Audio recording interference method and system based on random frequency hopping and resonant coupling

Abstract

The present application belongs to the technical field of recording interference, and particularly relates to a recording interference method and system based on random frequency hopping and resonance coupling, which comprises the following steps: collecting a spatial acoustic response signal for spectral analysis, calculating the product of the peak amplitude and the quality factor to obtain a resonance strength index for screening a resonance frequency band; converting the normalized average resonance strength into a shift register initial state, performing nonlinear cyclic shift on the frequency hopping index by using Hamming weight, and generating a frequency hopping sequence coupled with the environmental resonance characteristics; taking the square root of the quality factor as a bias coefficient, non-uniformly biasing the frequency hopping points to the peak frequency points, and enabling energy concentration in the resonance zone; and adaptively adjusting the transition time of the raised cosine envelope according to the number of frequency bands spanned by adjacent frequency hopping points, to realize smooth switching of the interference signal. The present application can solve the problems of low energy utilization rate of traditional interference, fixed pseudo-random sequence structure and frequency point switching transient impact.

Description

Technical Field

[0001] This invention relates to the field of audio recording interference technology. More specifically, this invention relates to audio recording interference methods and systems based on random frequency hopping and resonant coupling. Background Technology

[0002] In sensitive scenarios such as confidential business meetings, government negotiations, and private conversations, anti-recording interference technology is a crucial means of ensuring the security of voice information. Current acoustic recording interference methods primarily employ the transmission of broadband white noise, conventional frequency sweep signals, or ultrasound to suppress normal speech through acoustic masking effects, or to utilize nonlinear effects to influence the microphone array and automatic gain control module of the recording equipment. In practical applications, the acoustic responses of different spatial environments and transmitting hardware vary significantly, and traditional interference methods typically exhibit indiscriminate coverage characteristics.

[0003] This type of indiscriminate interference across the entire frequency band or wideband has low energy utilization and can easily lead to increased power consumption of equipment. In order to achieve the shielding effect, it is often necessary to increase the transmission power, which in turn creates harsh background noise in the environment and affects the normal communication experience of people in the same space.

[0004] Referring to Chinese patent document CN105306162B, a mixing-based voice jamming device and its implementation method are disclosed. This technology includes a baseband control module, a baseband generation module, a cryptographic application module, a signal shaping and amplification module, and a transmission module. It further includes: a baseband control chip connected to a low-frequency DDS generator and a low-frequency amplifier; a baseband control chip connected to a high-frequency surface acoustic wave resonator and a high-frequency amplifier; a baseband control chip connected to a random number generation unit; a baseband control chip connected to a cryptographic unit; the low-frequency amplifier and the high-frequency amplifier respectively connected to a filtering and shaping unit; the filtering and shaping unit sequentially connected to a power secondary amplifier and an impedance gain regulator; and the impedance gain regulator connected to a low-frequency transmitter and a high-frequency transmitter respectively. This technical solution transmits the signal through a low-frequency transmitter or a high-frequency transmitter, creating mixing interference to disrupt voice equipment.

[0005] However, in frequency hopping sequence generation, existing technologies mostly employ traditional linear feedback shift registers, which have limited pseudo-randomness and fixed sequence structures. The generation logic is independent of the acoustic environment, making it difficult to deeply correlate with the real-time characteristics of environmental resonance. Regarding frequency hopping point allocation, traditional schemes typically use uniform distribution or simple random mapping, making it difficult to perform non-uniform energy bias allocation for resonant frequency bands with different bandwidths and quality factors, resulting in insufficient utilization of resonant peak frequencies. In the interference frequency switching stage, traditional frequency hopping signals usually employ hard switching or fixed-window smoothing, without considering the impact of the frequency hopping span on the transition state. Large-span jumps across different resonant frequency bands, if using a uniform switching duration, are prone to transient impulse noise and spectral leakage, which not only disrupts the acoustic smoothness of the interference signal but may also reduce the concealment of the interference process. Summary of the Invention

[0006] To address the technical problems of low energy utilization caused by blind coverage of interference signals and the ease with which the linearity of frequency hopping sequences can be cracked in existing technologies, this invention provides solutions in the following aspects.

[0007] In a first aspect, the present invention provides a recording interference method based on random frequency hopping and resonant coupling, comprising: S1, acquiring spatial acoustic response signals, performing spectral analysis to extract candidate resonant frequency bands and their peak frequency points, calculating the product of the peak amplitude and quality factor of the candidate resonant frequency bands to obtain a resonant intensity index, and retaining candidate resonant frequency bands whose resonant intensity index is greater than a preset threshold to generate a set of resonant frequency bands; S2. Configure a shift register based on the number of candidate frequency bands included in the resonant frequency band set and the preset total number of in-band representation intervals; convert the calculated normalized average resonant intensity index of the resonant frequency band set into a binary sequence and write it into the shift register; perform linear feedback shift based on the primitive feedback polynomial to continuously generate an initial frequency hopping index; count the Hamming weight of the current state of the shift register at a preset period, and perform a cyclic shift of the current initial frequency hopping index with a bit length equal to the Hamming weight to obtain a nonlinear frequency hopping sequence; S3. The nonlinear frequency hopping sequence is allocated to the resonant frequency band set through modulo operation to extract the frequency hopping points. The square root of the quality factor is used as the bias coefficient to bias the frequency hopping points toward the peak frequency point to generate an interference signal. The raised cosine envelope transition time is adjusted according to the number of frequency bands crossed by adjacent frequency hopping points and the interference signal is output. The interference signal is then output after electroacoustic conversion.

[0008] This invention extracts the acoustic response and calculates the product of peak amplitude and quality factor to obtain the resonance intensity index, locking the inherent resonance region in space and avoiding the energy waste caused by traditional blind frequency sweeping. It transforms the environmental resonance characteristics into the initial state of a shift register and dynamically performs cyclic shifting of the frequency hopping index according to Hamming weight, breaking the rule that conventional pseudo-random sequences have an algebraic linear structure, and constructing a nonlinear interference sequence that is extremely difficult to filter out by noise reduction algorithms. It uses the square root of the quality factor to non-uniformly offset the frequency hopping point to the resonance peak, greatly amplifying the effective interference sound pressure. On this basis, it adaptively adjusts the raised cosine envelope transition time according to the number of frequency bands crossed, effectively smoothing amplitude abrupt changes and suppressing out-of-band spectral leakage during cross-band transitions, and significantly shortening the delay response during in-band switching, achieving a unity of efficient utilization of interference energy and smooth acoustic output.

[0009] Preferably, the step of collecting the spatial acoustic response signal and performing spectral analysis to extract candidate resonant frequency bands and their peak frequencies includes: windowing the collected spatial acoustic response signal and converting it to the frequency domain using a fast Fourier transform to obtain an amplitude spectrum; extracting the frequency points corresponding to the local maxima in the amplitude spectrum as amplitude peak frequencies; arranging the amplitude peak frequencies in descending order according to their corresponding amplitudes, and extracting a preset number of the top frequency points to divide them into the candidate resonant frequency bands.

[0010] This invention improves the efficiency of environmental spatial resonance feature extraction by rapidly removing invalid frequency band noise through descending order extraction and local maximum filtering mechanisms.

[0011] Preferably, the step of calculating the product of the peak amplitude and the quality factor of the candidate resonant frequency band to obtain the resonant intensity index includes: determining the frequency corresponding to the maximum amplitude within the candidate resonant frequency band as the peak frequency point and obtaining the corresponding peak amplitude; searching for two frequency points where the amplitude drops to a preset proportion of the peak amplitude and subtracting them to obtain the peak bandwidth; dividing the peak frequency point by the peak bandwidth to obtain the quality factor, and multiplying the quality factor by the peak amplitude to obtain the resonant intensity index.

[0012] This invention combines peak bandwidth with refined calculation of the quality factor, providing an accurate mathematical dimension for evaluating the energy aggregation capability of each frequency band.

[0013] Preferably, converting the calculated normalized average resonant intensity index of the resonant frequency band set into a binary sequence and writing it into the shift register to generate an initial frequency hopping index includes: obtaining the average and maximum values ​​of the resonant intensity index within the resonant frequency band set; dividing the average value by the maximum value to obtain a normalized intensity value; multiplying the normalized intensity value by the full-scale unsigned integer limit value and rounding down to obtain a decimal representation value; and converting the decimal representation value into a binary bit string sequence with a length matching the order of the shift register.

[0014] This invention standardizes the mapping path between decimal intensity characteristics and the binary state of the underlying hardware, achieving seamless integration of environmental acoustic physical quantities into logic control parameters.

[0015] Preferably, after converting the decimal representation value into a binary bit string sequence with a length matching the order of the shift register, the method further includes: when the binary bit string sequence is determined to be an all-zero sequence, triggering state intervention logic to set the least significant bit of the binary bit string sequence to 1 to avoid the register falling into an all-zero deadlock state.

[0016] This invention eliminates the risk of abnormal shutdown of shift registers from the hardware security logic level, ensuring the absolute continuity of nonlinear frequency hopping sequence generation.

[0017] Preferably, the step of counting the Hamming weight of the current state of the shift register according to a preset period and performing a cyclic shift on the current initial frequency hopping index includes: setting a counter to count the number of generated frequency hopping points, and triggering the cyclic shift when the counter reaches the frequency hopping interval period determined by the ratio of the highest to the lowest resonance intensity index; reading the state word stored in the shift register at the trigger time, and counting the number of binary digits 1 contained in the state word as the Hamming weight using a bit counting instruction; performing a bit-by-bit cyclic right shift operation on the current initial frequency hopping index, wherein the number of bits shifted cyclically is equal to the Hamming weight, and synchronously resetting the counter to zero after the output is completed.

[0018] This invention introduces an adaptive nonlinear perturbation variable that is strongly coupled with the sequence's own running state, increasing the computational complexity of eavesdropping devices in reverse-engineering interference signals.

[0019] Preferably, the interference signal is generated by biasing the frequency hopping point towards the peak frequency point using the square root of the quality factor as a bias coefficient, including: calculating a first frequency difference between the high-frequency cutoff boundary of the candidate resonant frequency band and the peak frequency point, and a second frequency difference between the peak frequency point and the low-frequency cutoff boundary; dividing the first frequency difference and the second frequency difference into a predetermined number of representation intervals, wherein the frequency step size of the representation intervals increases in a geometric progression, and the common ratio of the geometric progression is the sum of 1 and the bias coefficient; performing the modulo operation on the nonlinear frequency hopping sequence to obtain the in-band interval number, mapping the in-band interval number to the corresponding representation interval, and using the center frequency within the representation interval as the frequency hopping point.

[0020] This invention utilizes a geometric series step size to construct a geometric series non-uniform mapping, which forces the generated physical frequency hopping points to naturally concentrate in the midpoint region of the high resonance response, thereby improving the effective energy conversion rate.

[0021] Preferably, adjusting the raised cosine envelope transition time based on the number of frequency bands traversed by adjacent frequency hopping points includes: calculating the absolute value of the difference between the resonant frequency bands of the previous and next frequency hopping points in the frequency sequence number arranged from low to high peak frequency points, as the number of frequency bands traversed; when the number of frequency bands traversed is greater than 0, multiplying the number of frequency bands traversed by the basic time constant to obtain the raised cosine envelope transition time, and limiting it to not exceed the fixed total dwell time threshold of a single frequency point allocated by the system; when the number of frequency bands traversed is 0, setting the raised cosine envelope transition time to a preset minimum transition time to shorten the same-band switching delay.

[0022] This invention adjusts the shunting time based on the actual physical span of the frequency hopping space, balancing the contradiction between cross-frequency leakage prevention smoothness and high-speed dwell response at the same frequency.

[0023] Preferably, outputting the interference signal includes: inputting the interference signal into a digital-to-analog converter to convert it into an analog signal; enhancing the voltage and current driving capability of the analog signal through an audio power amplifier circuit; and feeding the amplified analog signal to a speaker array to convert it into an acoustic interference wave and transmitting it to the target space.

[0024] Secondly, the present invention provides a recording interference system based on random frequency hopping and resonant coupling, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned recording interference method based on random frequency hopping and resonant coupling is implemented.

[0025] By adopting the above technical solution, the recording interference method based on random frequency hopping and resonant coupling is generated into a computer program and stored in a memory for loading and execution by a processor. This allows for the creation of a terminal device based on the memory and processor, facilitating its use.

[0026] The beneficial effects of this invention are as follows: This invention extracts the quality factor and resonance intensity index through environmental acoustic response, enabling the perception of environmental resonance bands. This perception data is then integrated into all core processes, including seed generation in hardware registers, Hamming weight perturbation displacement, and non-uniform frequency offset mapping. This achieves extremely high acoustic masking effect under low transmit power and effectively avoids the inherent algebraic linearity weakness of conventional frequency hopping signals, thus constructing a secure defense against advanced speech separation and noise reduction processing.

[0027] Furthermore, a span-aware raised cosine weighted smoothing control mechanism is introduced, which uses the number of frequency bands crossed as the reference coordinate to intelligently allocate the envelope transition time. This successfully eliminates the harsh auditory noise caused by sudden changes in the interference signal and the out-of-band spectrum leakage. While significantly improving the concealment of interference and the interception rate of eavesdropping, it also properly protects the power amplifier module and the sound array hardware components, making the system highly feasible for large-scale deployment and stable in long-term operation in commercial and confidential scenarios. Attached Figure Description

[0028] Figure 1 This is a flowchart of a recording interference method based on random frequency hopping and resonant coupling.

[0029] Figure 2 This is a schematic diagram of the peak and half-power points of the spectrum.

[0030] Figure 3 This is a schematic diagram of a smooth cosine transition. Detailed Implementation

[0031] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0032] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0033] This invention discloses a recording interference method based on random frequency hopping and resonant coupling, referring to... Figure 1 This includes steps S1-S3: S1. Acquire spatial acoustic response signals and calculate resonance intensity indices.

[0034] Acquire spatial acoustic response signals, perform spectral analysis to extract candidate resonant frequency bands and their peak frequencies, calculate the product of the peak amplitude and quality factor of the candidate resonant frequency bands to obtain the resonant intensity index, and retain candidate resonant frequency bands with resonant intensity indices greater than a preset threshold to generate a set of resonant frequency bands.

[0035] In one embodiment, a microphone array is used to receive the acoustic feedback generated by the target space in response to a specific test signal. This feedback is then processed by analog-to-digital conversion to obtain a digital spatial acoustic response signal. A Hamming window is applied to the spatial acoustic response signal, followed by spectral conversion to obtain a spectral amplitude sequence. (Before extraction...) Record the frequency point corresponding to the peak with the largest amplitude. For each peak frequency point, search to the left and right for frequencies where the amplitude drops by 3dB relative to the peak amplitude, or the amplitude drops to a value equal to the peak amplitude. The lower and upper frequency limits are used to divide the interval between the lower and upper frequencies into candidate resonant frequency bands centered on each peak frequency point. (Refer to...) Figure 2The figure marks the peak frequency corresponding to the peak amplitude and the half-power point reference line used to determine the boundary of the resonant frequency band, indicating the basis for dividing the candidate resonant frequency bands. The quality factor is calculated by dividing the peak frequency by the difference between the upper and lower limit frequencies. The resonant intensity index is obtained by multiplying the peak amplitude of each candidate resonant frequency band by the quality factor. Each resonant intensity index is compared with a preset intensity threshold under the same amplitude calibration scale. Candidate resonant frequency bands less than or equal to the preset threshold are removed, and the remaining candidate resonant frequency bands constitute the resonant frequency band set.

[0036] Specifically, in the process of acquiring spatial acoustic response signals, an omnidirectional microphone with a high signal-to-noise ratio is used to achieve a high sampling rate. Record environmental sweep signals or white noise response test signals to obtain discrete-time signal sequences. The discrete-time signal sequence is arranged according to frame length. and frame shift Frame division. Each frame of the signal is multiplied by a smoothing window function. To reduce spectral leakage, a fast Fourier transform was then used to convert the spectrum to the frequency domain, and the amplitude spectrum was obtained by taking the modulus. Perform a sliding window extremum search in the amplitude spectrum to find the condition that satisfies... and The local maximum points. Among them, Indicates digital frequency point.

[0037] An amplitude threshold is set to filter out invalid noise and extract candidate peaks. All candidate peaks are sorted in descending order using a quicksort algorithm based on their amplitude. The frequency points corresponding to the top-ranked positions are determined as the amplitude peak frequencies. To map digital frequencies to physical frequencies, a relationship between physical and digital frequencies is constructed, as follows:

[0038] in, Indicates the physical frequency point. The sampling rate is obtained through microphone hardware configuration. The frame length is indicated and is obtained through the frame splitting processing parameters.

[0039] For any given candidate resonant frequency band, locate the frequency corresponding to the index of the maximum amplitude and set it as the peak frequency. Record the value at the location corresponding to the peak frequency as the peak amplitude. The algorithm iterates point by point along both the peak frequency and the frequency increase directions to find the two frequency boundary points corresponding to the first drop in amplitude to a preset proportion. If the discrete frequency point does not fall exactly at the half-power boundary, a linear interpolation method is used to fit the amplitude and frequency of two adjacent frequency points to obtain the accurate lower limit frequency. With upper limit frequency .

[0040] To evaluate the distribution range of candidate resonant frequency bands, a relationship was constructed between the peak bandwidth and the upper and lower frequency limits, as follows:

[0041] in, This indicates the peak bandwidth.

[0042] To evaluate the sharpness of the candidate resonant frequency band, a relationship was constructed between the quality factor and the peak frequency and peak bandwidth, as follows:

[0043] in, Indicates the quality factor. The peak frequency is indicated by the maximum value of the amplitude spectrum.

[0044] The peak amplitude obtained above With quality factor Perform floating-point multiplication to obtain the resonance intensity index of the candidate resonant frequency band. The resonance intensity index represents the energy response level and bandwidth sharpness of a frequency band. When performing threshold screening, the resonance intensity index uses values ​​under the same sampling, window function, and amplitude calibration conditions, or is normalized first and then compared with the preset threshold to ensure the comparability of screening results between different frequency bands.

[0045] In this way, by collecting spatial acoustic response signals and calculating resonance intensity indices to screen candidate resonant frequency bands, the link that causes ineffective energy consumption due to blind coverage of the traditional full-band is cut off. By utilizing the inherent acoustic resonance characteristics of the target space, the actual acoustic response intensity generated by the interference signal in the target space is improved, and the conversion efficiency of transmitted energy into effective interference sound pressure is increased.

[0046] S2. Configure the shift register and update the nonlinear frequency hopping sequence.

[0047] Based on the radix of the resonant frequency band set, a shift register is configured. The normalized average resonant intensity index of the calculated resonant frequency band set is converted into a binary sequence and written into the shift register to generate an initial frequency hopping index. The number of bits with a value of one in the current state of the shift register is counted as the Hamming weight. The binary bits of the initial frequency hopping index are cyclically shifted and updated according to the Hamming weight to obtain a nonlinear frequency hopping sequence.

[0048] In one embodiment, the construction length is m-sequences, the order of the shift register satisfy ,in, The number of candidate frequency bands included in the resonant frequency band set is obtained by statistically counting the number of qualified frequency bands retained. This represents the total number of in-band representation intervals within a single resonant frequency band, obtained through the system's preset frequency offset resolution. The order of the shift register is indicated and obtained through system hardware parameter configuration. A structure based on this is then constructed using... A linear feedback shift register consisting of a series of flip-flops.

[0049] The process iterates through and retains qualified candidate resonant frequency bands, calculates the average value of all resonant intensity indices within the set of resonant frequency bands, and divides the average value by the maximum resonant intensity index to obtain the normalized average resonant intensity index. Based on the shift register order, the normalized average resonant intensity index is multiplied by the full-scale unsigned integer limit and rounded down. To map the acoustic intensity characteristics to hardware storage space, a formula for calculating the decimal integer representation is constructed, with the specific relationship as follows:

[0050] in, Represents the value as a decimal integer. This represents the normalized average resonance intensity index, which is obtained by dividing the average value by the proportion of the maximum value.

[0051] The decimal integer representation is converted into a binary sequence of length equal to the order of the shift register. The binary sequence is loaded bit by bit into the shift register according to its bit weight from high to low as the initial state. The loaded binary sequence is checked for all zeros. If an all-zero sequence is detected, the state intervention logic is triggered to flip the least significant bit to 1 to avoid all-zero deadlock. Driven by a clock pulse, the tapped register state is XORed according to the primitive feedback polynomial to generate a feedback bit, which is then input to the first-stage flip-flop, converting the shifted and updated register state into the initial frequency hopping index. For example, multiple bits in the shift register can be arranged and combined in high-low order to read a multi-bit unsigned binary integer. The tapped register refers to the register node in the cascaded chain containing flip-flops that has its signal deliberately extracted; this specially extracted register node participates in the XOR operation. The primitive feedback polynomial is the characteristic polynomial corresponding to the m-sequence feedback function.

[0052] The frequency hopping disturbance interval parameter is obtained by comparing the highest and lowest resonance intensity indices and rounding down. If the calculated result is less than 1, the frequency hopping disturbance interval parameter is set to 1; if the calculated result is greater than the preset maximum interval, the frequency hopping disturbance interval parameter is set to the preset maximum interval. An internal accumulator counter is used. When the accumulator counter continuously generates discrete frequency hopping points reaching the frequency hopping disturbance interval parameter, the status word stored in the shift register at the current time is read. The absolute number of binary 1s in the status word is counted to obtain the Hamming weight. The initial frequency hopping index obtained from the register state transition is extracted. The binary bits of the initial frequency hopping index are then cyclically shifted right, with the shifted bits equal to the Hamming weight, thus outputting a nonlinear frequency hopping sequence. After the output is complete, a reset signal is applied to the accumulator counter to synchronously return it to 0.

[0053] In this way, the spatial acoustic environment assessment characteristics are deeply integrated into the initial state configuration of the hardware shift register. The initial frequency hopping index is updated by nonlinear cyclic shifting by combining the intensity index ratio and Hamming weight. This enhances the pseudo-random characteristics of interference signal generation and increases the difficulty for the post-recording processing algorithm to perform spectrum tracking and elimination of interference signals.

[0054] S3. Adjust the transition time of the raised cosine envelope and output the interference signal.

[0055] The nonlinear frequency hopping sequence is allocated to the set of resonant frequency bands through modulo operation to extract the frequency hopping points. The square root of the quality factor is used as the bias coefficient to bias the frequency hopping points towards the peak frequency point to generate an interference signal. The raised cosine envelope transition time is adjusted according to the number of frequency bands crossed by adjacent frequency hopping points and the interference signal is output.

[0056] In one embodiment, a descending order of resonant intensity is used for initial frequency hopping index allocation. Simultaneously, a separate frequency sequence number is established based on peak frequencies from low to high. When calculating the number of frequency bands traversed, the number of bands traversed is determined by the difference in frequency sequence number between the bands belonging to the previous and next frequency hopping points, or by the ratio of the actual frequency difference to a preset frequency interval. Simultaneously, the initial frequency hopping index is moduloed by the total number of represented intervals within the target resonant frequency band to obtain the in-band interval number. This in-band interval number is used to determine the actual frequency hopping points within the target resonant frequency band.

[0057] Furthermore, the square root of the quality factor of the target resonant frequency band is calculated. The square root calculation result is normalized or restricted to a preset bias coefficient range as the bias coefficient. Combined with the geometric series-based non-uniform frequency discretization method, the uniformly distributed frequency points within the target resonant frequency band are mapped to a non-uniform frequency distribution densely packed towards the peak frequency point according to the bias coefficient. The actual frequency hopping point is selected from the non-uniform frequency distribution. A digital cosine interference signal sequence is generated based on the actual frequency hopping point frequency and the preset initial phase. When switching between adjacent frequency hopping points, the absolute value of the difference between the frequency band number of the previous frequency hopping point and the frequency band number of the next frequency hopping point is calculated as the number of frequency bands crossed. The number of frequency bands crossed is multiplied by the basic time constant to obtain the raised cosine envelope transition time. When the number of frequency bands crossed is 0, a preset minimum transition time is assigned. The calculated raised cosine envelope transition time is limited to a preset proportion of the total dwell time of the frequency point to retain a stable dwell time for output.

[0058] During the transition time of the raised cosine envelope, a raised cosine window function is used for weighted smoothing. The previous frequency signal sequence is multiplied by the weight of the falling cosine edge, and the current frequency signal sequence is multiplied by the weight of the rising cosine edge. The two weighted signals are added to achieve smooth fade-in and fade-out switching. After smooth splicing, the digital signal is input to the digital-to-analog converter to be converted into an analog signal. The voltage driving capability and current driving capability are improved by the audio power amplifier circuit, and then fed to the speaker array to be converted into acoustic interference wave emission.

[0059] Specifically, the frequency hopping index obtained from the nonlinear frequency hopping sequence is regarded as a binary value, which is divided into a high-order part and a low-order part. The high-order part is modulo the number of candidate frequency bands included in the resonant frequency band set, and the remainder is used as the index of the target resonant frequency band to determine the current resonant frequency band for frequency point allocation. The low-order part is modulo the total number of preset representation intervals within the target resonant frequency band to obtain the in-band interval number.

[0060] The frequency hopping indices of the nonlinear frequency hopping sequence are modulo-operated and assigned to a set of resonant frequency bands arranged in descending order to determine the upper and lower cutoff frequencies of the resonant frequency bands. To evaluate the high-end frequency offset span, a relationship is constructed between the first frequency difference, the upper cutoff frequency, and the peak frequency. The specific relationship is as follows:

[0061] in, Indicates the first frequency difference. This indicates the upper cutoff frequency, which is obtained by reading the high-frequency boundary parameters of the resonant band. The peak frequency is indicated by reading the parameters of the central energy concentration point.

[0062] To evaluate the low-end frequency offset span, a relationship was constructed between the second frequency difference and the peak frequency and the lower cutoff frequency. The specific relationship is as follows:

[0063] in, Indicates the second frequency difference. The lower cutoff frequency is indicated by reading the low-frequency boundary parameters of the resonant band and extracting the square root of the quality factor of the target resonant band as the bias coefficient.

[0064] The first frequency difference and the second frequency difference are each divided into a predetermined number of representation intervals. To construct a frequency distribution densely concentrated towards the peak frequency, the frequency step size of the representation intervals is set to increase in a geometric progression. The relationship between the common ratio of the geometric progression and the bias coefficient is established, as follows:

[0065] in, This represents the common ratio of a geometric sequence. This represents the bias coefficient, which is obtained by calculating the square root of the quality factor.

[0066] Taking the low-frequency band as an example, calculate the initial fundamental frequency step size. This divides the second frequency difference into Adjacent intervals of unequal width, closer to the peak frequency point The interval has a low series number and a small frequency step size. The narrower the band appears and the higher the distribution density, the more likely it is to form a spacing from the peak frequency to the cutoff frequency that is proportional to the common ratio throughout the lower sideband. The exponentially expanding non-uniform distribution state; the same processing steps are also applied symmetrically to the upper sideband represented by the first frequency difference.

[0067] This represents the interval index, obtained through in-band interval number mapping. When the in-band interval number obtained by pseudo-random generation and modulo operation falls within the resonant frequency band, the in-band interval number is mapped to 2... The interval selection value is obtained by taking the modulo; when the interval selection value is less than When it is located on the low-frequency side and corresponds to the first interval number; when the interval selection value is not less than When determining its location on the high-frequency side and its corresponding second interval number, a segmented lookup table or boundary comparison method is used to determine the specific interval corresponding to that number. The frequency hopping point is then identified as either low-frequency or high-frequency, and the midpoint frequency within the corresponding frequency range is extracted as the actual frequency hopping point. This actual frequency hopping point is assigned to the frequency control parameters of the digital frequency synthesizer or audio signal generator, serving as the actual output acoustic frequency hopping point. A digital cosine interference signal sequence is then generated based on the frequency hopping point.

[0068] When switching between adjacent frequency hopping points, the absolute value of the difference in frequency sequence number between the resonant frequency bands of the previous and next frequency hopping points is calculated. To calculate the frequency hopping span, a relationship is established between the number of frequency bands traversed and the frequency band numbers of the previous and next frequency hopping points. The specific relationship is as follows:

[0069] in, Indicates the number of frequency bands spanned. This indicates the frequency band number of the previous frequency hop, which is obtained by reading the sequential position of the current frequency hop. This indicates the frequency band number of the next hop frequency point, which is obtained by reading the sequential position of the target frequency hop frequency point within the frequency band.

[0070] To achieve smooth fade-in and fade-out switching between adjacent frequency points, a raised cosine envelope window function was constructed, with the specific relationship as follows:

[0071] in, This represents the raised cosine envelope window function. This represents a time variable, obtained through system clock timing. The transition time of the raised cosine envelope is represented by adjusting the number of frequency bands crossed.

[0072] When the number of frequency bands crossed is greater than 0, the number of frequency bands crossed is multiplied by the fundamental time constant to obtain the raised cosine envelope transition time. A minimum value limiting function is used to ensure that the calculation result does not exceed the system's allocated fixed total dwell time threshold for a single frequency point. This gradual transition reduces out-of-band energy leakage and transient burst noise. When the number of frequency bands crossed is 0, it means that the next frequency point still belongs to the same local resonant frequency band as the current one, and the frequency jump is small. Therefore, the transition time is set to the preset minimum transition time. In this case, the envelope function shortens the same-band switching time while ensuring a basically smooth transition, ensuring a compact frequency point winding without wasting output power and a dwell period under the same resonance peak. (Refer to...) Figure 3 The cross-band transition uses a longer smoothing time to suppress spectrum leakage, while the same-band transition uses a shorter transition time to ensure the real-time performance of frequency switching.

[0073] Thus, by dividing the non-uniform representation interval based on the bias coefficient, the generated frequency hopping points are concentrated near the center peak frequency point with the largest resonant energy, which improves the conversion rate of transmitted electrical power to effective sound pressure in spatial acoustics. The raised cosine envelope transition time is adaptively adjusted according to the number of cross-band frequency jumps, which suppresses amplitude abrupt changes caused by cross-band frequency jumps and transient out-of-band spectrum leakage, thus avoiding transient current impact on the loudspeaker array hardware.

[0074] With a volume of 50m 3 In the acoustic testing chamber, a test environment was constructed using a high signal-to-noise ratio omnidirectional microphone and a wideband speaker system with a sampling rate of 48kHz. The total dwell time for a single frequency point was fixed at 20ms, the number of candidate resonant frequency bands was 5, the number of representation intervals was 16, and the basic transition time constant was 2ms. Comparison schemes included a baseline group with traditional uniform frequency hopping and a rectangular window with a fixed transition time; ablation group one using only a non-uniform representation bias frequency point generation mechanism; ablation group two using only the raised cosine envelope transition time adjusted based on the number of frequency bands crossed; and an experimental group using a complete combined mechanism.

[0075] The baseline group had a resonant energy coupling coefficient of 42%, an out-of-band spectral leakage peak of -35 dB, and an average in-band switching delay of 2 ms. The resonant energy coupling coefficient of ablation group one improved to 78%, but the out-of-band spectral leakage peak worsened to -28 dB, while the average in-band switching delay remained at 2 ms. The resonant energy coupling coefficient of ablation group two was 45%, the out-of-band spectral leakage peak was suppressed to -55 dB, and the average in-band switching delay was shortened to 0.5 ms. The complete experimental group achieved a resonant energy coupling coefficient of 83%, a stable out-of-band spectral leakage peak at -62 dB, and an average in-band switching delay maintained at 0.5 ms.

[0076] The non-uniform representation based on the distance between the peak frequency and the bandwidth boundary causes the frequency hopping trajectory to exhibit a high-probability bias phenomenon of concentrating towards the resonant peak, which excites the resonant energy amplification effect of the acoustic cavity, leading to an improvement in energy coupling efficiency. The raised cosine weighted smoothing mechanism controls the envelope transition band length by analyzing the number of frequency bands crossed. When there are large-span jumps across bands, a gradual transition is used to strongly suppress out-of-band spectral leakage, while when switching within the same frequency band, an extremely short delay is automatically assigned to ensure the compactness of the frequency point winding. The two work together to enable the signal output to have both high resonant response strength and low out-of-band spectral leakage.

[0077] This invention also discloses a recording interference system based on random frequency hopping and resonant coupling, including a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement the recording interference method based on random frequency hopping and resonant coupling according to the present invention.

[0078] The aforementioned recording interference system based on random frequency hopping and resonant coupling also includes other components well known to those skilled in the art, such as communication buses and communication interfaces. Their settings and functions are known in the art and will not be described in detail here.

[0079] In the description of this specification, "multiple" or "several" means at least two, such as two, three or more, unless otherwise expressly and specifically defined.

Claims

1. A recording interference method based on random frequency hopping and resonant coupling, characterized in that, include: S1. Collect spatial acoustic response signals, perform spectral analysis to extract candidate resonant frequency bands and their peak frequencies, calculate the product of the peak amplitude and quality factor of the candidate resonant frequency bands to obtain the resonant intensity index, and retain the candidate resonant frequency bands whose resonant intensity index is greater than a preset threshold to generate a set of resonant frequency bands. S2. Configure a shift register based on the number of candidate frequency bands included in the resonant frequency band set and the preset total number of in-band representation intervals; convert the calculated normalized average resonant intensity index of the resonant frequency band set into a binary sequence and write it into the shift register; perform linear feedback shift based on the primitive feedback polynomial to continuously generate an initial frequency hopping index; count the Hamming weight of the current state of the shift register at a preset period, and perform a cyclic shift of the current initial frequency hopping index with a bit length equal to the Hamming weight to obtain a nonlinear frequency hopping sequence; S3. The nonlinear frequency hopping sequence is allocated to the resonant frequency band set through modulo operation to extract the frequency hopping points. The square root of the quality factor is used as the bias coefficient to bias the frequency hopping points toward the peak frequency point to generate an interference signal. The raised cosine envelope transition time is adjusted according to the number of frequency bands crossed by adjacent frequency hopping points and the interference signal is output. The interference signal is then output after electroacoustic conversion.

2. The recording interference method based on random frequency hopping and resonant coupling according to claim 1, characterized in that, The acquisition of spatial acoustic response signals for spectral analysis to extract candidate resonant frequency bands and their peak frequencies includes: The acquired spatial acoustic response signal is windowed and converted to the frequency domain using a fast Fourier transform to obtain the amplitude spectrum; the frequency points corresponding to the local maxima in the amplitude spectrum are extracted as amplitude peak frequencies; the amplitude peak frequencies are arranged in descending order according to their corresponding amplitudes, and a preset number of the top frequencies are extracted to divide the candidate resonant frequency bands.

3. The recording interference method based on random frequency hopping and resonant coupling according to claim 1, characterized in that, The product of the peak amplitude and the quality factor of the candidate resonant frequency band is used to obtain the resonant intensity index, including: The frequency corresponding to the maximum amplitude within the candidate resonant frequency band is determined as the peak frequency point, and the corresponding peak amplitude is obtained. The peak bandwidth is obtained by subtracting the two frequency points where the search amplitude drops to a preset proportion of the peak amplitude. The quality factor is obtained by dividing the peak frequency by the peak bandwidth, and the resonance intensity index is obtained by multiplying the quality factor by the peak amplitude.

4. The recording interference method based on random frequency hopping and resonant coupling according to claim 1, characterized in that, The normalized average resonant intensity index of the calculated resonant frequency band set is converted into a binary sequence and written into the shift register to generate the initial frequency hopping index, including: Obtain the average and maximum values ​​of the resonance intensity index within the resonant frequency band set; The normalized intensity value is obtained by dividing the average value by the maximum value. The normalized intensity value is multiplied by the full-scale unsigned integer limit value and rounded down to obtain a decimal representation value. The decimal representation value is then converted into a binary bit string sequence with a length matching the order of the shift register.

5. The recording interference method based on random frequency hopping and resonant coupling according to claim 4, characterized in that, After converting the decimal representation value into a binary bit string sequence with a length matching the order of the shift register, the process further includes: When the binary bit string sequence is determined to be an all-zero sequence, the state intervention logic is triggered to set the least significant bit of the binary bit string sequence to 1 to prevent the register from falling into an all-zero deadlock state.

6. The recording interference method based on random frequency hopping and resonant coupling according to claim 1, characterized in that, The step of calculating the Hamming weight of the current state of the shift register at a preset period and performing a cyclic shift on the current initial frequency hopping index includes: A counter is set to count the number of frequency hopping points that have been generated. When the counter reaches the frequency hopping interval period determined by the ratio of the highest to the lowest resonance intensity index, the cyclic shift is triggered. The status word stored in the shift register at the trigger time is read, and the number of binary digits 1 contained in the status word is counted by a bit counting instruction as the Hamming weight. A bit-by-bit cyclic right shift operation is performed on the current initial frequency hopping index, and the number of bits shifted is equal to the Hamming weight. After the output is completed, the counter is synchronously reset to zero.

7. The recording interference method based on random frequency hopping and resonant coupling according to claim 1, characterized in that, Using the square root of the quality factor as a bias coefficient, the frequency hopping point is biased towards the peak frequency to generate an interference signal, including: Calculate the first frequency difference between the high-frequency cutoff boundary of the candidate resonant frequency band and the peak frequency point, and the second frequency difference between the peak frequency point and the low-frequency cutoff boundary; divide the first frequency difference and the second frequency difference into a preset number of representation intervals, and the frequency step size of the representation intervals increases in a geometric progression, wherein the common ratio of the geometric progression is 1 and the sum of the bias coefficient; The nonlinear frequency hopping sequence is subjected to the modulo operation to obtain the in-band interval number, which is then mapped to the corresponding representation interval. The center frequency within the representation interval is used as the frequency hopping point.

8. The recording interference method based on random frequency hopping and resonant coupling according to claim 1, characterized in that, The adjustment of the raised cosine envelope transition time based on the number of frequency bands spanned by adjacent frequency hopping points includes: Calculate the absolute value of the difference between the resonant frequency bands of the previous hop frequency point and the next hop frequency point in the frequency sequence number arranged from low to high peak frequency points, and use it as the number of frequency bands crossed; When it is determined that the number of frequency bands crossed is greater than 0, the number of frequency bands crossed is multiplied by the basic time constant to obtain the raised cosine envelope transition time, and it is restricted to not exceed the fixed total dwell time threshold of a single frequency point allocated by the system. When the number of frequency bands crossed is determined to be 0, the raised cosine envelope transition time is set to a preset minimum transition time to shorten the same-band switching delay.

9. The recording interference method based on random frequency hopping and resonant coupling according to claim 1, characterized in that, The output of the interference signal includes: The interference signal is input to a digital-to-analog converter and converted into an analog signal; the voltage and current driving capability of the analog signal is enhanced by an audio power amplifier circuit; the amplified analog signal is fed to a speaker array and converted into an acoustic interference wave and emitted into the target space.

10. A recording interference system based on random frequency hopping and resonant coupling, characterized in that, include: A processor and a memory, the memory storing computer program instructions that, when executed by the processor, implement the recording interference method based on random frequency hopping and resonant coupling according to any one of claims 1-9.

Images