RU2575484C1 - Method of protecting acoustic information - Google Patents

Abstract

FIELD: physics.

SUBSTANCE: one version of the method of protecting acoustic information comprises installing a nonlinear converter (11) of a secondary electromagnetic noise signal in a facility to be protected in the immediate vicinity of an emitter (4) of an acoustic noise signal with a frequency spectrum whose width corresponds to the spectrum of the acoustic information signal, which is simultaneously the source of the secondary electromagnetic noise signal with the same frequency spectrum and statistical properties, said converter changing its frequency spectrum and statistical properties, thereby preventing compensation of the protective noise signal in eavesdropping equipment. In another version of the method the facility to be protected, near an emitter of an acoustic noise signal with a frequency spectrum whose width corresponds to the spectrum of the acoustic information signal, which is simultaneously the source of a secondary electromagnetic noise signal with the same frequency spectrum and statistical properties, is fitted with a nonlinear re-emitter-converter of the secondary electromagnetic noise signal, connected a generator, which performs intermodulation conversion of the secondary electromagnetic noise signal, thereby changing its frequency spectrum and statistical properties, thereby preventing compensation of the protective noise signal in eavesdropping equipment.

EFFECT: high effectiveness of protecting acoustic information by preventing compensation in eavesdropping equipment of a signal generated by a protective acoustic noise emitter.

2 cl, 10 dwg

Description

The invention relates to the field of protection of confidential information (CI) and can be used to protect acoustic CI in the premises to be protected (PZP).

To ensure the protection of acoustic (speech) CI, it is important to identify and sequentially block all leakage channels emerging from the BEP into the external environment. Examples of PPPs are premises (office rooms, meeting rooms and booths, conference rooms) designed to work with KI during meetings, negotiations, conferences, etc. Examples of speech CI leakage channels are direct acoustic, acousto-vibrational (vibroacoustic), acoustoelectric, acousto-electromagnetic and acousto-optic channels [1-2]. It is known from the prior art that both passive (using the effects of screening, absorption, re-reflection, etc. of sound waves) and active methods are currently used to protect speech CI from leakage through these channels. The idea of active KI protection is based on the use of a special type of signal (intentional interference), called up energetically (for masking noise interference) or by causing maximum information damage (for simulating interference) to “suppress” KI signals in leakage channels to make it difficult for an attacker to intercept processing of CI using his technical means of interception (TSP) [3-4]. Known options for the implementation of active protection KI using generators of low-frequency signals of different physical nature - sound, electrical, electromagnetic (EM), etc. noise.

The closest in technical essence is the method of spatial noise, implemented in the device [5] (prototype of the invention), which provides active protection of KI through direct acoustic and vibro-acoustic leakage channels using a radiator (generator complete with a radiating element) of sound noise. The obvious advantages of the known prototype method are its simplicity and versatility, a non-obvious disadvantage, as will be shown below, is the ability to compensate for a noise signal in a TSP, which significantly reduces the effectiveness of the protection of speech KI.

The technical result of the invention is to increase the efficiency of protection of acoustic KI by eliminating the possibility of compensation in the TSP signal generated by the emitter of sound noise, which is proposed to be implemented in the following three ways:

- using an additional emitter (generator complete with antenna) EM noise signal, statistically independent of sound noise and secondary EM noise signal;

- by applying a nonlinear converter of the side EM noise signal;

- by using a nonlinear re-emitter-converter of the side EM noise signal, which is under the influence of a generator that performs intermodulation (IM) conversion of the secondary EM noise signal.

The essence of the proposed method for the protection of acoustic information, including the placement in the room to be protected near the source of the acoustic information signal of the emitter of the acoustic noise signal with a frequency spectrum corresponding to the width of the spectrum of the acoustic information signal, consists in the following:

- in the first case, there is an additional emitter (generator with antenna) of an EM noise signal that is statistically independent of sound noise and secondary EM noise signal, with a frequency spectrum corresponding in width to the spectrum of the acoustic information signal;

- in the second case, in the immediate vicinity of the emitter of the acoustic noise signal with a frequency spectrum corresponding to the spectrum of the acoustic information signal, which is at the same time a source of the side EM noise signal with the same frequency spectrum and statistical properties, there is a nonlinear converter of the side EM noise signal;

- in the third case, near the emitter of a noise acoustic signal with a frequency spectrum corresponding to the spectrum of an acoustic information signal, which is also a source of a side EM noise signal with the same frequency spectrum and statistical properties, there is a nonlinear re-emitter-converter of a side EM noise signal that is affected a generator performing IM conversion of a secondary EM noise signal.

FIG. 1 shows a diagram of the implementation of the prototype - a known method of spatial noise of the PPP, where 1 is the source of speech KI; 2 - sound noise generator; 3 - emitter of sound noise; 4 - emitter side EM noise signal; 5 - the external propagation environment of the speech KI signal, sound and EM noise signals; 6 - a device for receiving a speech KI signal and sound noise; 7 - a device for receiving EM noise signal; 8 - signal processing unit with two inputs and one output; moreover, the elements of the circuit 3 and 4 are structurally combined in one common device, and the elements 6; 7 and 8 are part of the TSP, as shown by dashed contour lines; the first input of block 8 is connected to the output of the device 6, the second input of block 8 is connected to the output of the device 7.

FIG. 2 illustrates a diagram of an implementation of a first embodiment of the proposed method for protecting acoustic CI, where elements 1-8 correspond to the circuit of FIG. 1, element 9 is an additional generator of an electric noise signal; 10 - emitter of an additional EM noise signal, affecting the external environment 5 of the propagation of the speech KI signal, sound and EM noise signals.

FIG. 3 shows a diagram of an implementation of a second embodiment of the proposed method for protecting acoustic KI, where elements 1-8 correspond to the circuit of FIG. 1, element 11 is a non-linear converter of the side EM noise signal located between the emitter 4 of the side EM noise signal and the external environment 5 of the propagation of the KI speech signal, sound and EM noise signals.

FIG. 4 illustrates an implementation diagram of a third embodiment of the proposed method for protecting acoustic KI, where elements 1-8 correspond to the circuit of FIG. 1, element 12 is a nonlinear re-emitter-converter of the side EM noise signal located between the emitter 4 of the secondary EM noise signal and the external environment 5 of the propagation of the speech KI signal, sound and EM noise signals; element 13 - a generator that performs the IM conversion of the secondary EM noise signal, acting on the nonlinear re-emitter-converter 12.

FIG. 5 shows a fragment of an experimental setup for determining the level of a secondary EM noise signal acting on a DSP, illustrating the relative position of its two elements: a) a VI-45 vibration emitter of SONATA-AV equipment [5]; b) - the active antenna AI5-0, receiving a noise EM signal generated by the VI-45 vibration emitter and connecting wires.

FIG. Figure 6 shows the combined spectrograms of the EM background level in the laboratory (lower curve) and the secondary EM noise signal received by the AI5-09 antenna (upper curve), showing the presence of the secondary EM noise signal generated by the VI-45 vibration emitter of SONATA-AV equipment [5] (elements 3 and 4 in the composition of the TSP in the diagrams of Fig. 1-4) and connecting wires.

FIG. 7 shows the combined waveforms of electrical signals corresponding to sound noise and a secondary EM noise signal at the input of the signal processing unit 8 as a part of the TSP, which prove the possibility of their mutual compensation in order to reduce the effectiveness of the active protection of acoustic KI.

FIG. 8 illustrates the result of the IM-conversion of the frequency spectrum of the side EM noise signal (lower curve) due to the interaction of an intense harmonic signal from it from the generator 13 — the spectrogram of the converted side EM noise signal corresponds to the upper curve.

FIG. 9 shows a variant of the simplest implementation of converter 11 (see the diagram of FIG. 3) in the form of a frame antenna RA loaded on a non-linear element NE, conventionally shown as a diode.

FIG. 10 shows two possible embodiments of a re-emitter-converter 12 (see the diagram of FIG. 4):

- in the form of a frame antenna RA, loaded on non-linear elements NE 1-2, connected through a transformer TP to the generator 13, the signal from which interacts with the secondary EM noise signal during IM conversion of the latter to NE 1-2;

- in the form of a frame antenna RA loaded on a nonlinear element NE 1-4 in the form of a four-element bridge, which is connected to the generator 13, the signal from which interacts with the secondary EM noise signal during the IM conversion of the latter to NE 1-4.

The known prototype method is as follows.

In the absence of active protection, the source of speech KI 1 in the diagram of FIG. 1 through the external environment 5 acts on the receiver 6 of the KI signal, an electric signal with a power level P _with the output of which goes to the first input of the signal processing unit 8 and then to the attacker acting as an unauthorized user of the KI - and depending on the properties of the external environment 5, a diagram of FIG. 1 can correspond to both direct acoustic and vibroacoustic, and other KI leakage channels. To prevent this, the circuit of FIG. 1, the generator 2 and the emitter 3 of sound noise are introduced — as a result, at the output of the receiver 6, simultaneously with the electric signal P _c corresponding to the KI speech signal, there is an electric signal corresponding to sound noise with a dispersion D _w , therefore the value of the “interference the signal »χ = D _w / P _s in the DSP sharply increases, speech intelligibility and the bandwidth of the interception channel of the CI fall. Due to its simplicity, versatility and effectiveness, this method of protecting acoustic KI is currently widely used.

The presented results of experimental measurements (see Fig. 5 and Fig. 6) show, however, that simultaneously with the electric noise signal D _W, the DSP is affected by an electric noise signal with a dispersion D _e corresponding to the side EM noise signal generated by the emitter 4 in the circuit of FIG. . 1 - since elements 3 and 4 are inextricably combined in one common design connected to the sound noise generator 2 (an acoustic signal emitter with connecting wires creates an external “spurious” EM field in the external environment with a spectrum that completely corresponds to the frequencies of the acoustic signal). If there is an EM-noise receiver 7 in the TSC, as shown in the diagram of FIG. 1, and block 8 implements the simplest operation of linear signal processing, the following situation takes place (see combined experimental graphs of Fig. 7):

- the first input of block 8 from the output of device 6 receives a signal P _s + D _w ;

- the second input of block 8 from the output of the device 7 receives a signal D _e ;

; R - коэффициент корреляции между сигналом, соответствующим звуковому шуму, и сигналом, соответствующим ЭМ-шуму [6];- block 8 performs the total-difference conversion of these signals and generates an output signal of the form P _c + D _∑ , where $$D_{\Sigma }=D_w\pm 2R\sqrt{D_w{D}_{\mathrm{uh}}}+D_{\mathrm{uh}}$$

; R is the correlation coefficient between the signal corresponding to sound noise and the signal corresponding to EM noise [6];

, если дисперсии шумовых сигналов, принятых устройствами 6 и 7 в ТСП выровнены по мощности, то есть D_э≈D_ш;- since the statistical properties of these noise signals (close to normal white noise, for the implementation of which the correlation coefficient R is a measure of mutual independence) with variances D _w and D _{e are} determined by the same sound generator 2 in the diagram of FIG. 1, the value of R≈1, and if block 8 implements the differential conversion of signals P _c + D _w and D _e , then the dispersion of the noise signal after conversion will be equal to $$D_{\Sigma }=D_w-R\sqrt{D_w{D}_{\mathrm{uh}}}+D_{\mathrm{uh}}\approx 2D_w\left(\mathrm{one}-R\right)<<D_w$$

if the variances of the noise signals received by devices 6 and 7 in the DSP are power equalized, that is, D _e ≈D _w ;

- the ratio "interference-signal» χ = D _Σ / P _{with w} << D / P _with a PMT falls, intelligibility and bandwidth interception channel CI increases, the efficiency of active protection CI speech signal sharply decreases, for example, when R ≥0.95 the value of χ decreases by 10 or more times (more than 10 dB).

The aim of the invention is to eliminate this disadvantage of the prototype method.

The proposed method is carried out in the following three options.

1) In the PPP, along with the source of 1 voice CI, generator 2 of sound noise, emitters 3 of sound and 4 side EM noise, an additional generator of electric noise signal 9 and emitter 10 of EM noise signal acting on the external environment 5 are additionally placed, as shown in the diagram of FIG. 2, creating a noise electrical signal with a dispersion D _n in the DSP, statistically independent of sound noise. The situation in the TSP is as follows:

- the first input of block 8 from the output of device 6 receives a signal P _s + D _w ;

- to the second input of block 8 from the output of the device 7 receives signals D _e and D _n ;

, поскольку дисперсии шумовых сигналов соответствуют условию D_э≈D_ш, дисперсия дополнительного шумового сигнала соответствует условию D_n≥D_ш, а коэффициент корреляции между сигналом, соответствующим звуковому шуму и ЭМ-шуму, с одной стороны, и дополнительным шумовым сигналом с другой стороны, равняется нулю;- block 8 performs a differential conversion of these signals and generates an output signal of the form P _c + D _∑ , where $$D_{\Sigma }=D_w-2R\sqrt{D_w{D}_{\mathrm{uh}}}+D_{\mathrm{uh}}+D_n\approx 2D_w\left(\mathrm{one}-R\right)+D_n\ge D_w$$

since the variances of the noise signals correspond to the condition D _e ≈D _w , the variance of the additional noise signal corresponds to the condition D _n ≥D _w , and the correlation coefficient between the signal corresponding to the sound noise and EM noise, on the one hand, and the additional noise signal, on the other hand equals zero;

- the value of the "noise-signal" χ = D _∑ / P _s ≥D _w / P _s in the DFT increases to a value determined by the power level of the additional noise signal (dispersion value D _n ), speech intelligibility and bandwidth of the interception channel of the CI are reduced, the effectiveness of the active protection of the speech KI signal increases to any desired level.

From a comparison of the circuits of FIG. 2 and FIG. 1 it can be seen that the first version of the proposed method for protecting acoustic CI significantly exceeds the prototype method in terms of efficiency, since it virtually eliminates the possibility of compensating for the protective (masking) noise signal in the DSP, which is a significant drawback of the prototype method.

A relative disadvantage of the first option is the need to place an additional generator 9 and EM emitter 10 of the noise signal in the PPP, which can negatively affect the environmental and ergonomic safety of personnel jobs due to the EM radiation factor inside the PPP.

2) In the PZP, along with a source of 1 voice CI, a generator of 2 sound noise, emitters 3 of sound and 4 side EM noise, in the immediate vicinity of the emitter 4 of side EM noise there is a non-linear converter 11 of the side EM signal. Due to the electromagnetic interaction with the converter 11, the load of the emitter 4 of the side EM noise signal becomes non-linear, due to which the frequency spectrum and the statistical properties of the side EM noise affecting the external environment 5 are transformed - as a result of which they acquire significant differences from the frequency spectrum and statistical properties sound noise used to protect CI.

, а побочный ЭМ шумовой сигнал представлен суммой спектральных составляющих $$U_{вх}={\displaystyle \sum _{n=1}^N{U}_n\left({\omega }_n\right)}$$

, где N>>1, то различие между U_вых и U_вх по частотному спектру и статистическим свойствам является очевидным. Поэтому компенсация шумовых электрических сигналов, которую иллюстрирует Фиг. 7, на входе блока 8 в ТСП становится невозможной и снижения эффективности защиты акустической КИ не происходит. Реализацию нелинейного преобразователя 11 побочного ЭМ шумового сигнала в виде рамочной антенны РА, нагруженной на диодный нелинейный элемент НЭ, демонстрирует Фиг. 9. Сохраняя все преимущества первого варианта реализации предлагаемого способа перед прототипом, позитивным отличием от него в данном случае является отсутствие в ПЗП дополнительного генератора 9 и излучателя 10 ЭМ шумового сигнала, что не ухудшает эколого-эргономическую безопасность рабочих мест персонала по фактору ЭМ-излучения внутри ПЗП.For example, if the characteristic of the set of elements 4 and 11 is approximated by a polynomial of the form $$U_{\mathrm{at}sx}={\alpha }_{\mathrm{one}}{U}_{\mathrm{at}x}+{\alpha }_2{U}_{\mathrm{at}\mathrm{<figure-callout\; id="2"\; label="x"\; filenames="00000010.png,00000011.png"\; state="\{\{state\}\}">x</figure-callout>}}^2+{\alpha }_3{U}_{\mathrm{at}\mathrm{<figure-callout\; id="3"\; label="x"\; filenames="00000010.png,00000011.png"\; state="\{\{state\}\}">x</figure-callout>}}^3...$$

and the side EM noise signal is represented by the sum of the spectral components $$U_{\mathrm{at}x}={\displaystyle \sum _{n=\mathrm{one}}^N{U}_n\left({\omega }_n\right)}$$

Where N >> 1, the difference between U _out and U _Rin of the frequency spectrum and the statistical properties is obvious. Therefore, the compensation of noise electrical signals, which is illustrated in FIG. 7, it becomes impossible at the input of block 8 to the TSP, and the effectiveness of protection of acoustic KI does not decrease. The implementation of the non-linear converter 11 of the side EM noise signal in the form of a frame antenna RA loaded on a diode non-linear element NE is shown in FIG. 9. Keeping all the advantages of the first embodiment of the proposed method over the prototype, the positive difference from it in this case is the absence of an additional generator 9 and emitter 10 of the EM signal in the PZP, which does not impair the environmental and ergonomic safety of personnel’s workplaces by the EM radiation factor PZP.

A relative disadvantage of the second option may be the insufficient efficiency of the nonlinear conversion in the elements 4 and 11 of the side EM noise signal due to the small values of the coefficients α ₁ inherent in them; α ₂ ; α ₃ ..., due to which the difference between the noise electrical signals at the input of block 8 in the DSP, which is illustrated in FIG. 7 may not be sufficient to prevent mutual compensation.

3) In the PPP, along with the source 1 of the voice CI, the generator 2 of sound noise, the emitters 3 of the sound and 4 side EM noise, between the radiator 4 of the side EM noise and the external environment 5 is a nonlinear re-emitter-converter 12 of the side EM noise signal connected to the generator 13, which acts on it and carries out the IM conversion of the secondary EM noise signal. By changing the level of the signal supplied to the re-emitter-converter 12 from the generator 13, it is possible to control the efficiency of the IM conversion - therefore, to place the element 12 in the immediate vicinity of the emitter 4 of the secondary EM noise signal is not necessary in this case.

, а побочный ЭМ шумовой сигнал и сигнал генератора 13 могут быть представлены суммами спектральных составляющих, соответственно, $$U_{вх}={\displaystyle \sum _{n=1}^N{U}_n\left({\omega }_n\right)}$$

и $$U_{ген}={\displaystyle \sum _{m=1}^M{U}_m\left({\omega }_m\right)}$$

, то продукты ИМ-преобразования порядка mn будут иметь вид $$U_{вых}^{mn}={\alpha }_{mn}{U}_{вх}{U}_{ген}$$

- при этом различие между $$U_{вых}^{mn}$$

и U_вх по частотному спектру и статистическим свойствам может быть существенно «усилено» путем рационального выбора U_ген. Графики Фиг. 8 показывают, что даже в самом простом случае: когда U_ген является гармоническим сигналом, преобразование частотного спектра побочного шумового сигнала является весьма заметным - причем с помощью изменения уровня сигнала U_ген эффективностью ИМ-преобразования можно управлять.If the characteristic of element 12 is still approximated by a polynomial $$U_{\mathrm{at}sx}={\alpha }_{\mathrm{one}}{U}_{\mathrm{at}x}+{\alpha }_2{U}_{\mathrm{at}\mathrm{<figure-callout\; id="2"\; label="x"\; filenames="00000010.png,00000011.png"\; state="\{\{state\}\}">x</figure-callout>}}^2+{\alpha }_3{U}_{\mathrm{at}\mathrm{<figure-callout\; id="3"\; label="x"\; filenames="00000010.png,00000011.png"\; state="\{\{state\}\}">x</figure-callout>}}^3...$$

and the side EM noise signal and the signal of the generator 13 can be represented by the sums of the spectral components, respectively, $$U_{\mathrm{at}x}={\displaystyle \sum _{n=\mathrm{one}}^N{U}_n\left({\omega }_n\right)}$$

and $$U_{gen}={\displaystyle \sum _{m=\mathrm{one}}^M{U}_m\left({\omega }_m\right)}$$

, then the products of the IM transform of order mn will have the form $$U_{\mathrm{at}sx}^{mn}={\alpha }_{mn}{U}_{\mathrm{at}x}{U}_{gen}$$

- the difference between $$U_{\mathrm{at}sx}^{mn}$$

and U _Rin of the frequency spectrum and the statistical properties can be substantially "reinforced" by the rational choice U _gene. Graphs of FIG. 8 show that even in the simplest case: when the U _gene is a harmonic signal, the conversion of the frequency spectrum of the secondary noise signal is very noticeable - and by changing the signal level of the U _{gene, the} efficiency of the IM conversion can be controlled.

и U_вх на входе блока 8 в ТСП (см. Фиг. 7) становится невозможной и снижения эффективности защиты акустической КИ в данном случае не происходит.As a result of such a controlled IM conversion, the frequency spectrum and the statistical properties of the incident EM noise acquire significant differences from the frequency spectrum and the statistical properties of sound noise used to protect KI (see Fig. 8). As a result, the compensation of noise electrical signals $$U_{\mathrm{at}sx}^{mn}$$

and U _I at the input of block 8 in the TSP (see Fig. 7) becomes impossible and in this case there is no decrease in the efficiency of protection of acoustic KI.

Two options for the implementation of a nonlinear re-emitter-converter 12 of the side EM noise signal: in the form of a frame antenna RA loaded on non-linear elements NE 1-2 connected through a transformer TP to the generator 13, the signal of which interacts with the side EM noise signal during IM conversion of the latter to NE 1-2 (see Fig. 10a) and in the form of a frame antenna RA loaded on a nonlinear element NE 1-4 in the form of a four-element bridge, which is connected to the generator 13, the signal from which interacts with the secondary EM noise signal IM-converting the latter to NE 1-4 (see. Fig. 10b), is subject to control MI-conversion process by changing the signal level input to the member 12 by a generator 13.

The practical effectiveness of the proposed method for the protection of acoustic KI is confirmed, on the one hand, by experimental results, examples of which are presented in FIG. 6-8, on the other hand, by theoretical and calculated data of publications [7-9]. The proposed method for the protection of acoustic KI is not inferior to the prototype method in terms of versatility and simplicity, but significantly surpasses it in efficiency, since it eliminates the possibility of eliminating the protective (masking) signal generated by the sound emitter in block 8 of the TSP.

LITERATURE

1. Buzov G.A., Kalinin S.V., Kondratiev A.V. Protection against information leakage through technical channels. M .: Hot line - Telecom, 2005 .-- 416 p.

2. Horev A.A. Technical information security. Volume 1. Technical channels of information leakage. M .: SPC "Analytics", 2008. - 436 p.

3. Sobolev A.N., Kirillov V.M. Physical basis of technical means of ensuring information security. M .: Helios ARV, 2004 .-- 224 p.

4. Zaitsev A.P., Shelupanov A.A. Technical means and methods of information protection. Tomsk: V-Spectrum, 2006 .-- 384 p.

5. System of vibroacoustic and acoustic protection "SONATA-AV". Model 1M. Operating Instructions APEM.468781.003-02. RE. Moscow, 2004 .-- 15 p.

6. Levin B.R. Theoretical foundations of statistical radio engineering. Prince 1. M .: Sov. Radio, 1969 .-- 752 p.

7. Maslov O.N. Low-energy information protection of random antennas // Telecommunication. No. 1, 2014 .-- S. 32-38.

8. Maslov O.N., Scherbakova T.A. Complex modeling of active information protection systems // Information Protection. No. 6, 2013. - S. 34-39.

9. Maslov O.N. Random antennas: theory and practice. Samara: Publishing house of PSUTI-OFORT, 2013 .-- 480 p.

Claims (2)

1. A method of protecting acoustic information, including placing in a room to be protected near the source of the acoustic information signal the emitter of the acoustic noise signal with a frequency spectrum corresponding to the width of the spectrum of the acoustic information signal, which is also a source of secondary electromagnetic noise signal with the same frequency spectrum and statistical properties characterized in that in the immediate vicinity of the emitter of a noise acoustic signal with cha -frequency spectrum corresponding to the width of the spectrum of the acoustic information signal that is simultaneously a source of natural electromagnetic noise signal with the same frequency spectrum and statistical properties is nonlinear converter spurious electromagnetic noise signal. 2. A method of protecting acoustic information, comprising placing in a room to be protected near the source of the acoustic information signal the emitter of the acoustic noise signal with a frequency spectrum corresponding to the width of the spectrum of the acoustic information signal, which is also a source of a secondary electromagnetic noise signal with the same frequency spectrum and statistical properties characterized in that in the immediate vicinity of the emitter of a noise acoustic signal with cha -frequency spectrum corresponding to the width of the spectrum of the acoustic information signal that is simultaneously a source of natural electromagnetic noise signal with the same frequency spectrum and statistical properties is a non-linear re-radiator converter spurious electromagnetic noise signal, connected to the generator, conducting the intermodulation conversion of natural electromagnetic noise signal.

Images