Electromagnetic sound information leakage prevention device, method, and program

Abstract

Provided are an electromagnetic sound information leakage prevention device, method, and program which are capable of effectively preventing leakage of sound information through electromagnetic waves. The present invention comprises: a reproduced sound input means 2; a noise input means 3; a first synthesized sound signal generating means 4; a second synthesized sound signal generating means 5; and a synthesized sound signal output means 6. The reproduced sound input means 2 inputs reproduced sound from a sound source 7. The noise input means 3 receives an input of noise 80. The first synthesized sound signal generating means 4 generates a first synthesized sound signal by adding the noise 80, the input of which has been received by the noise input means 3, to the reproduced sound signal of the reproduced sound inputted to the reproduced sound input means 2. The second synthesized sound signal generating means 5 generates a second synthesized sound signal by adding noise, which has a phase opposite to that of the noise 80 and the input of which has been received by the noise input means 3, to a reproduced sound signal which has the same phase as the reproduced sound inputted to the reproduced sound input means 2. The synthesized sound signal output means 6 simultaneously outputs the first synthesized sound signal and the second synthesized sound signal to a speaker 9.

Description

Device, method and program for preventing electromagnetic sound information leakage

[0001] The present invention relates to a technique for preventing leakage of sound information via electromagnetic waves.

[0002] One method of ensuring the security of information devices has been to connect devices handling highly confidential information to private networks isolated from public networks such as the Internet, or to use them standalone without any network connection. This eliminates potential attack vectors and ensures safety. However, when information devices process information, time-varying electromagnetic signals such as current and voltage inevitably occur, and the resulting electromagnetic waves are emitted from the device. Because the emitted electromagnetic waves contain internal processing information, there is a risk of confidential information being leaked. In particular, input / output signals from information devices intended for humans are not encrypted, meaning that attackers could instantly ascertain information by measuring and analyzing them. Furthermore, with the rapid increase in the use of sound information due to remote work and other factors since the COVID-19 pandemic, numerous reports of electromagnetic wave leaks have emerged, but no definitive countermeasures have yet been reported. Furthermore, while measures in the field of environmental electromagnetic compatibility (EMC) have been considered as a means of suppressing electromagnetic waves, these countermeasures are being developed from the perspective of electromagnetic compatibility, making it difficult to fully suppress the electromagnetic waves that cause information leaks.

[0003] Conventional countermeasures against electromagnetic information leakage from audio devices are limited to measures to suppress the emission of leaked electromagnetic waves containing sound information. For example, Non-Patent Documents 1 to 3 cite techniques for shielding audio devices as a countermeasure. In addition, Non-Patent Document 1 cites a design that reduces the electromagnetic coupling between a regulator IC and an audio amplifier, which is the dominant factor in leakage. Furthermore, techniques that utilize out-of-phase sound signals have been disclosed as a method for suppressing sound leakage to the outside (see Patent Document 1) and improving the directionality of a sound field (see Patent Document 2). These methods are used to improve the performance of speakers and the like.

[0004] Conventional countermeasures have focused on the attenuation of electromagnetic wave intensity during the propagation of leaked electromagnetic waves containing sound information. However, because electromagnetic waves emitted from audio devices are generated by temporal fluctuations during the transmission of electrical signals, it is difficult to suppress the generation of these electromagnetic waves themselves according to the laws of physics. Therefore, while conventional countermeasures can suppress the intensity of electromagnetic radiation emitted from devices to a certain level, it is extremely difficult to suppress it below background noise, which makes it impossible for attackers to recover sound information. Furthermore, as measuring instruments become smaller and more affordable, attackers can now use portable setups to launch attacks in various locations. As a result, countermeasures against information leakage through electromagnetic waves, which require the installation of equipment to account for attenuation due to walls, buildings, and distance, are no longer as effective as they were in the past.

[0005] JP 2022-099194 A Patent No. 7398647 A

[0006] J. Choi, H.-Y. Yang, and D.-H. Cho, “TEMPEST Comeback: A Realistic Audio Eavesdropping Threat on Mixed-signal SoCs,” in Proceedings of the 2020 ACM SIGSAC Conference on Computer and Communications Security, New York, NY, USA: Association for Computing Machinery, 2020, pp. 1085-1101. NDSS 2024.

[0007] In view of the above circumstances, an object of the present invention is to provide an apparatus, method and program for preventing electromagnetic sound information leakage that can effectively prevent sound information from leaking via electromagnetic waves.

[0008] The present inventors have discovered that, rather than making measurement difficult by suppressing the radiation intensity of electromagnetic waves, it is possible to make it difficult for an attacker to restore the sound information by taking measures applicable to a "leakage source" from which sound information leaks as electromagnetic waves, without impairing the ease with which a listener perceives the sound. That is, an electromagnetic sound information leakage prevention device according to a first aspect of the present invention is a device for preventing leakage of sound information via electromagnetic waves, and comprises: a playback sound input means for inputting a playback sound from a sound source; a noise input means for accepting input of noise; a first synthetic sound signal generation means for adding noise to a playback sound signal of the input playback sound to generate a first synthetic sound signal; a second synthetic sound signal generation means for adding noise of an opposite phase to the noise added by the first synthetic sound signal generation means to a playback sound signal that is in phase with the playback sound signal, to generate a second synthetic sound signal; and a synthetic sound signal output means for simultaneously outputting the first synthetic sound signal and the second synthetic sound signal to a speaker. This enables the realization of a countermeasure to "prevent sound information from being included in electromagnetic waves at the source of electromagnetic waves." In other words, the present invention is a countermeasure applied to the "leakage source" of sound information leaking through electromagnetic waves, making it difficult for an attacker to recover the sound information even under conditions where the attacker can observe the electromagnetic waves emitted from an audio device. The sound source here is intended to broadly encompass any source of sound, including data containing sound information, such as audio files and video files, as well as media on which such data is recorded. Furthermore, although the first synthetic sound signal and the second synthetic sound signal are output simultaneously to one speaker, they may also be output simultaneously to multiple speakers. Note that electromagnetic sound information leakage refers to the leakage of sound information via electromagnetic waves.

[0009] An electromagnetic sound information leakage prevention device according to a second aspect of the present invention is a device for preventing leakage of sound information via electromagnetic waves, and includes: a playback sound input means for inputting a playback sound from a sound source; a noise input means for accepting input of noise; a first synthetic sound signal generation means for adding a first noise to a playback sound signal of the input playback sound to generate a first synthetic sound signal; a second synthetic sound signal generation means for adding a second noise having a phase difference of 120° from the first noise to a playback sound signal that is in phase with the playback sound signal to generate a second synthetic sound signal; a third synthetic sound signal generation means for adding a third noise having a phase difference of 240° from the first noise to a playback sound signal that is in phase with the playback sound signal to generate a third synthetic sound signal; and a synthetic sound signal output means for simultaneously outputting the first synthetic sound signal, the second synthetic sound signal, and the third synthetic sound signal to a speaker. Since the sound sources to which three different types of noise, each with a phase difference of 120°, are added are output simultaneously, the three types of noise, each with a phase difference of 120°, cancel each other out when they reach the listener's ear, making it possible to deliver clear sound to the listener. Note that for the three types of noise, each with a phase difference of 120°, to cancel each other out, the distance from the position of each sound source to the position of the listener must be the same, but if the distance from each sound source to the listener is sufficiently far and the sound sources are sufficiently close to each other so that they can be considered to be essentially the same sound source, the noises cancel each other out and clear sound can be delivered to the listener.

[0010] An electromagnetic sound information leakage prevention device according to a third aspect of the present invention is a device for preventing leakage of sound information via electromagnetic waves, and comprises: a playback sound input means for inputting a playback sound from a sound source; a noise input means for accepting input of noise; a first synthetic sound signal generation means for adding a first noise to a playback sound signal of the input playback sound to generate a first synthetic sound signal; an nth synthetic sound signal generation means for adding an nth noise having a phase difference of 360°×(n−1) / m (n is an integer of 2 or greater and m is an integer of 3 or greater) from the first noise to a playback sound signal that is in phase with the playback sound signal to generate the nth synthetic sound signal; and a synthetic sound signal output means for repeating the nth synthetic sound signal generation means from n=2 up to m by adding 1 at a time, and for simultaneously outputting the first to mth synthetic sound signals to a speaker. With this configuration, even when m different types of noise are added, the noises can be canceled out, preventing leakage of electromagnetic sound information and delivering clear sound to the listener. Here, n corresponds to the index of the synthetic sound signal (or speaker), and m corresponds to the total number of synthetic sound signals (or speakers). For example, if the total number of speakers is four (m = 4), the first speaker adds a first noise to the playback sound signal of the playback sound to generate and output a first synthetic sound signal. The second speaker (n = 2) adds a second noise, which is in phase with the playback sound signal and has a phase difference of 360° × (2 - 1) / 4 = 90° from the first noise, to generate and output a second synthetic sound signal. The third (n=3) speaker adds the first noise and a third noise with a phase difference of 360°×(3−1) / 4=180° to a reproduction sound signal that is in phase with the reproduction sound signal, to generate and output a third synthetic sound signal. The fourth (n=4) speaker adds the first noise and a fourth noise with a phase difference of 360°×(4−1) / 4=270° to a reproduction sound signal that is in phase with the reproduction sound signal, to generate and output a fourth synthetic sound signal.

[0011] In the electromagnetic sound information leakage prevention device according to the first aspect of the present invention, the first or second synthetic sound signal generating means may be configured in an amplifier circuit such that the noise signal to be added passes through an inverting stage of the output stage, and the noise signals are in phase or out of phase between the output terminals. More specifically, in a class D amplifier BTL (bridge-tied load) configuration, the first or second synthetic sound signal generating means is configured so that the noise signal does not pass through an inverting amplifier in the amplifier, and so the noise signals are in phase between the output terminals, thereby easily realizing noise cancellation.

[0012] The electromagnetic sound information leakage prevention device according to any one of the first to third aspects of the present invention may be provided with a noise generating means for generating noise within the device, instead of a noise input means for inputting noise from outside the device. By generating noise within the device, it is possible to eliminate the need for noise input from outside the device, and to simplify the external interface of the device body.

[0013] In the electromagnetic sound information leakage prevention device according to any one of the first to third aspects of the present invention, the noise generating means may generate noise whose amplitude or frequency components are dynamically changed according to the characteristics of the input audio information. This makes it possible to generate noise that makes it difficult to recognize specific audio components, thereby more effectively preventing leakage of audio information via electromagnetic waves.

[0014] In the electromagnetic sound information leakage prevention device according to any one of the first to third aspects of the present invention, the noise generating means may generate wideband white noise having frequency components within the audible range, thereby more effectively preventing leakage of sound information via electromagnetic waves.

[0015] In the electromagnetic sound information leakage prevention device according to the first aspect of the present invention, the synthetic sound signal output means may simultaneously output a first synthetic sound signal to the first speaker or the first speaker group and a second synthetic sound signal to the second speaker or the second speaker group. That is, the synthetic sound signal output means may output the first synthetic sound signal to the first speaker or the first speaker group and the second synthetic sound signal to the second speaker or the second speaker group, and these outputs may be performed simultaneously. As a result, even if different synthetic sound signals are output to different speakers, opposite-phase noises cancel each other out before the signals reach the listener's ears, thereby ensuring clear sound to the listener. The first speaker and the second speaker are not limited to one each, and multiple speakers may be provided. In such a case, the synthetic sound signal output means outputs the first synthetic sound signal to the first speaker group and the second synthetic sound signal to the second speaker group, and these outputs are performed simultaneously. In the electromagnetic sound information leakage prevention device of the second aspect of the present invention, the synthetic sound signal output means may simultaneously output the first synthetic sound signal to the first speaker or the first speaker group, the second synthetic sound signal to the second speaker or the second speaker group, and the third synthetic sound signal to the third speaker or the third speaker group.

[0016] The electromagnetic sound information leakage prevention device according to any one of the first to third aspects of the present invention may include a detection means for detecting the level of electromagnetic radiation indicating leakage, and a control means for automatically adjusting at least one of the amplitude and spectrum of the noise generated by the noise generation means based on the detection result by the detection means. Increasing the amplitude value of the noise generation means improves the effectiveness of preventing leakage of electromagnetic sound information, but at the same time increases power consumption. Furthermore, if there is another electronic device in close proximity to the device being prevented from leaking, the electromagnetic waves emitted from that electronic device may contribute to preventing leakage of electromagnetic sound information. In such cases, a sufficient leakage prevention effect can be achieved without the noise generation means generating noise with a high amplitude value. Therefore, by including such a detection means and control means, the device can be made to have both leakage prevention performance and energy-saving performance. Furthermore, adding noise with a high amplitude increases power consumption, which in turn increases the intensity of unwanted radiation, potentially causing self-poisoning or electromagnetic interference with other devices. Therefore, suppressing the amplitude of the noise generated by the noise generating means also helps to ensure the stable operation of other electronic devices.

[0017] In the electromagnetic sound information leakage prevention device according to any one of the first to third aspects of the present invention, the detection means may continuously monitor the level of electromagnetic radiation indicating leakage, and the control means may dynamically select at least one from a plurality of noise control modes based on the monitoring results, and automatically adjust at least one of the amplitude and spectrum of the noise generated by the noise generation means. This makes it possible to provide a device that is highly convenient and has high energy-saving performance.

[0018] The electromagnetic sound information leakage prevention method of the present invention is a method for preventing leakage of sound information via electromagnetic waves, and causes a computer to perform the following steps: a playback sound input step for inputting playback sound from a sound source; a noise input step for accepting noise input; a first synthetic sound signal generation step for adding noise to a playback sound signal of the input playback sound to generate a first synthetic sound signal; a second synthetic sound signal generation step for adding noise of an opposite phase to the noise added in the first synthetic sound signal generation step to a playback sound signal that is in phase with the playback sound signal to generate a second synthetic sound signal; and a synthetic sound signal output step for simultaneously outputting the first synthetic sound signal and the second synthetic sound signal to a speaker.

[0019] The electromagnetic sound information leakage prevention method of the present invention may include a noise generation step of generating noise inside the leakage prevention device instead of the noise input step of inputting noise from outside the leakage prevention device. By generating noise inside the leakage prevention device, it is possible to eliminate the need to input noise from outside the leakage prevention device and simplify the external interface of the leakage prevention device main body.

[0020] In the electromagnetic sound information leakage prevention method of the present invention, the noise generation step may generate noise that makes it difficult to recognize specific audio components by dynamically changing the amplitude or frequency components of the noise according to the characteristics of the input audio information, or may generate wideband white noise having frequency components within the audible range.

[0021] In the electromagnetic sound information leakage prevention method of the present invention, the synthetic sound signal output step may be to simultaneously output a first synthetic sound signal to a first speaker or a first group of speakers and a second synthetic sound signal to a second speaker or a second group of speakers.

[0022] The electromagnetic sound information leakage prevention program of the present invention causes a computer to execute each step of any of the above-described electromagnetic sound information leakage prevention methods.

[0023] The electromagnetic sound information leakage prevention device, method, and program of the present invention have the effect of effectively preventing sound information leakage via electromagnetic waves.

[0024] 1. Functional block diagram of the electromagnetic sound information leakage prevention device of Example 1. 2. Configuration image diagram of the electromagnetic sound information leakage prevention device of Example 1. 3. Flow diagram of the electromagnetic sound information leakage prevention method of Example 1. 4. Functional explanatory diagram of the electromagnetic sound information leakage prevention device of Example 1. 5. Image diagram of sound perceived by a listener. 6. Image diagram of sound restored from electromagnetic waves. 7. Configuration image diagram of evaluation experiment of the electromagnetic sound information leakage prevention device of Example 1. 8. Evaluation results of sound reproduced from a speaker. 9. Evaluation results of sound restored from electromagnetic waves. 10. Explanation diagram of the mechanism of sound information leakage. 11. Functional block diagram of the electromagnetic sound information leakage prevention device of Example 2. 12. Configuration image diagram of the electromagnetic sound information leakage prevention device of Example 2. 13. Flow diagram of the electromagnetic sound information leakage prevention method of Example 2. 14. Functional block diagram of the electromagnetic sound information leakage prevention device of Example 3. 15. Configuration image diagram of the electromagnetic sound information leakage prevention device of Example 4. 16. Functional block diagram of the electromagnetic sound information leakage prevention device of Example 5. 17. Functional block diagram of the electromagnetic sound information leakage prevention device of Example 6. 18. Configuration image diagram of the electromagnetic sound information leakage prevention device of Example 6.

[0025] First, we will explain the general mechanism of sound information leakage. Figure 10 is an explanatory diagram of the mechanism of sound information leakage, where (1) shows the leak source side and (2) shows the attacker side. As shown in Figure 10 (1), for example, a signal is output from an audio amplifier IC (integrated circuit) 100 and sound is reproduced in a speaker 9. In such a case, V ＋ and V － The desired sound is reproduced using these two voltage values, but as shown by the arrows in Figure 10 (1), electromagnetic waves are generated at the moment of rise or fall. ＋ and V － Since the electromagnetic waves generated from each of the two transmission signals (V ＋ , V － ) Since electromagnetic waves leak from each of them, the restored sound is the absolute value of the reproduced sound.

[0026] In contrast to this, the present invention implements measures to "prevent sound information from being included in electromagnetic waves at the source of the electromagnetic waves" by taking measures against the "leakage source" of sound information leaking through electromagnetic waves. In other words, the present invention employs a method that makes it difficult for an attacker to restore sound information even under conditions where the attacker can observe electromagnetic waves emitted from audio devices.

[0027] An example of an embodiment of the present invention will be described in detail below with reference to the drawings. Note that the scope of the present invention is not limited to the following examples and illustrated examples, and many modifications and variations are possible.

[0028] FIG. 1 shows a functional block diagram of an electromagnetic sound information leakage prevention device according to a first embodiment. As shown in FIG. 1, the electromagnetic sound information leakage prevention device 1 includes a reproduced sound input means 2, a noise input means 3, a first synthetic sound signal generation means 4, a second synthetic sound signal generation means 5, and a synthetic sound signal output means 6. The reproduced sound input means 2 inputs a reproduced sound from a sound source 7. The sound source 7 may be, for example, an audio file stored on a PC or the like. The noise input means 3 receives input of noise 80. The first synthetic sound signal generation means 4 generates a first synthetic sound signal by adding noise 80 received by the noise input means 3 to a reproduced sound signal of the reproduced sound input to the reproduced sound input means 2. The second synthetic sound signal generation means 5 adds noise of opposite phase to the noise 80 received by the noise input means 3 to a reproduced sound signal of the reproduced sound input to the reproduced sound input means 2, the reproduced sound signal being in phase with the reproduced sound signal of the reproduced sound input to the reproduced sound input means 2, thereby generating a second synthetic sound signal. The synthetic sound signal input means 6 outputs the first synthetic sound signal and the second synthetic sound signal to the speaker 9 simultaneously.

[0029] FIG. 2 shows a schematic diagram of the electromagnetic sound information leakage prevention device of the first embodiment. As shown in FIG. 2, a noise generator 18 is connected to an audio amplifier IC 10. The audio amplifier IC 10 and a battery 11 are connected via a power line 12. An audio file stored in a computer 13a is used as the sound source 7. The audio file is output as a digital audio signal, converted into an analog audio signal by a digital-to-analog converter (DAC) 17, and input to the audio amplifier IC 10. The audio amplifier IC 10 is also connected to a speaker 9. The audio amplifier IC 10 functions as a playback sound input means 2, a noise input means 3, a first synthesized sound signal generating means 4, a second synthesized sound signal generating means 5, and a synthesized sound signal output means 6, and can be implemented using a stereo amplifier IC used in known audio devices. The audio amplifier IC 10 utilizes two stereo output ports and combines them when connecting to the speaker 9, thereby enabling simultaneous output of the first synthesized sound signal and the second synthesized sound signal. However, since the audio must be in phase, it is output as a monaural signal. The reproduced sound input to the audio amplifier IC 10 is processed within the audio amplifier IC 10 to generate a signal with added noise, which is then output to the speaker 9. Specifically, a program in the audio amplifier IC 10 uses noise 80 input from the noise generator 18 to the reproduced sound input from the computer 13a via the DAC 17 to generate a synthesized sound signal with added in-phase and out-of-phase noise, and outputs the synthesized sound signal to the speaker 9. By performing processing to prevent leakage of electromagnetic sound information, it becomes extremely difficult to reproduce the reproduced sound, even if an attempt is made to restore it by measuring the electromagnetic waves leaking from the power line 12, for example.

[0030] FIG. 3 shows a flow diagram of the electromagnetic sound information leakage prevention method of Example 1. The electromagnetic sound information leakage prevention method of Example 1 is a method in which a computer prevents leakage of sound information via electromagnetic waves. As shown in FIG. 3 , first, a reproduced sound is input from a sound source 7 (step S01: reproduced sound input step). Next, a noise input means 3 accepts input of noise 80 (step S02: noise input step). A first synthetic sound signal generation means 4 adds noise to the reproduced sound signal of the input reproduced sound to generate a first synthetic sound signal (step S03: first synthetic sound signal generation step). Furthermore, a second synthetic sound signal generation means 5 adds noise of an opposite phase to the noise in the first synthetic sound signal generation step to a reproduced sound signal in phase with the reproduced sound signal to generate a second synthetic sound signal (step S04: second synthetic sound signal generation step). Thereafter, the synthetic sound signal input means 6 simultaneously outputs the first synthetic sound signal and the second synthetic sound signal to the speaker 9 (step S05: synthetic sound signal output step).

[0031] 4 is a diagram illustrating the function of the electromagnetic sound information leakage prevention device of the first embodiment. In the electromagnetic sound information leakage prevention device 1 of the present invention, noise having the following characteristics is added to a normal reproduction sound and reproduced. Specifically, as shown in FIG. 4, the first synthetic sound signal generating means 4 generates a reproduction sound signal S of the normal reproduction sound input to the reproduction sound input means 2. １ On the other hand, the noise N １ The second synthetic sound signal generating means 5 generates a first synthetic sound signal A by adding the above-mentioned signal to the reproduced sound input means 2. １ The reproduced sound signal S ２ On the other hand, the noise N １ and anti-phase noise N ２ to generate a second synthesized sound signal B.

[0032] Specifically, a normal reproduced sound signal (S １ , S ２ ) and antiphase noise (N １ , N ２) and outputs a synthesized sound signal (A, B) by adding the respective signals. Then, in the synthesized sound signal input means 6, the signals to be output are synthesized as an input signal for the same speaker, and output to the speaker 9. As a result, when the listener 31 perceives the sound generated from the speaker, the added noise (N １ , N ２ ) cancel each other out in phase and out of phase, so noise (N １ , N ２ ) becomes difficult to recognize, and only normal reproduced sound can be recognized. On the other hand, when a malicious third party 32 perceives the sound restored from the leaked electromagnetic waves, noise (N １ , N ２ ) is superimposed on the absolute value of the reproduced sound, making it difficult to recognize only the reproduced sound containing information.

[0033] 5 shows an image diagram of the sound perceived by the listener. As shown in FIG. 5, the sound perceived by the listener is firstly obtained by inputting the reproduced sound signal S １ and a reproduced sound signal S ２ are mutually reinforcing, and the reproduced sound signal S ３ On the other hand, the noise N １ And the opposite phase noise N ２ Since the noise N ３ Then, the first synthetic sound signal A and the second synthetic sound signal B are simultaneously output to the speaker 9 by the synthetic sound signal input means 6, thereby generating a reproduced sound signal S ３ and noise N ３ A reproduced sound signal S consisting of ４ The reproduced sound 41 based on this is output from the speaker 9 and reaches the ears of the listener 31 as clear sound without noise. In this way, when the listener 31 perceives the sound generated from the speaker, the normal reproduced sound is in phase and therefore reinforces each other, while the noise is in opposite phase and therefore cancels each other out, so the listener perceives only the normal sound, and the audio device can be used without impairing ease of listening.

[0034] Figure 6 shows an image of the sound restored from electromagnetic waves. As mentioned above, the V ＋ and V － Since each of these generates an electromagnetic wave, the reproduced sound leaks as such an absolute value component. １ [s(t)] + N １ [n(t)], a reproduced sound signal in phase with this, and a noise S ２ [s(t)] + N ２ [n(t)] is added as an absolute value component, and the electromagnetic wave E １ [|s(t) + n(t)|] and electromagnetic wave E ２ [|s(t)-n(t)|]. Then, the first synthetic sound signal A and the second synthetic sound signal B are simultaneously output to the speaker 9 by the synthetic sound signal output means 6, and the leakage electromagnetic wave 42 is ３ The resulting signal is [|s(t) + n(t)| + |s(t) - n(t)|], making it difficult to restore the information. In this way, when a malicious third party 32 perceives the sound restored from the leaked electromagnetic waves, the absolute value components of each output signal leak through the electromagnetic waves, so that for both the normal playback sound and the noise, sounds that are added together as in-phase signals with no phase difference are leaked. Therefore, the sound restored from the electromagnetic waves is normal playback sound with noise added to it, making it difficult to recognize the information.

[0035] That is, the restored sound from the electromagnetic waves is ultimately cancelled out when it comes out of the normal speaker 9, but since the restored sound from the electromagnetic waves leaks out in the form of a sum of noise components, it is possible to effectively prevent the leakage of sound information from the electromagnetic waves without impairing the ease of hearing the sound coming out of the speaker.

[0036] (Evaluation Experiment of Electromagnetic Sound Information Leakage Prevention Device) Next, experiments were conducted on the electromagnetic sound information leakage prevention device of Example 1 to determine whether the audibility of the sound emitted from the speaker is impaired (audibility), and (b) whether it is difficult to restore sound information via leaked electromagnetic waves, in other words, whether the sound becomes inaudible in the event of an actual attack by an attacker (effectiveness of the countermeasure). As an evaluation method, no noise was added to the comparative example, and white noise was added to the example, and a chirp sound was played for 5 seconds and swept over the frequency range of the human audible range to evaluate whether all sounds within the human audible range were protected.

[0037] (Evaluation Experiment Configuration) Figure 7 shows an image of the evaluation experiment configuration for the electromagnetic sound information leakage prevention device of Example 1. As shown in Figure 7, the equipment used in the evaluation experiment consisted of a waveform generator 70 as a sound source 7, an audio amplifier IC 10, a current probe 21, a battery 11 (DC 6V), a speaker 9, an amplifier 22, a spectrum analyzer 23, and a smartphone 24. The setup was constructed to simulate an actual attacker. The waveform generator 70 was a Trueform waveform generator from the "33600A Series" manufactured by Keysight. The audio amplifier IC 10 was the "AE-CD8755" class D power amplifier module, the current probe 21 was the "F-35A-L" manufactured by Fisher Custom Communications, and the amplifier 22 was the "R&K-LA101-0S" manufactured by R&K Corporation. The spectrum analyzer 23 used to evaluate the effectiveness of the countermeasures was an R&S® FSV4 manufactured by Rohde & Schwarz, with a center frequency of 3.634 MHz, a sampling rate of 1 MHz, and a bandwidth of 800 kHz. The spectrum analyzer 23 frequency-converted the input signal (real signal) to I / Q data (In-Phase / Quadrature-Phase signal), and then performed AM demodulation to extract the original information signal from the amplitude-modulated (AM) signal. The smartphone 24 was an iPhone® 12 manufactured by Apple Inc. In the evaluation experiment, a chirp sound and noise 80 were first input from a waveform generator 70 to the audio amplifier IC 10. To evaluate (a) above, the chirp sound coming from the speaker 9 was recorded without noise added to the comparative example and with white noise added to the example, and the SNR (Signal to Noise Ratio) was calculated for each. The chirp sound was recorded from a location approximately 15 cm away from the speaker 9 using the voice memo function of the smartphone 24.

[0038] (Evaluation Experiment Results) Figure 8 shows the evaluation results of the sound reproduced from the speaker. (1) shows a comparative example without white noise added, and (2) shows an example with white noise added. Comparing the comparative example without white noise shown in Figure 8(1) and the example with white noise added shown in Figure 8(2), it was found that both displayed similar spectrograms, and no significant difference was observed. Furthermore, the SNR was 28.4 dB for the comparative example and 28.8 dB for the example, with almost no difference observed. The bar graph on the right side of the graph, shown in shaded gray, indicates the sound pressure level, which is a quantity expressed as the common logarithm of the ratio of sound pressure to a reference value. For sounds in the audible range of the same frequency, the higher the sound pressure, the louder the sound is perceived. The unit is [dBA], which is corrected to take into account how sounds are heard in the human audible range. From the above, it was demonstrated that the use of the electromagnetic sound information leakage prevention device 1 does not impair the ease of hearing for the listener 31.

[0039] Regarding the evaluation of the effectiveness of the countermeasures (b) above, no noise was added to the comparative example, and white noise was added to the example, and electromagnetic waves leaked from the audio equipment were measured and audio was restored. Since the electromagnetic sound information leakage prevention device 1 aims to apply countermeasures to "leakage sources of electromagnetic waves containing sound information," measurements were made of electromagnetic waves leaked from a power line connected to an audio device in the vicinity of the audio device where the electromagnetic wave intensity can be observed without attenuation, i.e., the power line 12 connecting the audio amplifier IC 10 and the battery 11 shown in Figure 7. After measuring the signal with a current probe 21, signal processing similar to that of an actual attack was applied using a spectrum analyzer 23 via an amplifier 22, and the reproduced sound was restored.

[0040] FIG. 9 shows the evaluation results of sound restored from leaked electromagnetic waves, where (1) shows a comparative example without white noise added, and (2) shows an example with white noise added. Comparing the comparative example without white noise added shown in FIG. 9(1) with the example with white noise added shown in FIG. 9(2), it was confirmed that the restored sound of the example was more difficult to observe on a spectrogram than the comparative example. Furthermore, the SNR was 17.7 dB for the comparative example and 10.0 dB for the example, a 7.7 dB decrease. In the graph shown in FIG. 9(1), the line (α) represents the reproduced sound signal, and the line (β) represents the absolute value of the reproduced sound signal (with the frequency doubled). From the above, it was demonstrated that using the electromagnetic sound information leakage prevention device 1 makes it difficult to restore sound information via leaked electromagnetic waves.

[0041] FIG. 11 shows a functional block diagram of an electromagnetic sound information leakage prevention device according to a second embodiment. As shown in FIG. 11 , the electromagnetic sound information leakage prevention device 1a according to the second embodiment includes a reproduced sound input unit 2, a noise generating unit 8, a first synthetic sound signal generating unit 4, a second synthetic sound signal generating unit 5, and a synthetic sound signal output unit 6. That is, unlike the electromagnetic sound information leakage prevention device 1 according to the first embodiment, the electromagnetic sound information leakage prevention device 1a includes a noise generating unit 8 that generates noise internally in the device, instead of the noise input unit 3 that inputs noise from outside the device. Specifically, for example, noise is generated by processing within an audio amplifier IC, and a signal to which the noise has been added is generated. That is, since the reproduced sound input by the reproduced sound input unit 2 is transmitted to the audio amplifier IC as a digital signal, this can be achieved by adding a process for automatically adding noise to the digital signal processing. The other configurations are the same as those of the first embodiment.

[0042] FIG. 12 is a schematic diagram of an electromagnetic sound information leakage prevention device according to a second embodiment. As shown in FIG. 12 , an audio amplifier IC 10a functions as a playback sound input means 2, a first synthetic sound signal generating means 4, a second synthetic sound signal generating means 5, a synthetic sound signal output means 6, and a noise generating means 8. As in the first embodiment, a stereo amplifier IC used in known audio equipment can be used. A signal with noise added to the playback sound input to the audio amplifier IC 10a is generated by processing within the audio amplifier IC 10a, and the signal is output to the speaker 9. Specifically, a program included in the audio amplifier IC 10a uses the noise generated within the audio amplifier IC 10a to generate a synthetic sound signal with in-phase and anti-phase noise added to the playback sound input from the computer 13a via the DAC 17, and outputs the synthetic sound signal to the speaker 9. The remaining configuration is the same as in the first embodiment.

[0043] FIG. 13 shows a flow diagram of an electromagnetic sound information leakage prevention method according to a second embodiment. The electromagnetic sound information leakage prevention method according to the second embodiment is a method for preventing leakage of sound information via electromagnetic waves by a computer. As shown in FIG. 13 , first, a reproduced sound is input from a sound source 7 (step S11: reproduced sound input step). Next, noise 80 is generated by a noise generating means 8 included in the electromagnetic sound information leakage prevention device 1a (step S12: noise generation step). The first synthetic sound signal generating means 4 adds noise to the reproduced sound signal of the input reproduced sound to generate a first synthetic sound signal (step S13: first synthetic sound signal generation step). Furthermore, the second synthetic sound signal generating means 5 adds noise of an opposite phase to the noise generated in the first synthetic sound signal generation step to a reproduced sound signal in phase with the reproduced sound signal to generate a second synthetic sound signal (step S14: second synthetic sound signal generation step). Thereafter, the synthetic sound signal output means 6 simultaneously outputs the first synthetic sound signal and the second synthetic sound signal to the speaker 9 (step S15: synthetic sound signal output step).

[0044] FIG. 14 shows a functional block diagram of an electromagnetic sound information leakage prevention device according to a third embodiment. As shown in FIG. 14, the electromagnetic sound information leakage prevention device 1a according to the third embodiment differs from the device according to the second embodiment in that a synthetic sound signal is output to a first speaker 9a and a second speaker 9b. That is, the synthetic sound signal output means 6 outputs a first synthetic sound signal to the first speaker 9a and a second synthetic sound signal to the second speaker 9b, and these outputs are performed simultaneously. The other configurations are the same as those of the second embodiment. Even if separate synthetic sound signals are output to multiple speakers, the outputs are performed simultaneously. Therefore, in terms of ease of listening, out-of-phase noises cancel each other out before the signals reach the listener's ears, and clear audio can be delivered to the listener.

[0045] FIG. 15 is a schematic diagram of an electromagnetic sound information leakage prevention device according to a fourth embodiment. As shown in FIG. 15, the electromagnetic sound information leakage prevention device according to the fourth embodiment further includes a computer 13b connected to an audio amplifier IC 10b. Other configurations are the same as those of the second embodiment. The computer 13b functions as a first synthetic sound signal generating means 4, a second synthetic sound signal generating means 5, and a noise generating means 8. The audio amplifier IC 10b also functions as a reproduced sound input means 2 and a synthetic sound signal output means 6. The reproduced sound input from the computer 13a to the audio amplifier IC 10b via the DAC 17 is subjected to a noise-adding electromagnetic sound information leakage prevention process and output to the speaker 9. Specifically, a program included in the computer 13b uses the noise generated by the computer 13b to generate a synthetic sound signal by adding in-phase and out-of-phase noise to the reproduced sound input to the audio amplifier IC 10b, and the audio amplifier IC 10b outputs the synthetic sound signal to the speaker 9. According to this configuration, even if the audio amplifier IC 10b used does not have a function such as noise addition, by connecting the computer 13b, the electromagnetic sound information leakage prevention device 1a of Example 2 can be realized. Note that the computers 13a and 13b may be the same computer.

[0046] FIG. 16 is a functional block diagram of an electromagnetic sound information leakage prevention device according to a fifth embodiment. As shown in FIG. 16, an electromagnetic sound information leakage prevention device 1b includes a reproduced sound input means 2, a noise input means 3, a first synthetic sound signal generation means 4, a second synthetic sound signal generation means 5a, a third synthetic sound signal generation means 5b, and a synthetic sound signal output means 6. The reproduced sound input means 2 inputs a reproduced sound from a sound source 7. The noise input means 3 accepts input of noise 80. The first synthetic sound signal generation means 4 generates a first synthetic sound signal by adding noise 80 received by the noise input means 3 to a reproduced sound signal of the reproduced sound input to the reproduced sound input means 2. The second synthetic sound signal generation means 5a adds noise having a phase difference of 120° from the noise 80 received by the noise input means 3 to a reproduced sound signal that is in phase with the reproduced sound signal of the reproduced sound input to the reproduced sound input means 2, thereby generating a second synthetic sound signal. Furthermore, the third synthetic sound signal generation means 5b generates a third synthetic sound signal by adding noise having a phase difference of 240° with the noise 80 input by the noise input means 3 to a reproduction sound signal that is in phase with the reproduction sound signal of the reproduction sound input to the reproduction sound input means 2. The synthetic sound signal input means 6 outputs the first synthetic sound signal, the second synthetic sound signal, and the third synthetic sound signal to the speaker 9 simultaneously.

[0047] Unlike the example shown in FIG. 16 , a configuration may be adopted in which first to third speakers are provided and a synthetic sound signal is output to each speaker. Specifically, the synthetic sound signal output means 6 outputs the first synthetic sound signal to the first speaker, the second synthetic sound signal to the second speaker, and the third synthetic sound signal to the third speaker, and these outputs are performed simultaneously. Note that the distance from each speaker to the listener is approximately uniform. Even if separate synthetic sound signals are output to the three speakers, the outputs are performed simultaneously. Therefore, in terms of ease of listening, three types of noise with a phase shift of 120° cancel each other out when they reach the listener's ear, thereby delivering clear sound to the listener. Therefore, if the distance from each of the multiple speakers to the listener is sufficiently large and the speakers are close enough to each other that they can be considered to be speakers in essentially the same position, the noises will cancel each other out, and even in the case of four or more speakers, clear sound can be delivered to the listener.

[0048] FIG. 17 shows a functional block diagram of an electromagnetic sound information leakage prevention device according to a sixth embodiment. As shown in FIG. 17 , the electromagnetic sound information leakage prevention device 1c is configured by adding a detection unit 81 and a control unit 82 to the electromagnetic sound information leakage prevention device 1a according to the second embodiment. The detection unit 81 detects the level of electromagnetic radiation indicating leakage, and the control unit 82 automatically adjusts at least one of the amplitude and spectrum of the noise generated by the noise generation unit 8 based on the detection result by the detection unit 81. The detection unit 81 and the control unit 82 may be operated temporarily or intermittently as needed, or the detection unit 81 may continuously monitor the level of electromagnetic radiation indicating leakage, and the control unit 82 may dynamically select at least one of a plurality of noise control modes based on the monitoring result and automatically adjust at least one of the amplitude and spectrum of the noise generated by the noise generation unit. Alternatively, the detection unit 81 and the control unit 82 may be operated temporarily or intermittently as needed.

[0049] FIG. 18 shows a schematic diagram of an electromagnetic sound information leakage prevention device according to a sixth embodiment. The trust region 19 shown in FIG. 18 is an area where information leakage is not a problem and refers to a room or space that the user considers safe. Areas outside the trust region 19 are areas that the user cannot trust and represent areas where countermeasures should be taken. The audio amplifier IC 10 is an amplifier IC with built-in noise injection and signal synthesis functions. The power lines (12a, 12b) refer to cables extending from the audio amplifier IC 10 to the outside and are also used as communication cables. The power line 12a indicates a portion located within the trust region 19, while the power line 12b indicates a portion located outside the trust region 19. The current probe 21 is an antenna that picks up leaked information from signals flowing through the circuit, and the array antenna 16 is an antenna that captures signals leaking through space (air). The current probe 21 and the array antenna 16 are located within the trust region 19 and near the boundary with the outside of the trust region 19 to receive leaked signals. As a result, if the leakage signal is not received by the current probe 21 and the array antenna 16, it can be said that leakage outside the trust region 19 has been prevented. In this way, by using both the current probe 21 and the array antenna 16, more accurate measurements are possible.

[0050] The measurement and analysis device 14 includes a device that converts the received electrical signal from analog to digital and a device that performs signal processing to determine whether a leak exists. The conversion device and the leakage determination device may be separate devices. The conversion device may be a software-defined radio or amateur radio device, and the leakage determination device may be a computer equipped with a CPU. The network transmission path 15 is a path for transmitting the determination results of the measurement and analysis device to the computer 13c and may be either a wired or wireless connection. It may also be via the Internet or a local area network. The computer 13c may receive only the measurement data and determine whether a leak exists using its own analysis program. The computer 13c receives the leakage determination results or the measurement data and dynamically controls parameters such as the amplitude and frequency of the noise generation means 8. This configuration enables a highly energy-efficient leakage prevention device.

[0051] The present invention is useful as a technique for preventing leakage of sound information via electromagnetic waves.

[0052] REFERENCE SIGNS LIST 1, 1a to 1c Electromagnetic sound information leakage prevention device 2 Reproduced sound input means 3 Noise input means 4 First synthetic sound generation means 5, 5a Second synthetic sound generation means 5b Third synthetic sound generation means 6 Synthetic sound signal input means 7 Sound source 8 Noise generation means 9 Speaker 9a First speaker 9b Second speaker 10, 10a, 10b, 100 Audio amplifier IC 11 Battery 12, 12a, 12b Power line 13a to 13c Computer 14 Measurement analysis device 15 Network transmission path 16 Array antenna 17 DAC 18 Noise generator 19 Trust region 21 Current probe 22 Amplifier 23 Spectrum analyzer 24 Smartphone 31 Listener 32 Attacker (malicious third party) 40 Waveform 41 Reproduced sound 42 Leaked electromagnetic waves 70 Waveform generator 80, N １ ~N ３ Noise 81 Detection means 82 Control means A First synthesized sound signal B Second synthesized sound signal E １ ~E ３ Electromagnetic wave S １ ~S ３ Playback sound signal

Claims

1. A device for preventing leakage of sound information via electromagnetic waves, comprising: a playback sound input means for inputting playback sound from a sound source; a noise input means for accepting noise input; a first synthetic sound signal generation means for adding noise to a playback sound signal of the input playback sound to generate a first synthetic sound signal; a second synthetic sound signal generation means for adding noise of an opposite phase to a playback sound signal of the same phase as the noise to generate a second synthetic sound signal; and a synthetic sound signal output means for simultaneously outputting the first synthetic sound signal and the second synthetic sound signal to a speaker.

2. A device for preventing the leakage of sound information via electromagnetic waves, comprising: a playback sound input means for inputting playback sound from a sound source; a noise input means for accepting noise input; a first synthetic sound signal generation means for adding a first noise to a playback sound signal of the input playback sound to generate a first synthetic sound signal; a second synthetic sound signal generation means for adding a second noise having a phase difference of 120° from the first noise to a playback sound signal that is in phase with the playback sound signal to generate a second synthetic sound signal; a third synthetic sound signal generation means for adding a third noise having a phase difference of 240° from the first noise to a playback sound signal that is in phase with the playback sound signal to generate a third synthetic sound signal; and a synthetic sound signal output means for simultaneously outputting the first synthetic sound signal, the second synthetic sound signal, and the third synthetic sound signal to a speaker.

3. A device for preventing leakage of sound information via electromagnetic waves, comprising: a playback sound input means for inputting playback sound from a sound source; a noise input means for accepting input of noise; a first synthetic sound signal generation means for adding a first noise to a playback sound signal of the input playback sound to generate a first synthetic sound signal; an nth synthetic sound signal generation means for adding an nth noise having a phase difference of 360°×(n-1) / m (n is an integer of 2 or greater, and m is an integer of 3 or greater) from the first noise to a playback sound signal that is in phase with the playback sound signal to generate the nth synthetic sound signal; and a synthetic sound signal output means for repeating the nth synthetic sound signal generation means from n=2 up to m by incrementing by 1, and for simultaneously outputting the first to mth synthetic sound signals to a speaker.

4. The electromagnetic sound information leakage prevention device according to claim 1, characterized in that the first or second synthetic sound signal generating means is configured such that the noise signal to be added passes through an inversion stage of the output stage in the amplifier circuit, and the noise signal is in phase or out of phase between the output terminals.

5. The electromagnetic sound information leakage prevention device according to any one of claims 1 to 3, further comprising noise generating means for generating noise instead of the noise input means.

6. The electromagnetic sound information leakage prevention device according to claim 5, characterized in that the noise generating means generates noise whose amplitude or frequency components are dynamically changed according to the characteristics of the input audio information.

7. The electromagnetic sound information leakage prevention device according to claim 5, wherein said noise generating means generates wide-band white noise having frequency components within the audible range.

8. The electromagnetic sound information leakage prevention device according to claim 1, characterized in that the synthetic sound signal output means simultaneously outputs a first synthetic sound signal to a first speaker or a first group of speakers, and a second synthetic sound signal to a second speaker or a second group of speakers.

9. The electromagnetic sound information leakage prevention device according to claim 2, characterized in that the synthetic sound signal output means simultaneously outputs a first synthetic sound signal to a first speaker or a first group of speakers, a second synthetic sound signal to a second speaker or a second group of speakers, and a third synthetic sound signal to a third speaker or a third group of speakers.

10. An electromagnetic sound information leakage prevention device as described in claim 5, characterized in that it comprises a detection means for detecting the level of electromagnetic radiation indicative of leakage, and a control means for automatically adjusting at least one of the amplitude and spectrum of the noise generated by the noise generation means based on the detection results by the detection means.

11. The electromagnetic sound information leakage prevention device described in claim 10, characterized in that the detection means continuously monitors the level of electromagnetic radiation indicative of leakage, and the control means dynamically selects at least one from a plurality of noise control modes based on the monitoring results and automatically adjusts at least one of the amplitude and spectrum of the noise generated by the noise generation means.

12. A method for preventing leakage of sound information via electromagnetic waves, comprising having a computer perform the following steps: a playback sound input step of inputting playback sound from a sound source; a noise input step of accepting noise input; a first synthetic sound signal generation step of adding noise to a playback sound signal of the input playback sound to generate a first synthetic sound signal; a second synthetic sound signal generation step of adding noise of an opposite phase to a playback sound signal of the same phase as the noise to generate a second synthetic sound signal; and a synthetic sound signal output step of simultaneously outputting the first synthetic sound signal and the second synthetic sound signal to a speaker.

13. The method for preventing electromagnetic sound information leakage according to claim 12, further comprising a noise generating step for generating noise instead of the noise input step.

14. The method for preventing electromagnetic sound information leakage described in claim 13, characterized in that the noise generation step generates noise whose amplitude or frequency components are dynamically changed according to the characteristics of the input audio information, or generates wideband white noise having frequency components within the audible range.

15. The electromagnetic sound information leakage prevention method according to claim 12, characterized in that the synthetic sound signal output step simultaneously outputs a first synthetic sound signal to a first speaker or a first group of speakers, and a second synthetic sound signal to a second speaker or a second group of speakers.

16. A program for preventing information leakage due to electromagnetic sound, which causes a computer to execute each step of the method for preventing information leakage due to electromagnetic sound according to any one of claims 11 to 15.

Images