A method and system for noise management

Abstract

The application discloses a muting management method and system. With the help of FPGA and DPS, FFT signals, voice signals and noise signals are extracted from the intermediate frequency voice signals after frequency down conversion, and the signal-to-noise ratio and the carrier-to-noise ratio of the voice signals are calculated. The signal-to-noise ratio and the carrier-to-noise ratio are compared with the set signal-to-noise ratio opening noise threshold, the signal-to-noise ratio closing noise threshold, the carrier-to-noise ratio opening noise threshold and the carrier-to-noise ratio closing noise threshold to make the muting decision. Furthermore, the maximum energy value is taken as the benchmark to divide into multiple energy levels, and the sum of the energy of the frequency points whose energy does not reach the energy level under each energy level and the maximum energy value are taken to calculate the carrier-to-noise ratio. The carrier-to-noise ratio is corrected according to the average value of the pure carrier carrier-to-noise ratio under the corresponding condition. The carrier-to-noise ratio opening and closing noise thresholds are dynamically adjusted according to the average value of the carrier-to-noise ratio in the first time length and the first threshold value. The application solves the problem of incorrect muting decision in the single carrier interference, and the dynamic adjustment of the threshold realizes the muting management in the complex electromagnetic environment.

Description

Technical Field

[0001] This invention relates to the field of wireless voice communication, and in particular to a squelch management method and system. Background Technology

[0002] In wireless voice communication, squelch functions can reduce, suppress, or eliminate channel noise during transmission, making the voice clear and understandable. Commonly used squelch methods fall into six main categories: pilot squelch, noise squelch, carrier detection squelch, sub-audio codec squelch, speech spectrum squelch, and combined carrier and speech squelch.

[0003] Among the six noise reduction methods mentioned above, carrier detection squelch is widely used due to its simple implementation. However, carrier detection squelch is prone to misjudgment when single-carrier interference occurs due to the low carrier-to-noise ratio. At the same time, under low-frequency AM, signal and squelch management will be disordered, and AM and FM will suffer from severe signal trailing when judging large signals. Summary of the Invention

[0004] The purpose of this invention is to provide a squelch management method to address all or part of the aforementioned problems, thereby solving the problem that the carrier-to-noise ratio is too low during single-carrier interference in the carrier squelch process of wireless voice communication, which is not conducive to squelch determination.

[0005] The technical solution adopted in this invention is as follows:

[0006] A noise suppression management method, which calculates the signal-to-noise ratio (SNR) and carrier-to-noise ratio (CNR) of a speech signal and compares them with preset SNR noise-on threshold, SNR noise-off threshold, CNR noise-on threshold, and CNR noise-off threshold to determine noise suppression; the method further includes:

[0007] The calculated carrier-to-noise ratio is corrected;

[0008] Dynamically adjust the carrier-to-noise ratio (CNR) noise-on threshold and carrier-to-noise ratio (CNR) noise-off threshold.

[0009] Furthermore, methods for calculating the carrier-to-noise ratio include:

[0010] Calculate the sum of energy at each frequency point within each unit of time, using a predetermined time period as the unit.

[0011] Determine the frequency point and energy value of the maximum energy value;

[0012] Based on the maximum energy value, it is divided into multiple energy levels;

[0013] Calculate the noise variable for each energy level, where the noise variable is the count of frequency points where the energy level has not been reached;

[0014] The noise energy sum is obtained by taking the sum of the energy of the frequency points corresponding to the noise variables, and the maximum energy value is taken as the carrier energy. The difference between the logarithms of the two is used to obtain the carrier-to-noise ratio.

[0015] Furthermore, the noise variable is a count of frequency points where the energy level does not reach the required level within the effective signal bandwidth.

[0016] Furthermore, when the energy level is less than the maximum noise value of the left or right band, the speech signal at that energy level is determined to be full noise.

[0017] Furthermore, the correction of the calculated carrier-to-noise ratio includes:

[0018] Calculate the carrier-to-noise ratio (CNR) per unit time.

[0019] If the difference between the carrier-to-noise ratio (CNR) value in the current unit time and the CNR value in the previous unit time is greater than the first threshold, and the difference between the CNR value in the previous unit time and the CNR value in the unit time before that does not exceed the first threshold, then the CNR value in the previous unit time is taken as the CNR value in the current unit time.

[0020] Furthermore, the dynamic adjustment of the carrier-to-noise ratio (CNR) noise-on threshold and carrier-to-noise ratio (CNR) noise-off threshold includes:

[0021] For each set of carrier-to-noise ratio values ​​calculated per unit time, the statistical variable N2 is accumulated until it is cleared to zero when the voice signal is transmitted.

[0022] When the time corresponding to the statistical variable N2 is greater than the first duration, the average carrier-to-noise ratio of the first modulation mode within the first duration is calculated.

[0023] When the carrier noise dynamic adjustment switch is active, the carrier noise ratio on threshold and carrier noise ratio off threshold are dynamically adjusted based on the calculated average carrier noise ratio in the first modulation mode.

[0024] Furthermore, the step of dynamically adjusting the carrier-to-noise ratio (CNR) on-noise threshold and carrier-to-noise ratio (CNR) off-noise threshold based on the calculated average CNR value under the first modulation mode includes:

[0025] When the average carrier-to-noise ratio (CNR) under the first modulation mode is less than or equal to the current CNR noise-off threshold, the CNR on / off threshold is restored to the initially set CNR on / off threshold. When the average CNR under the first modulation mode is greater than the current CNR noise-off threshold, and the average CNR under the first modulation mode plus the first threshold is less than the current CNR noise-on threshold, the average CNR under the first modulation mode is used as the new CNR noise-on threshold. When the average CNR under the first modulation mode is greater than the current CNR noise-off threshold, and the average CNR under the first modulation mode plus the first threshold is greater than the current CNR noise-on threshold, the average CNR under the first modulation mode is used as the new CNR noise-off threshold, and the average CNR under the first modulation mode plus the first threshold is used as the new CNR noise-on threshold.

[0026] Furthermore, the correction of the calculated carrier-to-noise ratio includes:

[0027] Calculate the carrier-to-noise ratio (CNR) and signal-to-noise ratio (SNR) for each group per unit time under the current carrier center. If the first judgment condition for the instantaneous CNR and SNR is met, the statistical variable N1 under the current carrier center is accumulated; otherwise, the statistical variable N1 is cleared to zero.

[0028] When the time corresponding to the statistical variable N1 under the current carrier center is greater than the second duration, the final carrier-to-noise ratio value under the current carrier center is assigned as: the average carrier-to-noise ratio value of each energy level with a bandwidth of 4KHz under the current carrier center per unit time.

[0029] Furthermore, the correction of the calculated carrier-to-noise ratio also includes:

[0030] When the sum of the times corresponding to the statistical variable N1 under the current carrier center and the statistical variable N1 under the adjacent carrier center is greater than or equal to the second duration, the final carrier-to-noise ratio (CNR) under the current carrier center and the final CNR under the adjacent carrier center are respectively assigned as: the average CNR of each energy level with a bandwidth of 4KHz under the current carrier center and the adjacent carrier center per unit time.

[0031] Furthermore, the sliding window method is used to calculate the download noise ratio per unit time for each group.

[0032] To address all or some of the above problems, the present invention also provides a noise suppression management system, including an FPGA and a DSP;

[0033] The FPGA is configured to: extract the FFT signal before demodulation of the speech signal, as well as the demodulated speech signal and noise signal; and transmit the FFT signal, speech signal and noise signal to the DSP;

[0034] The DPS is configured to: calculate the signal-to-noise ratio (SNR) and carrier-to-noise ratio (CNR) of the speech signal based on the FFT signal, the speech signal, and the noise signal; compare the SNR with the set noise-on threshold, noise-off threshold, CNR noise-on threshold, and CNR noise-off threshold to make a noise reduction decision; correct the calculated CNR; and dynamically adjust the CNR noise-on threshold and CNR noise-off threshold.

[0035] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0036] 1. This invention ensures the accuracy of the carrier-to-noise ratio by correcting the carrier-to-noise ratio, and to a certain extent solves the problem of incorrect squelch judgment caused by the inconvenience of detecting the small carrier-to-noise ratio when there is single-carrier interference.

[0037] 2. This invention improves the squelch judgment accuracy in response to real-time environmental influences by dynamically adjusting the carrier-to-noise ratio (CNR) and the noise-to-noise ratio (NNR) to match the CNR of real-time voice signals and the corresponding modulation methods (AM, FM), and realizes squelch management in complex electromagnetic environments. Attached Figure Description

[0038] The present invention will be described by way of example and with reference to the accompanying drawings, wherein:

[0039] Figure 1 This is the architecture diagram of the noise reduction management system.

[0040] Figure 2 It is a schematic diagram of the spectrum before the superposition of a set of spectra.

[0041] Figure 3 It is a schematic diagram of the spectrum after the superposition of a set of spectra.

[0042] Figure 4 This is a schematic diagram of the spectrum classification.

[0043] Figure 5 It is a flowchart for dynamically adjusting the carrier-to-noise ratio (CNR) noise-on / noise-off threshold. Detailed Implementation

[0044] All features disclosed in this specification, or all steps in all disclosed methods or processes, may be combined in any way, except for mutually exclusive features and / or steps.

[0045] Any feature disclosed in this specification (including any appended claims and abstract) may be replaced by other equivalent or similar features, unless specifically stated otherwise. That is, unless specifically stated otherwise, each feature is merely one example of a series of equivalent or similar features.

[0046] A squelch management method is disclosed. This method calculates the signal-to-noise ratio (SNR) and carrier-to-noise ratio (CNR) of a speech signal and compares them with preset SNR noise-on threshold, SNR noise-off threshold, CNR noise-on threshold, and CNR noise-off threshold to determine whether to enable / disable squelch or maintain the current state. Furthermore, this method also corrects the calculated CNR to dynamically adjust the CNR noise-on / off threshold, thereby achieving squelch management in complex electromagnetic environments.

[0047] This method processes the intermediate frequency signal after down-conversion, such as... Figure 1 As shown, in terms of execution, it can be implemented using a system composed of a modem module FPGA+DSP. Therefore, in the embodiments of the present invention, the description of the method can also be applied to the description of the corresponding configuration of the FPGA and DSP.

[0048] In the signal-to-noise ratio (SNR) and carrier-to-noise ratio (CNR) calculation process, the FPGA extracts the demodulated 32kHz voice signal and noise signal; the FPGA also extracts the undemodulated intermediate frequency (IF) signal, performs 127-point decimation, and passes it through a bandpass filter to obtain a 256kHz FFT signal. These three signals—voice signal, noise signal, and FFT signal—are transmitted to the DSP via a bus interface. Upon receiving the signal, the DSP performs calculations every 8ms (the example unit of time), calculating the voice data (256 data points in 8ms, with signal energy calculated as the arithmetic mean of the squares of these 256 data points), the noise data (256 data points in 8ms, with noise energy calculated as the arithmetic mean of the squares of these 256 data points), and the FFT data (256 data points in 8ms; due to the symmetry of the FFT data, the first 128 data points from each group are summed to obtain the FFT spectrum). Based on these three calculation results, the SNR and CNR are calculated.

[0049] Furthermore, this method only corrects the carrier-to-noise ratio under pure carrier conditions. This is because during normal conversations, modulation between words is intermittent, resulting in a lower calculated carrier-to-noise ratio, which is detrimental to squelch determination.

[0050] The following details the process of correcting the carrier-to-noise ratio in this method (and in the DSP):

[0051] 1. Data Collection

[0052] a) Baseband speech signal acquisition

[0053] Generally, the frequency range of speech signals is 300Hz to 3.4kHz. The baseband speech signal recovered after analog demodulation is output at a sampling rate of 32kHz, with a total of 1024 points collected in 1024 / 32 = 32ms.

[0054] b) Intermediate Frequency FFT Data Acquisition

[0055] The demodulated speech signal was analyzed by a 256-point FFT and then output at a sampling rate of 256kHz, collecting 8192 points in 8192 / 256 = 32ms.

[0056] 2. Calculate the initial value of the carrier-to-noise ratio.

[0057] Each FFT calculation consists of 256 data points, each 8ms long, with a width of 256kHz. Eight calculations are performed, resulting in eight groups and a total of 2048 data points. The statistical variable N² is incremented by 1 after each calculation unless the zeroing condition is met. Due to the symmetry of the FFT data, only the first 128 data points from each group are used. To reduce the number of logarithmic calculations, the FFT data is first analyzed to determine the ratio of carrier energy to noise energy, and then the logarithm is calculated. (The carrier center frequency is 63kHz without offset). After passing through a bandpass filter, the effective frequency range is approximately 17kHz to the left and right of the center frequency. Therefore, the calculation focuses on data within this 17kHz range around the carrier center.

[0058] The initial value of the carrier-to-noise ratio is calculated as follows:

[0059] a) Calculate the sum of energy for each frequency point within each unit time (8ms as an example). The starting point of the data, _s_sign_star, is 50kHz, and the ending point, _s_sign_end, is 76kHz. The summed data can be placed in the array aS_sum_1k[i], where i corresponds to each frequency point. Therefore, aS_sum_1k[0] represents the sum of energy at 50kHz, and so on. aS_sum_1k

[27] represents the sum of energy at 76kHz. Each calculation increments the statistical variable N2 by 1, representing the number of groups of received voice signals. The statistical variable N2 is cleared to zero when the clearing condition is met in the calculation of each group. Taking pure carrier as an example, the spectrum of 1ms data is shown in the attached figure. Figure 2 As shown. Due to spectral symmetry, amplitude values ​​from 0 to 127 points are used for spectrum calculation. The spectrum obtained by superimposing the 8ms spectra is shown in the attached figure. Figure 3 As shown.

[0060] As attached Figure 4As shown, firstly, a search is performed within the passband to find the point with the maximum energy value. The frequency of the maximum energy value is denoted as Max_Point, and the amplitude of the maximum energy value is denoted as Max_Value. Based on the maximum energy value Max_Value, the energy value is divided into 16 (or other) energy levels. The first level division line L1 (i.e., the first energy level) is taken as Max_Value / 4 (when dividing into other levels, the denominator 4 needs to be adjusted accordingly), which means the maximum energy value is reduced by 6dB. The second level division line L2 = L1 / 4, which means the energy value of the first level division line L1 is reduced by 6dB. The third level division line L3 = L2 / 4, which means the energy value of the second level is reduced by 6dB. The fourth level division line L4 = L3 / 4, which means the energy value of the third level is reduced by 6dB... The sixteenth level division line L16 = L15 / 4, which means the energy value of the fifteenth level is reduced by 6dB.

[0061] Then, using each energy level as a baseline, the signal and noise variables are calculated separately:

[0062] Before proceeding, let's define the effective signal bandwidth of the spectrum. The effective signal bandwidth is the bandwidth within 10 kHz to the left and right of the carrier center, starting from _s_sign_L (53 kHz) and ending at _s_sign_R (73 kHz), totaling 21 kHz. The spectrum outside the effective signal bandwidth, located to the left of the effective signal bandwidth, is called left-side noise; similarly, the spectrum to the right of the effective signal bandwidth is called right-side noise.

[0063] Taking the third segmentation line L3 as an example, the energy value of the third segmentation line L3 is compared with the maximum noise value of the left sideband (Max_left), and also compared with the maximum noise value of the right sideband (Max_right). If the energy value of the third segmentation line L3 is less than either of the maximum noise values ​​of the left and right sidebands, the speech signal at this energy level is considered to be pure noise, and the process exits. Otherwise, the process iterates within the effective signal bandwidth, counting the number of frequency points with energy values ​​greater than the energy value of the third segmentation line L3 and recording them in the signal variable `sign_bandwidth`, and counting the number of frequency points with energy values ​​less than the energy value of the third segmentation line L3 and recording them in the noise variable `noise_bandwidth`. By repeating this method for each segmentation line, the statistical analysis of the signal and noise variables for all 16 segmentation lines can be completed.

[0064] Using each segmentation line as a reference, the sum of the amplitude values ​​of the noise variable noise_bandwidth at the corresponding frequency point for each baseline (i.e., the currently calculated segmentation line) is calculated, which is the noise energy sum, and the logarithm is denoted as N. The center point of the signal is Max_Point, and the corresponding amplitude value is the maximum energy value, which is taken as the carrier energy, and the logarithm is denoted as C. The difference between the two is the carrier-to-noise ratio (CNR). The CNR values ​​are sorted in ascending order according to the bandwidth of each segmentation line. In subsequent calculations, the CNR under pure carrier conditions is the CNR value with a bandwidth (calculated symmetrically about the carrier center) of 4kHz, the CNR value for AM operating mode is AM_CN with a bandwidth of 14kHz, and the CNR value for FM operating mode is FM_CN with a bandwidth of 18kHz. For ease of explanation, and to distinguish it from the pure carrier condition, AM and FM operating modes are collectively referred to as the first modulation mode. The specific modulation and demodulation are based on the specific operating mode, and the CNR of the latter two (AM and FM operating modes) is recorded as the applied CNR.

[0065] 3. Correct the carrier-to-noise ratio.

[0066] As mentioned earlier, this method only corrects the carrier-to-noise ratio under pure carrier conditions.

[0067] a) Pure carrier determination

[0068] Based on the calculated instantaneous (i.e., current unit time) carrier-to-noise ratio (CNR) and instantaneous signal-to-noise ratio (SNR), determine whether it falls under the pure carrier condition (i.e., the first criterion based on the instantaneous CNR and SNR values). The method for determining the pure carrier condition is as follows:

[0069] (1) The carrier-to-noise ratio at a bandwidth of 4KHz is greater than 20, and the signal-to-noise ratio at this time is less than or equal to 11;

[0070] (2) The carrier-to-noise ratio at a bandwidth of 4KHz is greater than 15, and the signal-to-noise ratio at this time is less than or equal to 12.

[0071] (3) The carrier-to-noise ratio at a bandwidth of 4KHz is greater than 10, and the signal-to-noise ratio at this time is less than or equal to 13.

[0072] When the current group is determined to be a pure carrier condition, the count of the statistical variable N1 is incremented by 1; otherwise, the count of the statistical variable N1 is cleared to zero.

[0073] b) Carrier-to-noise ratio correction

[0074] If the difference between the carrier-to-noise ratio (CNR) calculated by the current group and the CNR calculated by the previous group does not exceed 6 (an adjustable first threshold), no correction is performed. If the difference between the CNR calculated by the current group and the CNR calculated by the previous group is greater than 6, and the difference between the CNR calculated by the previous group and the CNR calculated by the group before that does not exceed 6, then correction is performed, and the CNR calculated by the previous group is used as the CNR calculated by the current group. No correction is performed in all other cases.

[0075] c) Conditions for clearing statistical variables

[0076] When the current group is under non-pure carrier conditions, statistical variable N1 is reset to zero, as are the energy and value of the current group. Additionally, during speech signal transmission, both statistical variables N1 and N2 are reset to zero.

[0077] d) Handling long-term interference points

[0078] The average carrier-to-noise ratio (CNR) is calculated based on an 8-second time segment (the first settable duration). This includes both the average CNR under pure carrier conditions and the average CNR for application. When the time corresponding to statistical variable N2 (one unit of statistical variable N2 corresponds to 8ms) is greater than or equal to 8 seconds, the average CNR for application within that 8-second time segment, sCN_8s, is calculated. The average CNR for pure carrier conditions within that 8-second time segment, sCN10_8s, can also be calculated for later use. If the dynamic CNR adjustment switch is active, the CNR switching threshold is dynamically adjusted based on the calculated average CNR for application, sCN_8s, with an adjustment period of 8 seconds. The adjustment process is detailed in the attached diagram. Figure 5 As shown. When sCN_8s is less than or equal to the current carrier-to-noise ratio (CNR) noise reduction threshold, the CNR on / off noise threshold is restored to the initially set CNR on / off noise threshold; when sCN_8s is greater than the current CNR noise reduction threshold, and sCN_8s+6 is less than the current CNR noise activation threshold, sCN_8s is used as the new CNR noise activation threshold; in other cases (sCN_8s is greater than the current CNR noise reduction threshold, and sCN_8s+6 is greater than or equal to the current CNR noise activation threshold), if sCN_8s+6 is greater than the current CNR noise activation threshold, then sCN_8s is used as the new CNR noise reduction threshold, and sCN_8s+6 is used as the new CNR noise activation threshold; otherwise, no adjustment is made.

[0079] When the time corresponding to the statistical variable N1 at a certain carrier center (e.g., 63kHz as described in the previous embodiment, referring to the voice signal using that carrier center) exceeds 16s (an adjustable second duration), the final carrier-to-noise ratio (CNR) value at that carrier center is assigned the average of the pure carrier CNR values ​​(i.e., the CNR value at the energy level with a bandwidth of 4kHz) per unit time. If, within the signal band, the sum of the statistical variable N1 at the current carrier center for 16s and the sum of the statistical variable N1 at adjacent carrier centers using adjacent frequency points (e.g., the current carrier center is 63kHz, and the adjacent carrier centers are 62kHz and 64kHz) is greater than or equal to 16s, the CNR value at the current carrier center and the CNR value at the adjacent carrier centers are both assigned the average of the pure carrier CNR values ​​per unit time for their respective groups. Additionally, if the sum of these times lasts for less than half of 16 seconds, the speech signal under the current carrier center is determined to be non-single-carrier interference, and the carrier-to-noise ratio (CNR) under this carrier center is reset to zero. If none of the above conditions are met, the calculated mean CNR and the mean of the pure carrier are retained.

[0080] e) Pure carrier interference handling: Under pure carrier conditions, take the average carrier-to-noise ratio (CNR) of the frequency point of the maximum energy value (i.e., the carrier center) for 8 seconds. If the average CNR is greater than 0, and the CNR of the current group is less than the average CNR + 5, and the pure carrier elimination function is effective, and the pure carrier continuous time exceeds 256ms (adjustable threshold), then the adjustment of the CNR of the current group and the correction of the 10dBm bandwidth (referring to the bandwidth formed by the intersection of the maximum energy value and the spectrum when the maximum energy value is reduced by 10dBm) will be enabled.

[0081] In addition, to ensure the accuracy of carrier-to-noise ratio (CNR) calculation and correction, in this embodiment of the invention, a sliding window method is used to calculate the CNR value. Taking a window size of 4 as an example, after accumulating 4 consecutive CNR values, the average CNR value is taken as the CNR of the current group. Among the 4 groups, the last group is the current group.

[0082] This invention is not limited to the specific embodiments described above. The invention extends to any new feature or combination disclosed in this specification, as well as any new method or process step or combination disclosed herein.

Claims

1. A noise suppression management method, wherein the method calculates the signal-to-noise ratio (SNR) and carrier-to-noise ratio (CNR) of a speech signal and compares them with preset SNR noise-on threshold, SNR noise-off threshold, CNR noise-on threshold, and CNR noise-off threshold to determine noise suppression; characterized in that, Also includes: The calculated carrier-to-noise ratio is corrected, including: calculating the carrier-to-noise ratio value under the current unit time; if the difference between the carrier-to-noise ratio value under the current unit time and the carrier-to-noise ratio value under the previous unit time is greater than the first threshold, and the difference between the carrier-to-noise ratio value under the previous unit time and the carrier-to-noise ratio value under the unit time before that does not exceed the first threshold, then the carrier-to-noise ratio value under the previous unit time is used as the carrier-to-noise ratio value under the current unit time. Dynamically adjusting the carrier-to-noise ratio (CNR) noise-on threshold and carrier-to-noise ratio (CNR) noise-off threshold includes: accumulating the statistical variable N2 for each set of CNR values ​​per unit time until it is cleared to zero when the voice signal is transmitted; calculating the average CNR under the first modulation mode within the first duration when the time corresponding to the statistical variable N2 is greater than a first duration; and, with the carrier-to-noise ratio dynamic adjustment switch active, restoring the CNR noise-on / noise-off threshold to the initially set threshold when the average CNR under the first modulation mode is less than or equal to the current CNR noise-off threshold, when the CNR noise-off threshold under the first modulation mode is less than or equal to the current CNR noise-off threshold; and adjusting the CNR noise-on / noise-off threshold when the CNR noise-off threshold under the first modulation mode is less than or equal to the current CNR noise-off threshold. When the average carrier-to-noise ratio (CNR) is greater than the current CNR noise-off threshold, and the average CNR plus the first threshold under the first modulation mode is less than the current CNR noise-on threshold, the average CNR under the first modulation mode is used as the new CNR noise-on threshold. When the average CNR under the first modulation mode is greater than the current CNR noise-off threshold, and the average CNR plus the first threshold under the first modulation mode is greater than the current CNR noise-on threshold, the average CNR under the first modulation mode is used as the new CNR noise-off threshold, and the average CNR under the first modulation mode plus the first threshold is used as the new CNR noise-on threshold.

2. The noise control method as described in claim 1, characterized in that, Methods for calculating the carrier-to-noise ratio include: Calculate the sum of energy at each frequency point within each unit of time, using a predetermined time period as the unit. Determine the frequency point and energy value of the maximum energy value; Based on the maximum energy value, it is divided into multiple energy levels; Calculate the noise variable for each energy level, where the noise variable is the count of frequency points where the energy level has not been reached; The noise energy sum is obtained by taking the sum of the energy of the frequency points corresponding to the noise variables, and the maximum energy value is taken as the carrier energy. The difference between the logarithms of the two is used to obtain the carrier-to-noise ratio.

3. The noise control method as described in claim 2, characterized in that, The noise variable is a count of frequency points where the energy level is not reached within the effective signal bandwidth.

4. The noise control method as described in claim 1, characterized in that, The correction of the calculated carrier-to-noise ratio includes: Calculate the carrier-to-noise ratio (CNR) and signal-to-noise ratio (SNR) for each group per unit time under the current carrier center. If the first judgment condition for the instantaneous CNR and SNR is met, the statistical variable N1 under the current carrier center is accumulated; otherwise, the statistical variable N1 is cleared to zero. When the time corresponding to the statistical variable N1 under the current carrier center is greater than the second duration, the final carrier-to-noise ratio value under the current carrier center is assigned as: the average carrier-to-noise ratio value of each energy level with a bandwidth of 4KHz under the current carrier center per unit time.

5. The noise control method as described in claim 4, characterized in that, The correction of the calculated carrier-to-noise ratio also includes: When the sum of the times corresponding to the statistical variable N1 under the current carrier center and the statistical variable N1 under the adjacent carrier center is greater than or equal to the second duration, the final carrier-to-noise ratio (CNR) under the current carrier center and the final CNR under the adjacent carrier center are respectively assigned as: the average CNR of each energy level with a bandwidth of 4KHz under the current carrier center and the adjacent carrier center per unit time.

6. The noise control method as described in claim 5, characterized in that, The sliding window method was used to calculate the download noise ratio per unit time for each group.

7. A noise suppression management system, comprising an FPGA and a DSP, characterized in that, The FPGA is configured to: extract the FFT signal before demodulation of the speech signal, as well as the demodulated speech signal and noise signal; and transmit the FFT signal, speech signal and noise signal to the DSP; The DSP is configured to: calculate the signal-to-noise ratio and carrier-to-noise ratio of the speech signal based on the FFT signal, the speech signal and the noise signal, and compare them with the set signal-to-noise ratio noise-on threshold, signal-to-noise ratio noise-off threshold, carrier-to-noise ratio noise-on threshold and carrier-to-noise ratio noise-off threshold to make noise judgment. And to correct the calculated carrier-to-noise ratio; Dynamically adjust the carrier-to-noise ratio (CNR) noise-on threshold and carrier-to-noise ratio (CNR) noise-off threshold; where: The calculated carrier-to-noise ratio is corrected, including: calculating the carrier-to-noise ratio value under the current unit time; if the difference between the carrier-to-noise ratio value under the current unit time and the carrier-to-noise ratio value under the previous unit time is greater than the first threshold, and the difference between the carrier-to-noise ratio value under the previous unit time and the carrier-to-noise ratio value under the unit time before that does not exceed the first threshold, then the carrier-to-noise ratio value under the previous unit time is used as the carrier-to-noise ratio value under the current unit time. Dynamically adjusting the carrier-to-noise ratio (CNR) noise-on threshold and carrier-to-noise ratio (CNR) noise-off threshold includes: accumulating the statistical variable N2 for each set of CNR values ​​per unit time until it is cleared to zero when the voice signal is transmitted; calculating the average CNR under the first modulation mode within the first duration when the time corresponding to the statistical variable N2 is greater than a first duration; and, with the carrier-to-noise ratio dynamic adjustment switch active, restoring the CNR noise-on / noise-off threshold to the initially set threshold when the average CNR under the first modulation mode is less than or equal to the current CNR noise-off threshold, when the CNR noise-off threshold under the first modulation mode is less than or equal to the current CNR noise-off threshold; and adjusting the CNR noise-on / noise-off threshold when the CNR noise-off threshold under the first modulation mode is less than or equal to the current CNR noise-off threshold. When the average carrier-to-noise ratio (CNR) is greater than the current CNR noise-off threshold, and the average CNR plus the first threshold under the first modulation mode is less than the current CNR noise-on threshold, the average CNR under the first modulation mode is used as the new CNR noise-on threshold. When the average CNR under the first modulation mode is greater than the current CNR noise-off threshold, and the average CNR plus the first threshold under the first modulation mode is greater than the current CNR noise-on threshold, the average CNR under the first modulation mode is used as the new CNR noise-off threshold, and the average CNR under the first modulation mode plus the first threshold is used as the new CNR noise-on threshold.

Images