Anti-monitoring safe dialogue system

Abstract

The invention discloses an anti-monitoring safe dialogue system, which is applied to the field of communication, and aims to solve the problems that professional eavesdropping equipment cannot be effectively resisted in the prior art, and especially in a dynamically changing noise environment, the noise reduction and anti-eavesdropping capabilities are limited. According to the noise reduction earphone, environmental noise can be rapidly reduced without a learning process by predicting a noise rule, and especially for a rapidly changing noise source, an instant noise reduction capability which is not possessed by a traditional noise reduction earphone is provided, an eavesdropping behavior is effectively resisted, and the security of communication content is ensured.

Description

Technical Field

[0001] The present invention belongs to the field of communication technology, and in particular relates to a noise reduction and anti-eavesdropping technology. Background Art

[0002] In the field of modern communication technology, the confidentiality of audio communications is of vital importance. Especially in military communications, corporate high-level meetings and other conversations with high security requirements, the security of call content faces severe challenges from modern electronic devices such as recording devices and smartphones. These devices have high recording capabilities and can easily capture and record sensitive conversations, resulting in a sharp increase in the risk of information leakage.

[0003] To improve the security of audio communication, interference sources can be installed in special environments. With the help of special noise-canceling headphones, the two parties can achieve effective communication. However, people or recording devices without noise-canceling headphones cannot recognize useful audio information due to excessive noise. However, the design of traditional noise-canceling headphones mainly focuses on reducing environmental background noise and improving call quality through passive sound insulation or active noise reduction technology. These traditional noise reduction technologies mainly focus on reducing environmental background noise and lack the ability to adapt to rapidly changing noise. In a dynamic noise environment, the noise reduction effect of traditional headphones is limited and cannot effectively resist external noise interference.

[0004] The existing technologies mainly include passive and active noise reduction headphones. The passive type blocks the influence of noise pollution as much as possible through physical isolation or special structures on the headphones. The principle is to use the physical properties of passive devices to achieve noise reduction and sound insulation effects, such as silicone ear caps, leather head covers, etc., while the active type generates a reverse 180° phase sound wave after processing by the active noise cancellation (Active Noise Cancellation, ANC) chip, which is transmitted to the human ear canal through the headphone speaker, creating a phase difference of 180° with the noise, offsetting the environmental noise that the human ear should hear, thereby achieving the purpose of noise reduction. The structural diagram is shown in the figure below. Figure 1 shown.

[0005] The corresponding parameters of active noise-cancelling headphones are often fixed, so they cannot perform rapid adaptive adjustments and phase control according to changes in the external environment. Therefore, their active noise-cancelling performance will be affected by external factors and become unstable, and they will only have a good noise-cancelling effect for some stable noises.

[0006] Passive noise-cancelling headphones mainly target high-frequency noise pollution, and are not very effective for noise pollution below 800Hz and lower frequency bands. Summary of the invention

[0007] In order to solve the above technical problems, the present invention proposes a secure conversation system that prevents eavesdropping. For pseudo-random dynamic interference sources, it can not only effectively reduce in-band interference and improve call quality, but also effectively interfere with illegal recording equipment by introducing controllable dynamic noise sources, thereby preventing the call content from being illegally recorded or eavesdropped.

[0008] The technical solution adopted by the present invention is: a secure conversation system for preventing eavesdropping, which is intended to provide a highly secure environment for communication. The system includes an external dynamic noise source and a headset used in conjunction with the external dynamic noise source. The external dynamic noise source is set in a conversation environment, such as a conference room, and is used to generate dynamic noise during the conversation to interfere with potential eavesdropping. The headset is designed to use the same pseudo-random number seed as the external dynamic noise source to achieve accurate noise cancellation.

[0009] The external dynamic noise source is a core innovation of the present invention. It can generate dynamic noise based on a preset pseudo-random number seed. This noise is designed to simulate and interfere with possible eavesdropping. If the eavesdropper cannot obtain the preset pseudo-random number seed, it will not be able to eliminate the dynamic noise in the recorded signal, thereby preventing illegal recording at the source. At the same time, the high-sensitivity microphone is responsible for capturing all external noises including dynamic noise and environmental noise in real time, as well as the dialogue sound signals of the dialogue scene.

[0010] The headset includes: a power module, a microphone, an external dynamic noise source, a dynamic noise cancellation speaker, an environmental noise cancellation speaker, an ANC adaptive filter and an audio communication speaker; the power module supplies power to the microphone, the external dynamic noise source, the dynamic noise cancellation speaker, the environmental noise cancellation speaker, the ANC adaptive filter and the audio communication speaker.

[0011] The dynamic noise cancellation speaker uses the same pseudo-random number seed as the external dynamic noise source to generate a reverse signal corresponding to the current dynamic noise in real time, accurately eliminating dynamic noise. This process does not require the learning phase required by traditional noise-cancelling headphones to achieve instant noise reduction. In addition, the ANC adaptive filter uses spectral subtraction to generate cancellation signals for ambient noise, which are then input into the ambient noise cancellation speaker to eliminate background noise and further purify the call environment.

[0012] Finally, the audio communication speaker is responsible for outputting a pure sound signal after double noise reduction processing, ensuring the clarity and naturalness of communication. The noise reduction headphone system of the present invention, through this unique component collaborative working mechanism, not only significantly improves the adaptability and noise reduction efficiency to rapidly changing noise, but also provides an unprecedented level of protection for communication security, especially in communication occasions that require high confidentiality.

[0013] Beneficial effects of the present invention: Compared with the prior art, the anti-eavesdropping secure dialogue system of the present invention can effectively prevent intentional eavesdropping while maintaining the clarity and naturalness of audio communications.

[0014] The design of this invention focuses on developing a system that enables noise-canceling headphones to work in conjunction with an external noise source to generate reverse dynamic noise based on a pseudo-random number seed to adapt to different communication environments. The design of this dynamic noise source can effectively interfere with potential eavesdropping without affecting normal communication, providing necessary protection for high-security communications.

[0015] In the application scenario, each participant wears a headset of the present invention, which can be synchronized with an external dynamic noise source to generate a reverse dynamic noise signal to eliminate the interference generated by the noise source in real time. Unlike traditional noise reduction headsets, the headset of the present invention can quickly adapt to the changes of the dynamic noise source after matching with the external dynamic noise source, and can maintain clear communication quality even in an environment where the noise changes frequently.

[0016] Through the implementation of the present invention, users can obtain a higher level of communication security while enjoying clear sound quality, effectively preventing information leakage and eavesdropping risks, and meeting the needs of military, business and other occasions with special requirements for communication confidentiality. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 This is a schematic diagram of the structure of active noise reduction headphones.

[0018] Figure 2 It is the system application scenario graph.

[0019] Figure 3 This is the system structure module diagram.

[0020] Figure 4 Schematic diagram of dynamic noise signal and cancellation signal.

[0021] Figure 5 The following are the mixed images of dynamic noise, band noise, dynamic reverse signal and band noise.

[0022] Figure 6 This is the principle diagram of spectral subtraction.

[0023] Figure 7 This is a timing comparison chart for low-frequency noise elimination.

[0024] Figure 8 This is a comparison chart of high-frequency noise elimination timing.

[0025] Fig. 9 This is a working diagram for noise-canceling headphones. DETAILED DESCRIPTION

[0026] To facilitate those skilled in the art to understand the technical content of the present invention, the present invention is further explained below with reference to the accompanying drawings.

[0027] like Figure 2 The system application scenario diagram shown reveals the layout and operation of the present invention in an actual conversation environment. In a sensitive conversation environment that may be eavesdropped, such as a conference room or a business negotiation area, the present invention creates a protective audio barrier by deploying an external dynamic noise source. The noise source generates unpredictable dynamic noise based on a preset pseudo-random number seed, which is sufficient to interfere with potential monitoring devices, making it impossible for unauthorized personnel to hear the conversation content.

[0028] In this configuration, each authorized conversation participant wears a specially made noise-canceling headset (hereinafter referred to as the "invented headset"). By sharing the same pseudo-random number seed with the external noise source, these headsets can accurately predict and generate signals that are opposite to the dynamic noise emitted by the noise source. In this way, the headset can not only eliminate the interference generated by the noise source in real time, but also effectively reduce other background noise to ensure the clarity and privacy of the conversation.

[0029] Figure 3 The system structure module diagram of the present invention is shown, revealing the noise source and the headphones and the internal components and their collaborative working methods. The core innovation of the headphones lies in its dynamic noise cancellation speaker, which is specially designed to eliminate the noise generated by the external dynamic noise source. The noise source generates unpredictable dynamic noise based on a preset pseudo-random number seed, which is designed to interfere with potential eavesdropping.

[0030] During the communication between the headset and the noise source, the noise source sends a pseudo-random number seed to the headset through the MCU. Before the noise source is activated, it sends a control command to the headset to notify the upcoming noise pattern change. The headset adjusts its working state in advance to ensure that it can respond to noise changes in a timely manner. In some cases, the noise source may suspend noise generation when switching noise channels and only send control commands. After the headset receives these commands and responds, the noise source starts to generate noise.

[0031] In addition, the earphones also contain an ANC adaptive filter to generate cancellation signals for ambient noise. These signals are then input into the ambient noise cancellation speakers to eliminate background noise and further purify the call environment. However, compared with the dynamic noise cancellation process, the role of the ANC adaptive filter is auxiliary and mainly targets more stable ambient noise.

[0032] The noise-cancelling headphones of the present invention break through the limitations of traditional ANC technology. Through a pre-set pseudo-random number seed, they can instantly generate a cancellation signal that is completely opposite to the dynamic noise source, and can achieve efficient noise reduction without a learning process. This fast response mechanism enables the headphones to instantly adjust the noise reduction strategy when the noise source changes, effectively reducing the impact of fast-changing noise on communication quality; the noise-cancelling headphones generate a reverse cancellation signal through the noise-cancelling speaker, such as Figure 4 As shown:

[0033] The pseudo-random number seed for generating the dynamic noise signal is the same as that for the elimination signal. The noise-canceling headphones generate a reverse 180° phase sound wave, which is transmitted to the human ear canal through the headphone speakers, creating a 180° phase difference with the noise, thus canceling out the dynamic noise that the human ear should have heard, thereby achieving the purpose of noise reduction.

[0034] like Figure 5 As shown, a schematic diagram of a dynamic noise signal, a noisy signal, and a mixture of a noisy signal and a dynamic reverse signal is given. It can be clearly seen that only environmental noise remains after mixing.

[0035] In the present invention, a classic speech enhancement technology, spectral subtraction, is used to eliminate noise. This technology can significantly improve the clarity of speech when processing additive mixtures of speech signals and noise. Spectral subtraction is based on a core assumption: speech signals and noise are statistically independent of each other, and noise has stable statistical characteristics. Using this assumption, the present invention can process noisy speech signals through short-time Fourier transform (STFT), first separating the spectral components of the noise from the signal, and then obtaining a clearer spectral representation of the speech signal by subtracting these noise components.

[0036] After completing the processing in the frequency domain, the present invention performs an inverse Fourier transform on the processed signal and converts it back to the time domain. This process enables the present invention to effectively restore the enhanced speech signal while maintaining the naturalness and clarity of the speech signal. In the present invention, special emphasis is placed on the statistical stationarity of the noise, which allows a stable noise model to be used to estimate the spectrum of the noise during a call. This estimation method is based on a mean algorithm, which assumes that the expected value of the amplitude spectrum of the noise remains unchanged throughout the call, whether there is voice activity or silence intervals.

[0037] In addition, the method of the present invention is also based on the short-term stationarity assumption of the speech signal, which means that in a short period of time, the statistical characteristics of the speech signal can be regarded as constant. In this way, the present invention can accurately estimate the short-time spectrum of the clean speech in the noisy speech signal, which is achieved by subtracting the short-time spectrum of the noise from the short-time spectrum of the noisy speech signal. In the mathematical model, if the time domain sampling signal of the clean speech is represented as s(n) and the time domain sampling signal of the noise is represented as d(n), then the time domain signal y(n) of the noisy speech can be represented as a linear combination of s(n) and d(n).

[0038] y(n)=s(n)+d(n),0≤n≤N-1 (1)

[0039] In the present embodiment, a pure speech signal is defined as s(n), and a stable additive noise signal is represented as d(n). In order to reduce the truncation effect introduced by the framing operation in signal processing, the present embodiment implements a windowing technique to the noisy speech signal y(n). Through this step, the present invention can process the edge of the signal more smoothly, thereby reducing the distortion in the processing process.

[0040] In the process of Fourier transform, the transformation result of the noisy speech signal y(n) is expressed as Y K , the transformation result of the pure speech signal s(n) is expressed as S K , and the transformation result of the noise signal d(n) is expressed as D K .

[0041] Figure 6 In the example, after y(n) is transformed by FFT, we have Y K =S K +D K , from which we can get:

[0042] |Y K | 2 =|S K | 2 +|D K | 2 +S K D K * +D K S K * (2)

[0043] For the noise signal D K and the speech signal S K , their conjugates are represented by D K * and S K *Since the speech signal s(n) and the noise signal d(n) are assumed to be statistically independent, this means that there is no correlation between them. This independence further means that the autocorrelation function S of the speech signal is K and the autocorrelation function Y of the noisy signal K are also independent of each other. In addition, the statistical properties of the noise signal d(n), such as its mean being zero and its Gaussian distribution, ensure that its autocorrelation function D K It also has the Gaussian distribution characteristics of zero mean, so:

[0044] E||Y K | 2 |=E||S K | 2 |+E||D K | 2 | (3)

[0045] For a short-term stationary process within an analysis frame, we have:

[0046] |Y K | 2 =|S K | 2 +λ n (k) (4)

[0047] λ n (k) is when there is no voice |D K | 2 The statistical average of the original speech is as follows: is the amplitude of the enhanced speech signal.

[0048] |S K | 2 =|Y K | 2 -λ n (k) (5)

[0049]

[0050]

[0051] definition and the posterior signal-to-noise ratio γ K =|Y K | 2 / λ n (k)

[0052]

[0053] G K =(1-1 / γ K ) 1 / 2(9)

[0054] This method involves decomposing the speech signal into its spectral components and carefully processing each component. During the processing, the attenuation coefficient is adjusted according to the presence of speech and the signal-to-noise ratio. When the signal-to-noise ratio of the speech signal is high, a smaller attenuation coefficient is used to retain more speech information; while in areas with low signal-to-noise ratio, the attenuation coefficient is increased to reduce the impact of noise. The power spectrum of the clean speech signal is obtained by calculating the power spectrum of the noisy speech and the noise, and subtracting the power spectrum of the noise from the power spectrum of the noisy speech. Then, by taking the square root of this power spectrum, the amplitude spectrum of the speech signal is obtained. Finally, these spectral components are resynthesized into a time domain signal through an inverse Fourier transform, resulting in an enhanced speech signal.

[0055] like Figure 7 and Figure 8 As shown in the figure, the noise reduction earphones of the present invention (hereinafter referred to as "the earphones of the invention") show significant technical advantages in processing dynamic noise, especially in low-frequency and high-frequency noise environments. These timing diagrams show in detail the noise processing strategies and effects of the traditional noise reduction earphones (hereinafter referred to as "the traditional earphones") and the earphones of the present invention at different frequencies.

[0056] Figure 7 The figure shows a time series comparison of the noise reduction processing of the noise reduction headphones of the present invention (hereinafter referred to as "the invented headphones") and the traditional noise reduction headphones (hereinafter referred to as "the traditional headphones") in a low-frequency dynamic noise environment. In the figure, the noise source generates noise 1 to noise n in sequence, and each noise event is represented by a digital mark. Traditional headphones require a certain learning time (represented by the red area) at the beginning to adapt to the noise pattern, and the headphones cannot provide effective noise reduction during this period of time. In contrast, the invented headphones use a pseudo-random number seed synchronized with an external noise source to directly send out a reverse cancellation signal without a learning phase, thereby achieving instant noise reduction. This significantly improves the noise reduction efficiency and ensures the clarity and naturalness of communication.

[0057] Figure 8 The case of a high-frequency dynamic noise environment is demonstrated, where the noise source changes much faster than traditional headphones can handle. In this case, by the time the traditional headphones try to learn and adapt to the current noise pattern, the noise has already changed, making it unable to effectively perform noise reduction.

[0058] In contrast, the Invention headphones, which do not require a learning phase, can continuously and quickly predict and eliminate high-frequency dynamic noise. This means that no matter how quickly the noise changes, the Invention headphones can maintain efficient noise reduction performance and provide users with a stable and clear listening environment.

[0059] The noise-canceling headphones of the present invention show excellent performance in both low-frequency and high-frequency dynamic noise environments. By adopting a pre-pairing strategy, the headphones of the present invention not only improve the noise reduction efficiency, but also ensure a clear auditory experience under dynamic noise conditions. These advantages make the headphones of the present invention a better choice in environments that require instant, clear and secure communication. Compared with traditional noise-canceling headphones, the present invention provides an innovative solution to meet the strict requirements of modern communication technology for high confidentiality and sound quality.

[0060] The anti-eavesdropping secure dialogue system of the present invention is designed as a highly complex and accurate audio management device, and its workflow starts with the activation of an external noise source. Fig. 9 As shown:

[0061] The external noise source generates dynamic noise based on a pseudo-random number seed. This noise is specially designed to interfere with illegal recording equipment and prevent the leakage of sensitive call content. When the headset is started, it is first paired with the external noise source. The noise source sends the pseudo-random number seed of this dynamic noise to the headset through the MCU. This signal also contains a clock synchronization signal. After the headset receives the signal from the noise source through its internal MCU, it will reply with a confirmation signal and send a cancellation signal synchronously under clock control. At this time, the pairing is considered successful, and the same pseudo-random number seed is used to ensure that the internal sound source of the headset can generate a cancellation signal that is completely opposite to the external noise.

[0062] After pairing is complete, the microphone inside the headset begins to collect ambient noise. These collected ambient noise signals are then mixed with the cancellation signals generated inside the headset in the signal processing unit. The mixing process is the key to the noise reduction effect and requires precise time synchronization and phase matching.

[0063] Next, the signal processing unit uses advanced adaptive filtering technology, especially spectral subtraction, to process the mixed signal. Spectral subtraction decomposes the signal into spectral components through short-time Fourier transform and separates noise and speech in the frequency domain. Since noise has statistical stationarity and speech signals have short-time stationarity, spectral subtraction can effectively subtract the power spectrum of noise from the noisy speech signal to estimate the pure speech power spectrum. Then, by taking the square root of the estimated pure speech power spectrum and performing an inverse Fourier transform, the signal is restored to the time domain to obtain the enhanced speech signal.

[0064] The adaptive filtering algorithm dynamically adjusts the cancellation signal based on the noise characteristics detected in real time, ensuring the best noise reduction effect even when the noise environment changes. The optimized signal is played through the headphone speakers to cancel out the noise in the environment, achieving active noise cancellation and outputting clear audio to the user.

[0065] During the entire process, the system continuously performs sound input and noise detection, and adjusts the noise reduction strategy in real time to ensure the clarity of sound quality and the security of communication. This is a collaborative working mechanism between internal and external sound sources.

[0066] The key to the noise-cancelling headphones of the present invention being able to significantly reduce in-band interference and effectively resist eavesdropping lies in their unique system design and collaborative working mechanism. Unlike the prior art that only uses spectral subtraction to enhance speech, the present invention adopts an innovative collaborative elimination strategy of dynamic noise sources and environmental noise. This strategy not only uses spectral subtraction to remove environmental noise, but also introduces an innovative dynamic noise generation and elimination mechanism, which is not seen in existing solutions.

[0067] Specifically, the dynamic noise source of the present invention can generate a matching noise signal based on a pseudo-random number seed to match the known dynamic noise source. This mechanism enables the headset to instantly adjust the noise reduction strategy when the known noise source changes, effectively reducing the impact of rapidly changing noise on communication quality.

[0068] Therefore, the noise reduction effect achieved by the noise reduction headphones of the present invention, through innovative system design, realizes the utilization and elimination of dynamic noise, which not only improves the noise reduction efficiency, but also enhances the security of communication, especially when countering professional eavesdropping equipment, showing significant technical advantages.

[0069] The creativity of the present invention lies not only in its comprehensive application of existing technologies, but also in its innovative anti-eavesdropping strategy using dynamic noise sources. The present invention is not only innovative in its solution, but also shows significant practical value and market potential in practical applications, especially in the military, business and other fields that require high-security communications.

[0070] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the principles of the present invention, and should be understood that the scope of protection of the present invention is not limited to such specific statements and embodiments. For those skilled in the art, the present invention may have various changes and variations. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of the claims of the present invention.

Claims

1. A secure conversation system for preventing wiretapping, characterized in that: The conversation environment of the application includes several conversation partners and a dynamic noise source set in the conversation environment. Each conversation partner wears headphones. Before starting a formal conversation, the dynamic noise source is matched with the headphones in the conversation environment. After the match is successful, the dynamic noise source and the headphones share a pseudo-random number seed. Once a conversation begins, the earphones cancel dynamic noise by generating a signal that is opposite to the source of the dynamic noise.

2. The anti-eavesdropping secure dialogue system according to claim 1, characterized in that: The sound signals collected by the earphone include conversation audio signals and external noise, and the external noise includes dynamic noise and environmental noise.

3. The anti-eavesdropping secure dialogue system according to claim 2, characterized in that: The headset comprises: a power module, a microphone, a dynamic noise cancellation speaker, an ambient noise cancellation speaker, an ANC adaptive filter and an audio communication speaker; the power module supplies power to the microphone, the dynamic noise cancellation speaker, the ambient noise cancellation speaker, the ANC adaptive filter and the audio communication speaker; The microphone is used to collect sound signals; The dynamic noise cancellation speaker shares a pseudo-random number seed with an external noise source, generating a signal opposite to the dynamic noise source in real time; The ANC adaptive filter uses spectral subtraction to generate a cancellation signal that is opposite to the ambient noise, and transmits the cancellation signal to the ambient noise cancellation speaker to eliminate the ambient noise; thus obtaining a clean conversation audio signal; The audio communication speaker is used to output a clean conversation audio signal in the headset.

4. The anti-eavesdropping secure dialogue system according to claim 3, characterized in that: After the conversation begins, the sound signal collected by the earphone is a mixed signal of three independent components: the conversation audio signal, dynamic noise, and environmental noise.

5. The anti-eavesdropping secure dialogue system according to claim 4, characterized in that: The timing of dynamic noise elimination is: after the dynamic noise source is successfully matched with the headphones in the conversation environment, the dynamic noise elimination speaker generates a signal opposite to the dynamic noise source in real time to eliminate the dynamic noise collected by the headphones.

6. The anti-eavesdropping secure dialogue system according to claim 5, characterized in that: The specific process of eliminating environmental noise is: Before the conversation begins, the earphones collect external noise, and the dynamic noise cancellation speaker generates a signal opposite to the dynamic noise source in real time; the dynamic noise collected by the earphones is eliminated; separate environmental noise is obtained; and the power spectrum of the environmental noise is calculated; After the conversation begins, the power spectrum of the conversation audio signal with ambient noise is calculated; Subtracting the power spectrum of the ambient noise from the power spectrum of the conversation audio signal with the ambient noise, thereby obtaining the power spectrum of the pure conversation audio signal; By taking the square root of the power spectrum of this pure conversation audio signal, the amplitude spectrum of the conversation audio signal is obtained; Finally, the enhanced dialogue audio signal is obtained through inverse Fourier transform.

Images