which is connected to the first input of the second adder 28. The output of the second F _{February 24} connected to the third input of the expander 26, whose output is connected Se second input of the second adder 28. The output of the second HPF _{on February 25} connected to the input of the third compressor 27, whose output is connected to a third input of the second adder 28. j In turn, the output of the adder 28 connected to the second input of the second BBCS ₂ 29, the output of which is connected to the input of the first DAC ₁ 7, the output of which is connected to the input of the BSK 8, the output of which is the output of the receiving side of the device.

The proposed method is carried out using the proposed device as follows (Fig. 1). On the transmitting side, the original analog audio broadcast signal from the source of audio signals 1 is fed to the input of BC 2, in which it is subjected to frequency correction. After that, the signal from the output of the BC 2 is fed to the input of the first limiter 3, in which the amplitude limit is applied to the signal. Next, the signal from the output of the first limiter 3 is fed to the input of the first ADC ₁ 4, which generates a digital broadcast transmission signal from the analog audio broadcast transmission signal. Then the signal from the output of the first ADC ₁ 4 is fed to the input of the first BFOS ₁ 9, in which a digital orthogonal transmission signal is formed from a digital broadcast signal of transmission. After that, the digital broadcast transmission signal and the digital orthogonal transmission signal, respectively, from the first and second outputs of the first BFOS ₁ 9, are received, respectively, to the first and second inputs of the first BMRS ₁ 10, in which from the complex signal thus obtained, consisting of S ( t), arriving at the first input of the first BMRS ₁ 10 and signal S ₁ (t), arriving at the second input of the first BMPCi 10, select a pair of parametric signals containing the cosine signal cos ϕ (t) of the transmission, arriving at the first output of the first BMRS ₁ 10 and signal: gi bertovskoy amplitude envelope A (t) of transmission input to the second output of the first ₁ BMRS 10. Further cosine phase signal cos φ (t) of transmission from the first output of the first BMRS ₁ 10 is fed to a first input of the first BMWs ₁ 17 and signal Hilbert envelope amplitude A (t) transmissions from the second output of the first BPCS ₁ 10 are fed to the parallel-connected inputs of the first LPF1 11, first PF1 12, and first HPF ₁ 13, in which the low-frequency, mid-frequency and high-frequency components are allocated by filtering, respectively, the Hilbert amplitude envelope th transfer. Then the low-frequency components of the transmission envelope from the output of the first low-pass filter ₁ 11 are fed to the input of the first compressor 5, in which they are compressed with a large compression ratio. Mid-frequency components of the transmission envelope from the output of the first PF ₁ 12 are fed to the input of the second compressor 14, in which they are compressed with a lower compression ratio. The high-frequency components of the envelope of the transmission from the output of the first high-pass filter ₁ 13 are fed to the input of the second expander 15, in which they are expanded. After that, the signals from the outputs of the first LPF ₁ 11, the first PF ₁ 12, and the first HPF ₁ 13, are fed, respectively, to the first, second and third inputs of the first adder 16, in which they are summed up and receive the processed Hilbert amplitude envelope A _o (t a) transfer. Further, this signal of the processed Hilbert envelope of the transmission from the output of the first adder 16 is fed to the second input of the first BBCS ₁ 17, in which it is multiplied by the cosine signal of the transfer phase cos ϕ (t) fed to the first input of the first BBCS ₁ 17 and is obtained at the output of the first BBCS ₁ 17 restored after processing digital broadcast signal S _in (t). Then the reconstructed digital broadcast signal from the output of the first BBCS ₁ 17 is fed to the input of the DAC ₂ 18, the output of which forms the output analog audio broadcast signal, which is the output signal of the transmitting side).

And on the receiving side (Fig. 1) the input analog audio broadcast signal is received at the input of the second limiter 19, in which the amplitude limit is applied to the signal. Next, the signal from the output of the second limiter 19 is fed to the input of the second ADC ₂ 20, in which the formation of a digital broadcasting signal from the analogue audio broadcasting signal is produced. Then the signal from the output of the second A / D converter ₂ 20 is fed to the input of the second BFOS ₂ 21, in which a digital reception signal is generated (Gilbert's harnessed orthogonal reception signal. After that, the signals from the first and second outputs of the second BFOS ₂ 21 are received, respectively on the first and second inputs of the second BMRS ₂ 22, in which of the complex signal thus obtained, consisting of S (t), arriving at the first input of the second BMRS ₂ 22 and signal S ₁ (t), arriving at the second input of the second BMRS ₂ 22 allocate a pair of 'raram signal signals containing the cosine signal cos ϕ (t) of the reception, which arrives at the first output of the second BMRS ₂ 22 and the signal of the Hilbert amplitude envelope A (t) of the signal, which arrives at the second output of the second BMRS ₂ 22. Further, the cosine signal cos (p ( t) receiving from the first output of the second BMRS ₂ 22 is fed to the first input of the second BMBC ₂ 29, and the signal of the Hilbert amplitude envelope A (t) of the reception from the second output of the second BMRS ₂ 22 is fed to parallel-connected inputs of the second LPF ₂ 2% of the second IF ₂ 24 and a second HPF on February _25, which is isolated by filtration, with responsible, low, medium and high frequency components of the Hilbert envelope reception. Then low-frequency components. The receiving envelope from the output of the second low pass filter ₂ 23 is fed to the input of the first expander 6, in which they are expanded with a large expansion coefficient. Mid-frequency components of the reception envelope from the output of the second PF ₂ 24 are fed to the input of the third expander 26, in which they are expanded with a lower coefficient of expansion. High-frequency components of the envelope of the reception output from the second HPF _{on February 25} are input to the third compressor 27, which compresses them. After that, the signals from the outputs of the second LPF ₂ 23, the second PF ₂ 24, and the second HPF ₂ 25, are respectively fed to the first, second and third inputs of the second adder 28, in which they are summed up and receive the processed Hilbert amplitude receiving envelope A _o ( t). Further, this signal is processed by the Hilbert envelope of the reception from the output of the second adder 28 is fed to the second input of the second BAMC ₂ 29, in which it is multiplied by the cosine signal of the receive phase cos ϕ (t) fed to the first input of the second BBCS ₂ 29 and is received by. the output of the second BBCS ₂ 29 recovered after processing the digital broadcast signal reception S _in (t). Then the reconstructed digital broadcast signal-reception from the output of the second BBCS ₂ 29 is fed to the input of the first DAC ₁ 7, the output of which forms an analogue sound broadcast signal of reception. After that, the analogue sound broadcast signal from the output of the first DAC ₁ 7 is fed to the input of the inverse correction unit 8, in which it is subjected to reverse frequency correction with a slight lowering of the high-frequency components of the signal. The output of the inverse correction unit 8 receives the analog audio output signal of the reception, which is the output signal of the receiving side.

The proposed companding device of sound broadcasting signals, unlike the prototype, allows to avoid distortion of the sound signal shape (retains the waveform), reduce modulation of the high frequency components of the signal and noise by the variable transmission coefficient, as well as reduce noise occupancy not only in the pause but also in the signal. As a result, it is possible to improve the quality of transmission of sound broadcasting signals.

A feature of the proposed companding device of sound broadcasting signals is that non-standard ones in it are: an orthogonal signal generation unit (BFOS), a modulation signal decomposition unit (BIRS) and a modulation signal recovery unit (BMLS).

An implementation example of an orthogonal signal generating unit (BFOS) 9, 21 is shown in FIG. 2 This block contains serially connected: Nuttall window function overlay (SSNOFN), Forward Discrete Fourier Transform (SPDPF) scheme, Phase of transform of transform coefficients (SPFCF), Inverse Discrete Fourier Transform (SODPF) scheme, Pattern of overlapping segments and irregularity compensation window function Nuttolla (SPSKNOFN). In addition, BFOS 9, 21 contains a sampling pulse frequency doubling circuit (SUDID) and a delay line. The first (code) input of the CCNOFN is connected to the input (code) of the BFOS 9, 21 and the first (code) input of the delay line, and the code output of the CCNOFN is connected through - connected in series SPDPP, SPFCN, SODPF to the code input of SPSKNOFN, the code output of which is connected to the second (coded) output BFOS 9, 21. The input of the sampling pulses BFOS 9, 21 (not shown in Fig. 1) is connected to the second input of the SSNOFN, the second input of the SPSKNOFN, the second input of the delay line and the input of the SOCHID, the output of which is connected to the third input of the SSNOFN, the third entrance SPSKNOFN, the second entrance SPDPF , the second entrance SPFCP and the second entrance SODPF. Code output: the delay line is connected to the first (code) output BFOS 9, 21.

The operation of the orthogonal signal generation unit (BFOS) 9, 21 is based on the expression for the direct and inverse discrete Fourier transform: (DFT)

where x (n) is a sequence of B time samples, X (k) is a sequence of B frequency samples.

Block BFOS 9, 21, operates as follows (Fig. 2). At the input (code) BFOS 9, 21 receive parallel code combinations from the output of the ADC 4 (Fig. 1). These code combinations inside BFOS 9, 21 are fed to the first (code) input of the delay line and to the first (code) input of SSNOFN, the second and third inputs of which receive, respectively, the sampling frequency pulses and the double sampling frequency pulses from the SCHID input and output . The sampling frequency pulses (CI from the ADC in Fig. 2) are fed to the input of the SCHIDID from the ADC 4 (in Fig. 1, the circuit for the sampling frequency pulses from the ADC 4 to BFOS 9, 21 is not shown). In CCNOFN, the formation of segments consisting of B parallel code combinations in each segment corresponding to B time discrete samples of the audio signal is carried out. Nuttall’s window function is then imposed on each segment. A digital signal in the form of segments from B parallel code combinations in each segment of the c code output of the CCNOFN is fed to the code input of the SPDPF, where the B point direct discrete Fourier transform of these B parallel code combinations in each segment is performed.

The need to superimpose the Nuttall window function was caused by the fact that the discrete Fourier transform (DFT) uses a rectangular window without overlap, which leads to the appearance of discontinuities of the analyzed functions. The resulting side lobes in the spectrum of the transform window, called percolation, will distort the amplitudes of the neighboring spectral components. To reduce the level of distortion and interference, it is necessary to minimize this leakage of side lobe energy into the main components of the signal. Obviously, the lower the side lobes of the window function in the frequency domain, the higher the accuracy of the direct discrete Fourier transform. The smallest level of side lobes, of the existing window functions, is exactly the Nuttall window.

As a result, In a point direct discrete Fourier transform, In code combinations in SPDP, B pairs of coefficients corresponding to the representation of a digital audio signal in the spectral region are formed, according to [8]. Next, the digital signal from the SPDPP code output is fed to the SPCCH code input, where the phase of the conversion coefficients is rotated by changing the coefficient sign at each pair of coefficients at jsin 2πnk / B, which corresponds to a 90 ° phase rotation of all spectral components in the time domain in the original analog sound signal.

Then, a digital signal from the SPCDT code output is fed to the SODPF code input, where a pointwise inverse discrete Fourier transform from B coefficient pairs to B code combinations in each segment is performed, according to [9].

After that, the digital signal from the SODPF code output goes to the SPSKNOFN code input. This scheme is necessary for better signal recovery in the case of the Nuttoll window, for which addition additionally with 50% overlap is performed. To this end, in SPSKNOFN carry out the addition of a 50% overlap of each segment with its previous segment, delayed for a duration equal to half the duration of the segment. Since the Nuttol window does not refer to the number of windows providing a single transmission coefficient when using 50% of overlaps, an additional increase in the accuracy of the reconstructed digital audio signal is performed by compensating for the non-uniformity of the Nuttall window function. This compensation allows you to increase the protection ratio, which characterizes the level of interference and distortion in the signal, up to 92 dB, which is essential for improving the accuracy of the formation of an orthogonal signal and the quality of signal processing in the device as a whole.

The digital signal from the code output SPSKNOFN is fed further to the second (code) output BFOS 9, 21.

Thus, in BFOS 9, 21, the Hilbert orthogonal transformation of the digital signal was carried out, corresponding to the phase rotation of all the spectral components of the analog audio signal by 90 °. However, this digital signal, after passing through CCNOFN, SPDPF, SPFCP, SODPF and SPSKNOFN, acquired a time delay. For normal operation of the modulation signal decomposition block (BPSS) 10, 22, it is necessary that the digital signal received at the (code) input of BFOS 9, 21 have exactly the same time delay as the digital signal at the first (code) output of this block on its second (code) output. For this purpose, a delay line is used in BFOS 9, 21.

A feature of BFOS 9, 21 is that nonstandard in it are CCNOFN and SPSKNOFN, which require additional disclosed. These blocks and time diagrams of their work are shown in FIG. 5 - FIG. eight.

The pulse frequency doubling frequency scheme (SUCHID) included in BFOS 9, 21 can be made in the form of a series-connected: a meander shaper, a differential circuit, a full-wave rectifier and a short pulse shaper.

An implementation example of the modulation signal decomposition block (BSRS) 10, 22 is shown in FIG. 3. BMRS 10, 22 consists of the first and second squaring schemes, an adder, a square root extraction circuit, a division circuit, and a delay line. The first (code) input BMRS 10,22 is connected to the first (code) input of the division circuit and the code input of the first squaring circuit (CBK ₂ ), and the code input of the second squaring circuit (SVK ₂ ) is connected to the second (code) the input BMRS 10, 22. Code outputs CBK ₁ and SVK _{2 are} connected, respectively, with the first and second (code) inputs of the adder. The code output of the adder is connected to the (code) input of the square root extraction circuit (CQMS), the code output of which is connected to the second (code) output of the BMRS 10, 22 and to the second (code) input of the division circuit. The code output of the division circuit is connected to the code input of the delay line, the code output of which is connected to the first (code) output of the BMRS 10, 22.

The operation of BMRS 10, 22 (Fig. 3), with the selection of the Hilbert amplitude envelope is carried out in accordance with the expression A (t) = [s ² (t) + s ₁ ² (t)] ^1/2 . For this, a digital signal is used, from the first (code) output of the BFOS 9, 21 (Fig. 1), corresponding to the analog audio signal s (t) and a digital signal from the second (code) output of the BFOS 9, 21, corresponding to the analog audio signal, but with the spectral components s ₁ (t) shifted by 90 °. In the first and second SVK (Fig. 3) is carried out in digital form squaring the numerical values of each parallel code combination (corresponding to the samples of the instantaneous amplitudes of the analog audio signal). Next, the digital signal in the form of parallel code combinations from the code outputs of the first CBK ₁ and second CBS ₂ are fed to, respectively, the first and second (code) inputs of the adder. In this scheme, in a digital form, the summation of the numerical values of code combinations arriving at 1 code input of the adder with their corresponding code combinations arriving at 2 code input of the adder is carried out. This operation corresponds to the expression s ² (t) + s ₁ ² (t). After that, the digital signal in the form of parallel code combinations from the code output of the adder is fed to the code input of the SCC. In this scheme, in digital form, the operation of extracting the square root from the numerical values of the code combinations obtained after summation is performed. This operation corresponds to the expression [s ² (t) + s ₁ ² (t)] ^1/2 . The digital signal, Corresponding to the allocated Hilbert amplitude envelope of the analog audio signal A (t), from the code output SIKK is fed to the second (code) output of BPSS 10, 22 and to the second (code) input of the division circuit. The division operation corresponds to the expression cos ϕ (t) = s (t) / A (t) = s (t) / [s ² (t) + s ₁ ² (t)] ^1/2 . The digital signal corresponding to the cosine analog audio signal cos ϕ (t) allocated to the cosine signal is output from the code output of the dividing circuit to the code input of the delay line. The delay line is necessary due to the fact that the digital signal from the first and second (code) inputs of BMRS 10, 22 after passing through CBK ₁ and SVK ₂ , the adder and SIKK acquired a time delay. In addition, the digital signal from the second (code) output BMRS 10, 22 (Fig. 1) then passes through the LPF 11, 23, PF 12, 24, HPF 13, 25, through the compressor 5, 14, 27, through the expander 15, 6 , 26, as well as through an adder 16, 28, then for the normal operation of the modulation signal recovery unit (BBCS) 17, 29 (Fig. 1), it is necessary that the digital signal received at the first (code) input of BBCS 17, 29, would have exactly the same time delay as the digital signal at the second (code) input of the BBCS 17, 29. For this purpose, a delay line is used in the BCRS 10, 22 (Fig. 3).

An implementation example of a modulation signal recovery unit (BMA) 17, 29 is shown in FIG. 4. BMVS 17, 29 consists of a multiplication scheme. The first (code) input of the BBCS 17, 29 is connected to the first (code) input of the multiplication circuit, and the second (code) input of the BBCS 17, 29 is connected to the second (code) input of the multiplication circuit, the code output of which is connected to the (code) output of the BBCS 17 29

The operation of the BBCS 17, 29 (Fig. 4), with the formation of the restored after processing the digital broadcast signal, is carried out in accordance with the expression [5] S _in (t) = A _о (t) ⋅cos ϕ (t). For this purpose, a digital signal is used in the form of parallel code combinations, from the first (code) output of BPCS 10, 22 (Fig. 1), corresponding to the cosine signal of the cos ϕ (t) phase, which is fed to the first (code) input of BHBC 17, 29. A inside the internal combustion engine 17, 29 (fig. 4), the digital signal from its first (code) input goes to the first (code) input of the multiplication circuit. In addition, it uses a digital signal in the form of parallel code combinations from the (code) output of the adder 16, 28 (Fig. 1), corresponding to the signal processed by the Hilbert amplitude envelope A _о (t), which is fed to the second (code) input of the BBCS 17, 29 And inside the internal combustion engine 17, 29 (Fig. 4), the digital signal from its second (code) input goes to the second (code) input of the multiplication circuit. After multiplying these two digital signals in the BCMS 17, 29, the digital broadcast signal S _in (t), which is restored after processing, is formed at its [code output, which is fed to the code output of the BCMS 17, 29.

An example of the implementation of the segmentation scheme and the imposition of Nuttall's window function (CSTOFN), which is part of BFOS 9, 21, is shown in FIG. 5. This circuit contains the first and second buffer memories, the multiplication circuit, the counter and the memory circuit. The first (code) input CCNOFN is connected to the first (code) input of the first buffer memory, the code output of which is connected through the second buffer memory to the 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 SSNOFN. The second, the CCNOFN input 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 of the CCNOFN is connected to the third input of the second buffer memory and to the second input of the memory circuit.

The scheme of segmentation and imposition of Nuttall's window function (Fig. 5) works as follows. In the initial state, the first and second buffer memories and the counter are reset. The memory circuit is also in the initial state, and at its code output there is a code combination corresponding to the transmission coefficient of the Natoll window for the first of the B code combinations (discrete samples) of the digital signal in the segment.

The first (code) input CCNOFN from the (code) input BFOS 9, 21 (Fig. 2) receives parallel code combinations, which are fed to the first (code) input of the first buffer memory (Fig. 5). At the same time, sampling frequency pulses, which are fed to the counter input and the second input of the first buffer memory (Fig. 5), are received from the input of the second SSNOFN input from the sampling pulse frequency doubling circuit input to BFOS 9, 21 (Fig. 2). Pulses with doubled sampling frequency, which are fed to the third input of the second buffer memory and the second input of the memory circuit (Fig. 5), are received from the output of the CCNOFN from the output of the sampling pulse frequency doubling circuit included in BFOS 9, 21 (Fig. 2). . In this case, the counter in CCNOFN is designed to count the number of code combinations equal to half the segment length (half-segment), which is then superimposed on the window-function of Natall. For example, from a digital signal having a sampling frequency of 48 kHz, it is necessary to form a sequence of half-segments, each of which must contain B / 2 = 480 discrete samples (code combinations). In addition, each discrete sample is, for example, a 16-bit code combination. Then, for the duration of each half-segment, 480 sixteen-bit code combinations will fit. It is after a given number of pulses of the sampling frequency that a short pulse appears at the output of the counter, indicating the end of this half-segment and the beginning of the next (Fig. 6 a, b). The pulses from the output of the counter 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. j

The first buffer memory in CCNOFN contains B / 2 = 480 code combinations (half-segment), and the second buffer memory consists of two halves and contains B = 960 code combinations (two half-segments with 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 code combinations appear on 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, B = 960 zero code combinations are read from the second buffer memory under the action of pulses with a double sampling rate. These zero 16 bit code combinations are sequentially received at the first code input of the multiplication circuit. At this time, the 16-bit code combinations corresponding to the transmission coefficients of the Natoll window are fed to the second code input of this scheme. After multiplying the code combinations, fed to 1 and 2 code inputs of the multiplication circuit, its output will also contain zero 16-bit code combinations.

Thus, during the period of filling the first buffer memory with code combinations corresponding to the first half-segment (1 ps in Fig. 6a), the output of the multiplication circuit is the formation of the first segment (0 ₁ -0 ₀ seg. Ha Fig. 6c ) from zero code combinations.

After 480 is filled with sixteen-bit code combinations of the first buffer memory, the first short pulse appears on the output of the counter (Fig. 6b) under the action of the leading edge of which the data code combinations from [the first buffer memory are recorded in the first half of the second buffer memory (1 p. In FIG. 6a). Under the action of the same short pulse, 480 null code combinations from the first half of the second buffer memory are shifted and written into the second half of this buffer memory (0 ps in Fig. 6a). Thus, the first segment is formed from the zero and first half-segments (1 segment in Fig. 6a).

Under the action of the decline of the same short pulse, the first buffer memory and the memory circuit are reset to the initial state. In this case, a code combination appears on the code output of the memory circuit, corresponding to the transmission coefficient of the Natoll window for the first of the B = 960 code combinations in the first segment (1 segment in Fig. 6a). It should be noted that the transmission coefficients of the Natoll window (and the corresponding code combinations) for the first half of the segment (for example, 0 ps in 1 segment. In Fig. 6a) are increasing, and for the second half of the segment (for example, 1 p.s. in 1 seg. in Fig. 6a) are decreasing.

Parallel code combinations that continue to arrive at 1 code input of the first buffer memory are written to this memory under the action of pulses with a sampling frequency. At the same time, under the action of pulses with doubled sampling frequency 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 go to the first and second code inputs of the multiplication circuit, respectively. Zero code combinations (from the second half of the second buffer memory) of the zero half-segment of the first segment (1 segment in Fig. 6a) are multiplied first, therefore only zero 16-bit code combinations appear on the code output of the multiplication circuit.

Then, 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. 6a) begin to multiply, so multiplied 16-bit code combinations appear on the code output of the multiplication circuit, corresponding to the original code combinations, but superimposed on these are the natolla window transmission coefficients.

So during the filling of the first buffer memory codewords corresponding to the second half portion of account (1 ps in FIG. 6a) at the output of the multiplication circuit is forming a second segment of the account (1 ₁ -0 ₂ Seg. FIG. 6c) consisting from the second time used by the zero half segment and the first time used by the first half segment (in which the transmission coefficients of the Natoll window are decreasing).

After filling in the next 480 code combinations of the first buffer memory, a second short pulse appears in the output of the counter (FIG. 6b) under the action of the leading edge of which the data code combinations are recorded in the first half of the second buffer memory. Under the action of the same short pulse 480, the previously recorded code combinations from the first half of the second buffer memory are shifted and recorded into the second half of this buffer memory.

Thus, from the first and second half-segments forms the second segment (2 segments in Fig. 6a).

Under the action of the decline of the same short pulse, the first buffer memory and the memory circuit are reset to the initial state. In this case, a code combination appears on the code output of the memory circuit, corresponding to the transmission coefficient of the Natoll window for the first of the B = 960 code combinations in the second segment (2 segments in Fig. 6a).

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 arrive at the first and second code inputs of the multiplication circuit, respectively. Code combinations (from the second half of the second buffer memory) of the first half-segment of the second segment (2 segments in Fig. 6a) are multiplied first. These multiplied code combinations appear at the output of the multiplication circuit. Next, code combinations (from the first half of the second buffer memory) of the second half-segment of the second segment (2 segments in Fig. 6a) begin to multiply. These multiplied code combinations also appear at the output of the multiplication circuit.

So at the output of the multiplication circuit, the third segment (2 ₁ -1 ₂ segments in Fig. 6c) is formed, consisting of the second time of the first half segment used (in which the transfer coefficients of the Nuttola window are increasing) and the first time of the second half segment used ( natoll's window transfer ratios are decreasing).

While reading the 16-bit code combinations from the second buffer memory, the code combinations corresponding to the third half-segment (3 ps in Fig. 6a) are written to the first buffer memory.

After filling with the next 480 code combinations of the first buffer memory, a third short pulse appears in the output of the counter (Fig. 6b) under the action of the leading edge of which the given code combinations are written in the first half of the second buffer memory. Under the action of the same short pulse 480, the previously recorded code combinations from the first half of the second buffer memory are shifted and recorded into the second half of this buffer memory. Thus, a third segment is formed from the second and third half-segments (3 segments in Fig. 6a).

Under the action of the decline of the same short pulse, the first buffer memory and the memory circuit are reset to the initial state. In this case, a code combination appears on the code output of the memory circuit, which corresponds to the transmission coefficient of the Natoll window for the first of the B = 960 code combinations in the third segment (3 segments in Fig. 6a).

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 arrive at the first and second code inputs of the multiplication circuit, respectively. Code combinations (from the second half of the second buffer memory) of the second half-segment of the third segment (3 segments in Fig. 6a) are multiplied first. These multiplied code combinations appear at the output of the multiplication circuit. Next, code combinations (from the first half of the second buffer memory) of the third half-segment of the third segment (3 segments in Fig. 6a) begin to multiply. These multiplied code combinations also appear at the output of the multiplication circuit.

So at the output of the multiplication circuit, the fourth segment (3 ₁ -2 _{2 2} segments in Fig. 6c) is formed, consisting of the second time of the second half segment used (in which the Natoll window transmission coefficients are increasing) and the first time of the third half segment used ( natoll's window transfer ratios are decreasing). Next, the work of CCNOFN occurs in a similar way.

An example of the implementation of the segment overlapping and non-uniformity compensation scheme of Nuttall's window function (SPSKNON), which is part of BFOS 9, 3, is shown in FIG. 7. This scheme contains: first, second, third and fourth buffer memory (PSU), adder, memory circuit (SP), multiplication circuit (SU), counter, trigger, driver, delay element (EZ). The first (code) input of the first buffer memory (BP ₁ ) is connected to the first (code) input SPSKNON, and its code output - 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 BP _{1 is} connected to the output of the EZ delay element, and the third input of BP _{1 is} connected to the second input of the SPSKNON, to which the counter input is also connected, the output of which is connected to the trigger input, the EZ input and to the second input of BP ₂ , the code output of which is connected with the first (code) input BP ₄ . The third input SPSKNON connected to the second input of the memory circuit (SP), the second input of BP ₃ and the second input of BP ₄ . The trigger output is connected to the driver input, the output of which is connected to the first SP input, to the third input of BP ₃ and to the third input of BP ₄ . Code outputs of BP ₃ and BP _{4 are} connected, respectively, to 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 (SU), the second code input of which is connected to the code output of the SP, and the code output of the SU is connected to the output SPSKNON.

SPSKNON (Fig. 7) works as follows. In the initial state, BP ₁ , BP ₂ , BP ₃ , BP ₄ , the counter, and the trigger are reset. The JV is also in the initial state when a code combination is present at its code output, which corresponds to the transmission coefficient to compensate for the unevenness of the Nuttall window function for the first of the B code combinations in the first segment.

Parallel code combinations from the code output of the inverse discrete Fourier transform circuit, which is part of BFOS 9, 21 (Fig. 2), arrive at the first (code) input SPSKNON (Fig. 7) and then to the first (code) input of the BS. At the same time, pulses with doubled sampling frequency, which are further fed to the third input of BP ₁ (Fig. 7), are received from the output of the SPSNKNON second output from the sampling pulse frequency doubling circuit included in BFOS 9, 21 (Fig. 2). Under the action of these pulses, the code combinations received at the input of the BP ₂ are recorded into it and appear on the code output of the BP ₁ . These code combinations are applied to the first (code) inputs of 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 double sampling rate, you need to form a sequence of half-segments, each of which must contain B / 2 = 480 discrete samples (code combinations). In addition, each discrete sample is a 16-bit code combination. Then, for the duration of each half-segment, 480 sixteen-bit code combinations will fit. 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 (Fig. 8 a, b).

BP ₁ , BP ₂ , BP ₃ , BP ₄ in our example, each contain 480 sixteen-bit code combinations (i.e. each half-segment), Code combinations from the code outputs of the adder, SU and SP are also 16-bit.

SPSNKN is designed to form segments of a digital signal from B code combinations in each segment and the addition with 50% overlap of each segment with its previous segment. In order to avoid gaps in the sequence of a digital signal that is formed after the segments overlap, it is necessary that the code combinations in BP ₁ be recorded at twice the sampling rate, and the code combinations from BP ₃ and BP ₄ are read at the sampling frequency. These pulses with a sampling frequency are fed to the third input of the SPSKNON from the input of the frequency doubling circuit of the sampling pulses, which is part of BFOS 9, 21 (Fig. 2).

Simultaneously with the recording of code combinations in BP ₁ , from BP ₃ and BP ₄ , zero code combinations are read under the action of pulses at their second inputs (Fig. 7). These zero 16 bit code combinations arrive at the first and second code inputs of the adder, the output of which will also be zero 16 bit code combinations that are fed to the first code input of the control system. At this time, the 16-bit code combinations corresponding to the transmission coefficients to compensate for the non-uniformity of the Nuttall window function are fed to the second code input of this scheme from the code output of the SP. After multiplying the code combinations applied to 1 and 2 code inputs of the control system, its code output will also contain zero 16-bit code combinations

So during the period when BP _{1 is} filled with code combinations corresponding to the first half-segment (0 ₀ ps in Fig. 8a), a half-segment (0 _n in Fig. 8 g) is formed from the zero code combinations on the SU code output.

After filling in 480 with sixteen-bit zero code combinations of the BS, corresponding to the 0 ₀ semi-segment (Fig. 8a), the first short pulse appears on the output of the counter (Fig. 8b) from which the trigger is triggered, and a short pulse also appears on the output of the former.

Under the action of the leading edge of the pulse from the shaper output, zero code combinations corresponding to the 0 ₀ half-segment, from the output of BP _{1 are} recorded in BP ₃ , and in BP ₄ , zero code combinations are also recorded, which were present in BP ₂ . Thus, from the 0 and 0 ₀ half-segments (Fig. 8a) the first segment is formed (1 segment. In Fig. 8a - below). At the same time, under the action of the same short pulse from the driver output, the AC is reset to its original state, when a code combination appears at its code output, corresponding to the transfer coefficient to compensate for the unevenness of the Nuttoll window function for the first code combination in the segment.

After that, under the effect of the decay of the pulse from the output of the counter, code combinations from the code output of BP ₁ , corresponding to the 0 ₀ semi-segment, are recorded in BP ₂ and appear on its code output. In addition, under the action of a short pulse delayed in an EZ, BP ₁ is zeroed out and starts recording code combinations corresponding to the following 0 ₁ half-segment (Fig. 8a).

Under the action of pulses at the second inputs of BP ₃ and BP ₄ , 16-bit zero code combinations from their code outputs go to, respectively, the first and second code inputs of the adder. Next, zero code combinations from the code output of the adder (0 ₀ ps +0 ps in Fig. 8a) are fed to the first code input of the SU, to the second code input of which code combinations from the SP output arrive. So at the output of the control system, the first segment is formed (0 ₀ +0 segments in Fig. 8c).

While from BP ₃ and BP ₄ , the reading of 16-bit code combinations is slowed by 2 times (compared to the writing speed in BP ₁ ), code combinations corresponding to the 0 ₁ half-segment are written to BP ₁ .

After filling 480 with zero code combinations of BP ₁ , a second short pulse appears at the output of the counter (Fig. 8b) under the action of which a trigger is triggered and at its output a “logical 0” (“log. 0”) appears, from which no shaper appears at the output of the driver There is no signal, and therefore no entries in BP ₃ and BP _{4, of} parallel code combinations from BP ₁ and BP ₂ . At this time, from BP ₃ and BP ₄ , reading, addition and multiplication of zero code combinations, corresponding to the 0 ₀ and 0 half-segments, continues and the 0 ₀ -0 segment is formed (Fig. 8c).

Under the action of a pulse recession from the output of the counter, code combinations from the code output of BP ₁ , corresponding to the 0 ₁ half-segment, are recorded in BP ₂ and appear at its code output. In addition, under the action of a short pulse delayed in an EZ, BP ₁ is zeroed out and starts recording code combinations corresponding to: the following 0 ₂ half-segment (Fig. 8a).

After filling with zero code combinations of BP ₁ (0 ₂ ps on Fig. 8a), a third short pulse appears at the output of the counter (Fig. 8b) under the action of which a trigger is triggered and at its output a “logical 1” (“log. 1 "), From which a second short pulse appears at the output of the shaper (Fig. 8c). 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 _{2 are} recorded in BP ₃ , and in BP ₄ , also zero code combinations corresponding to 'Oi are recorded and which were present in the mainboard. Thus, the second segment is formed from the O2 and Oi half-segments (2 segments in Fig. 8a - below). At the same time, under the action of the same short pulse from the driver output, the AC is reset to its original state, when a code combination appears at its code output, corresponding to the transfer coefficient to compensate for the unevenness of the Nuttoll window function for the first code combination in the segment.

After that, under the effect of the decay of the pulse from the output of the counter, code combinations from the code output of BP ₁ , the corresponding 0 ₂ -semi-segment, are recorded in BP ₂ and appear on its code output. In addition, under the action of a meek impulse delayed in an EZ, BP ₁ is zeroed out and starts recording code combinations corresponding to the following 1 _1- half segment (Fig. 8a).

Under the action of pulses at the second inputs of BP ₃ and BP ₄ , 16-bit zero code combinations from their code outputs go to, respectively, the first and second code inputs of the adder. Next, zero code combinations from the code output of the adder (0 ₂ ps + 0 ₁ ps in Fig. 8a) are fed to the first code input of the control system, to the second code input of which code combinations arrive from the output of the SP. On the code output of the CU, zero 16-bit code combinations appear. So the output of the control unit is the formation of the second segment (0 ₂ +0 ₁ segm. in Fig. 8 g).

While from PD ₃ and PD ₄ reads 16-bit codewords (0 ps ₂ and 0 ps ₁ in Fig. 8a) stored in the BS ₁ codewords corresponding half portion 1 ₁ (1 ₁ ps in Fig. 8a).

After filling with code combinations (1 ₁ ps in Fig. 8a), a fourth short pulse appears in the absence of the counter (Fig. 8b) under the action of which a trigger is triggered and at its output a “log. 0 ”, from which no signal is generated at the output of the driver, and therefore no code combinations from BP ₁ and BP ₂ are written to BP ₃ and BP ₄ .

At this time, BP ₃ and BS ₄ continues reading, compiling and multiplying the zero code combinations corresponding to 0 ₂ and 0 ₁ half-segments and O ₂ -O ₁ segment is formed (Fig. 8 g).

Under the action of a pulse recession from the output of the counter, code combinations from the code output of BP ₁ , the corresponding 1 ₁ half-segment are recorded in BP ₂ and appear at its code output. In addition, under the action of a short pulse delayed in an EZ, BP ₁ is zeroed out and starts recording code combinations corresponding to: the following 1 ₂ -semi-segment (Fig. 8a).

After filling with code combinations BP ₁ (1 ₂ ps and in Fig. 8a), a fifth short pulse appears at the output of the counter (Fig. 8b) under the action of which a trigger is triggered and at its output a “log. 1, from which a third short pulse appears at the shaper output (Fig. 8c). Under the action of this pulse, the code combinations from the code outputs of BP ₁ and BP _{2. are} recorded, respectively, in BP ₃ and BP ₄ . Thus, the third segment is formed from the 1 ₂ and 1 ₁ half-segments (3 segments in Fig. 8a - below). At the same time, under the action of the same short pulse, the memory unit is reset to its original state when a code combination appears on its code output, corresponding to the transfer coefficient to compensate for the unevenness of the Nuttoll window function for the first code combination in the third segment.

After that, under the action of the decay of the pulse from the output of the estimator, code combinations from the code output of BP ₁ , the corresponding 1 _2- half segment, are recorded in BP ₂ and appear on its code output. In addition, under the action of a short pulse delayed in an EZ, BP ₁ is zeroed out and starts recording code combinations corresponding to the following 2 ₁ half-segment (Fig. 8a).

Under the action of pulses with a sampling frequency at the second departures of BP ₃ and BP ₄ , the 16-bit information code combinations from their code outputs arrive at the first and second code inputs of the adder, respectively. When summing, the addition of code combinations included in the 1 ₂ half-segment (in which the transfer coefficients of the Natoll window are increasing) occurs with the same code combinations in the 1 ₁ half-segment (in which the transmission coefficients of the Nutoll window are decreasing), therefore the transfer coefficients the Natoll windows are aligned (become close to 1), although some unevenness remains.

Then, after summation, code combinations from the code output of the adder (1 ₂ ps + 1 ₁ ps in Fig. 8a) are fed to the first code input of the control system, to the second code input of which code combinations are received from the output of the SP. After multiplying the code combinations, the non-uniformity of the Nuttall window function is compensated. If we compare (1 ₁ -0 ₂ ) segment and (2 ₁ -1 ₂ ) segment (at the top of Fig. 8a) at the input of SPSKNON with 3 segment (3 segments, in Fig. 8a, or 1 ₂ +1 ₁ segments, in Fig. 8). 8d) at the output of the adder, it is clear that there is an addition with a 50% overlap of the segment with its previous segment.

The code output of the control unit receives 16-bit code combinations with compensated non-uniformity of the Nuttall window function. Thus, at the output of the CU, the third segment is formed (1 ₂ +1 ₁ segm. In Fig. 8 g).

While from PD ₃ and PD ₄ reads 16-bit codewords (1 ₂ and 1 ₁ ps ps in Fig. 8a) stored in the BS ₁ codewords corresponding 2 _one half portion (2 ₁ ps in Fig. 8a).

After filling with code combinations (2 ₁ pp in Fig. 13a) BP ₁ , a sixth short pulse appears at the output of the counter (Fig. 8b) under the action of which the trigger is triggered and at its output a “log. 0 ”, from which no signal is generated at the output of the driver, and therefore no recording in the BP ₃ and BP _{4 of} parallel code combinations from the code outputs of BP ₁ and BP ₂ does not occur.

At this time, from BP ₃ and BP ₄ , reading, additions, and multiplication of code combinations corresponding to 1 ₂ and 1 ₁ half-segments continue and 1 ₂ -1 ₁ segment is formed (Fig. 8 g).

Under the action of a pulse recession from the output of the counter, code combinations from the code output of BP ₁ , the corresponding 2 ₁ half-segments, are recorded in BP ₂ and appear at its code output. In addition, under the action of a short pulse delayed in an EZ, BP ₁ is zeroed out and starts recording code combinations corresponding to the following 2 ₁ half-segment (Fig. 8a).

After the PSU _{1 is} filled with code combinations corresponding to the 2 ₂ -semi-segment (2 ₂ ps in Fig. 8a), a seventh short pulse appears on the output of the counter (Fig. 8b), under the action of which a trigger is triggered and a “log. 1, from which a fourth short pulse appears at the shaper output (Fig. 8c).

Under the action of this pulse, code combinations from the code outputs of BP ₁ and BP _{2 are} recorded in BP ₃ and BP ₄ . Thus, the fourth segment is formed from 2 ₂ and 2 ₁ half segments (4 segments in Fig. 8a below). At the same time, under the action of the same short pulse, the memory unit is reset to its original state when a code combination appears on its code output, corresponding to the transfer coefficient to compensate for the unevenness of the Nuttoll window function for the first code combination in the fourth segment.

After that, under the action of the decay of the pulse from the output of the estimator, code combinations from the code output of BP ₁ , corresponding to a 2 ₂ half-segment. They are recorded in the BS and appear on its code output. In addition, under the action of a short pulse delayed in an EZ, BP ₁ is zeroed out and starts recording code combinations corresponding to the next 3 ₁ half-segment (Fig. 8a).

Under the action of pulses at the second inputs of BP ₃ and BP _4, 16-bit information code combinations from their code outputs go to the first and second code inputs of the adder, respectively. When summing, the addition of code combinations in the 2 ₂ half-segment (in which the transfer coefficients of the Natoll window are increasing) occurs with the same code combinations in the 2 ₁ half-segment (in which the transmission coefficients of the Nutoll window are decreasing), therefore, at the output of the adder, the transfer coefficients the Natoll windows are aligned (become close to 1), although some unevenness remains.

Further, after summation, code combinations from the code output of the adder (2 ₂ ps +2 ₁ ps in Fig. 8a) are fed to the first code input of the control system, to the second code input of which code combinations arrive from the output of the SP. After multiplying the code combinations, the non-uniformity of the Nuttall window function is compensated. If we compare (2 ₁ -1 ₂ ) segment and (3 ₁ -2 ₂ ) segment (at the top of Fig. 8a) at the entrance of SPSKNON with 4 segment (4 segments in Fig. 8a at the bottom or 2 ₂ + 2 ₁ segments in Fig. 8d) at the output of the adder, it can be seen that there is an addition with a 50% overlap of the segment with its previous segment.

The code output of the control unit receives 16-bit code combinations with compensated non-uniformity of the Nuttall window function. So at the output of the CU, the fourth segment is formed (2 ₂ + 2 ₁ segm. in Fig. 8d). Further work BPSKNON occurs in a similar way.

Thanks to this solution of the problem, the proposed method and companding audio broadcast signals, unlike the prototype, allows to avoid distortion of the sound signal shape (retains the waveform), reduce modulation of the high frequency components of the signal and noise by a variable transfer coefficient, and also reduce the visibility of noise not only during pauses but also in the signal. As a result, it is possible to improve the quality of transmission of sound broadcasting signals.

A feature of modern transmission channels is that, due to processing by existing compands, the transmitted sound broadcasting signals do not retain their shape and therefore cannot be qualitatively controlled by the existing metrological support focused on the measurement of the form. The proposed method and device retains the waveform, which allows the use of the existing metrological support in assessing the quality of the transmission (waveform). In addition, the preservation of the waveform and the resulting increase in quality is equivalent to increasing the bit depth of the digital representation not] 2 bits. This improvement in the quality of sound broadcasting signals allows a reduction in the transmission rate or volume of the signal during transmission and storage in exchange for a slight deterioration in this quality, which corresponds to the quality when processed by existing companders.

Using the proposed method and device can be transmitted as sound broadcasting signals, and speech signals, as well as any analog signals.

The proposed method and device can find application of existing analog and digital transmission channels, as well as in information storage systems. Their use will improve the quality of information transmission and reduce the transmission rate or signal volume in the communication channel.

The economic effect of the use of the present invention is expected to be obtained by ensuring the high quality of the transmission and reception of analog information signals. Preservation of the waveform allows the use of existing metrological assurance, rather than developing new measuring instruments in assessing the quality of transmission. The economic effect can also be obtained by reducing the speed of transmission or volume of the signal during its transmission and storage and increase as a result of this the number of channels.

Claims (3)

1. A method of companding audio broadcasting signals, including on the transmitting side a frequency correction of the original analogue broadcast audio signal, amplitude limitation of this signal, analog-to-digital conversion of the signal with the formation of a digital broadcast signal of transmission, amplitude compression of a digital signal, and on the receiving side digital expansion broadcast reception signal, digital-to-analog conversion with the formation of a reconstructed analog sound broadcasting system drive reception, inverse frequency correction and reception of the output analog audio broadcast signal, characterized in that on the transmitting side after the analog-digital conversion of the signal with the formation of a digital broadcast transmission signal, the Hilbert conjugate orthogonal transmission signal is formed and thus a complex transmission signal is obtained, from which the transmission phase cosine signal and the Hilbert amplitude envelope transmission signal are extracted from which the low-frequency, mid-frequency and high-frequency components of the transmission, and then the low-frequency the envelope components are compressed with a high compression ratio, the mid-frequency envelope components are compressed with a lower compression ratio, and the high-frequency components of the envelope are exposed, after which all three components of the Hilbert envelope transfer are summed up and the processed Hilbert amplitude transfer envelope is obtained, which is multiplied by the transmission phase cosine signal and the reconstructed after processing, a digital broadcast transmission signal, from which a digital-to-analogue transform form the output analog audio broadcast signal, and at the receiving side the amplitude limitation of the analog audio broadcast signal of the reception, the analog-digital conversion of this signal with the formation of the digital broadcast signal of the reception, the formation of the Hilbert conjugate orthogonal reception signal and the obtaining of a complex the reception signal, from which the cosine signal of the reception phase and the signal of the Hilbert amplitude reception envelope are extracted, from which By filtering, the low-frequency, mid-frequency and high-frequency components of the reception, and then the low-frequency components of the envelope are expanded with a large coefficient of expansion, the medium-frequency components of the envelope are expanded with a lower coefficient of expansion, and the high-frequency components of the envelope are compressed, after which the three components of the Hilbert envelope of the receive process are summed by the processing components of the envelope, and then the three components of the envelope are compressed, and then the three components of the Hilbert envelope of the reception will be processed by the total envelope processing components. the amplitude envelope of the reception, which is multiplied by the cosine signal of the reception phase and obtaining A reconstructed digital broadcasting reception signal is generated, from which a reconstructed analogue audio broadcasting signal is formed by digital-to-analog conversion, over which an inverse frequency correction is performed and an output analogue audio broadcasting signal is obtained. 2. A device for implementing a companding sound broadcasting method, comprising on the transmitting side serially connected sound source, a correction unit, a first limiter, a first analog-to-digital converter, and also a first compressor, and on the receiving side a first expander and serially connected first digital An analog converter and an inverse correction unit, characterized in that a first orthogonal signal generating unit is additionally introduced on the transmitting side, th unit modulation decomposition of the signal, the first low pass filter, the first band-pass filter, the first high pass filter, the second compressor, the second expander, the first adder, the first block of the modulation signal recovery and the second digital-to-analog converter, while the output of the first analog-to-digital converter is connected with the input of the first block forming an orthogonal signal, the first and second outputs of which are connected, respectively, with the first and second inputs of the first block of the modulation decomposition of the signal, ne the output of which is connected to the first input of the first modulation signal recovery unit, and its second output is connected to the input of the first low-pass filter, the input of the first band-pass filter and the input of the first high-pass filter, and the output of the first low-pass filter is connected to the input of the first compressor, the output of which connected to the first input of the first adder, the output of the first bandpass filter is connected to the input of the second compressor, the output of which is connected to the second input of the first adder, and the output of the first filter high frequencies connected to the input of the second expander, the output of which is connected to the third input of the first adder, the output of which is connected to the second input of the first modulation signal recovery unit, the output of which is connected to the input of the second digital-analog converter, the output of which is the output of the transmitting side of the device, and the second limiter, the second analog-to-digital converter, the second block forming the orthogonal signal, the second block of modulation times are additionally introduced to the receiving side signal, the second low pass filter, the second band pass filter, the second high pass filter, the third expander, the third compressor, the second adder and the second unit of the modulation signal recovery, while the input of the second limiter is the input side of the device, and its output is connected to the input of the second analog-to-digital converter, the output of which is connected to the input of the second block forming an orthogonal signal, the first and second outputs of which are connected, respectively, with the first and second inputs of the second b local modulation decomposition of the signal, the first output of which is connected to the first input of the second modulation signal recovery unit, and its second output is connected to the input of the second low-pass filter, the input of the second band-pass filter and the input of the second high-pass filter, with the output of the second low-pass filter connected to the input the first expander, the output of which is connected to the first input of the second adder, the output of the second bandpass filter is connected to the input of the third expander, the output of which is connected to the second input The second adder, and the output of the second high-pass filter is connected to the input of the third compressor, the output of which is connected to the third input of the second adder, the output of which is connected to the second input of the second modulation signal recovery unit, the output of which is connected to the input of the first digital-analog converter, the output of which is connected with the input of the reverse correction unit, the output of which is the output of the receiving side of the device.

Images