A signal frequency parameter estimation method and apparatus

Abstract

The application discloses a signal frequency parameter estimation method and device, and relates to the technical field of signal processing. The method comprises the following steps: acquiring a first center frequency, a first bandwidth and a signal-to-noise ratio, and determining a center frequency distribution range and a bandwidth distribution range in combination with a frequency interval; performing band-pass filtering processing and autocorrelation processing on an original signal, and then performing Fourier transform and frequency domain correction processing to obtain a target frequency domain signal; determining a second center frequency in the center frequency distribution range based on the target frequency domain signal, and then determining a corresponding bandwidth lower / upper limit distribution range; and determining a bandwidth start / end frequency in the bandwidth lower / upper limit distribution range based on the target frequency domain signal to obtain a second bandwidth. The application reduces the influence of noise interference through filtering and autocorrelation processing, adaptively limits the distribution range to avoid the limitation of Fourier transform length, and realizes accurate positioning of the frequency parameter through a difference method, so that the accuracy of frequency parameter estimation is significantly improved.

Description

Technical Field

[0001] This application relates to the field of signal processing technology, and in particular to a method and apparatus for estimating signal frequency parameters. Background Technology

[0002] The center frequency and bandwidth of a signal are core parameters describing its frequency domain characteristics. The center frequency refers to the frequency value corresponding to the center of the region where the signal energy is concentrated in the frequency domain. The bandwidth refers to the range of the signal in the frequency domain from the energy's starting frequency to its ending frequency, reflecting the width of the frequency band occupied by the signal. The center frequency and bandwidth of a signal together determine the signal's position and occupied range in the frequency domain. They are the foundation for effective signal analysis and processing in key scenarios such as communication system optimization, radar target detection, and spectrum detection. The accuracy of their estimation directly affects the performance of subsequent stages such as signal identification, demodulation, anti-interference, and spectrum resource allocation.

[0003] In current technology, the estimation of the center frequency and bandwidth of a signal is usually done by the time-frequency diagram observation method. The specific implementation process is as follows: first, the signal is captured by the signal acquisition device, then the acquired signal is subjected to a fast Fourier transform to obtain the corresponding time-frequency diagram, and finally the center frequency and bandwidth of the signal are estimated by manually observing the intuitive features of the time-frequency diagram.

[0004] Therefore, it can be seen that the current signal frequency parameter estimation method based on time-frequency diagram observation is affected by the length of the fast Fourier transform and noise, resulting in large estimation errors of the center frequency and bandwidth, which cannot meet the requirements for high-precision estimation of frequency parameters. Summary of the Invention

[0005] To address the aforementioned issues, this application provides a signal frequency parameter estimation method and apparatus. By filtering and autocorrelation processing, the influence of noise and interference is effectively reduced. Furthermore, by adaptively limiting the distribution range, the resolution limitation caused by the Fourier transform length is avoided. Finally, the frequency parameters are accurately located using the difference method, thereby reducing the estimation error of the center frequency and bandwidth and significantly improving the accuracy of the center frequency and bandwidth estimation, thus meeting the demand for high-precision frequency parameter estimation.

[0006] The embodiments of this application disclose the following technical solutions:

[0007] In a first aspect, an embodiment of this application provides a signal frequency parameter estimation method, comprising:

[0008] Acquire the original signal, and the first center frequency, first bandwidth, and signal-to-noise ratio corresponding to the original signal; wherein the first center frequency and the first bandwidth are determined based on the time-frequency diagram obtained by Fourier transform of the original signal, and the parameters of the Fourier transform include: frequency interval;

[0009] Based on the first center frequency, the first bandwidth, the signal-to-noise ratio, and the frequency interval, the center frequency distribution range and the bandwidth distribution range are determined.

[0010] Based on the bandwidth distribution range, the original signal is subjected to bandpass filtering to obtain a filtered signal, and the filtered signal is subjected to autocorrelation processing to obtain the target signal.

[0011] The target signal is subjected to Fourier transform and frequency domain correction to obtain the target frequency domain signal; wherein, the target frequency domain signal includes multiple frequency points and the power value corresponding to each frequency point;

[0012] Based on the target frequency domain signal, the frequency point with the largest power difference value is determined within the center frequency distribution range to obtain the second center frequency;

[0013] Based on the second center frequency, the first bandwidth, the signal-to-noise ratio, and the frequency interval, the corresponding lower bandwidth distribution range and upper bandwidth distribution range are determined;

[0014] Based on the target frequency domain signal, the frequency point with the largest power difference value is determined in the lower boundary distribution range and the upper boundary distribution range of the bandwidth, respectively, to obtain the bandwidth start frequency and the bandwidth end frequency, and based on the bandwidth start frequency and the bandwidth end frequency, the second bandwidth is obtained.

[0015] In one possible implementation, determining the center frequency distribution range and bandwidth distribution range based on the first center frequency, the first bandwidth, the signal-to-noise ratio, and the frequency interval includes:

[0016] Based on the signal-to-noise ratio, determine the corresponding distribution range parameters;

[0017] Based on the distribution range parameters, the first center frequency, and the frequency interval, the center frequency distribution range is obtained, and based on the distribution range parameters, the first bandwidth, and the frequency interval, the bandwidth distribution range is obtained.

[0018] In one possible implementation, determining the corresponding distribution range parameter based on the signal-to-noise ratio includes:

[0019] Based on the signal-to-noise ratio, the corresponding distribution range parameter is calculated using a preset distribution range calculation formula; wherein the distribution range parameter belongs to [1,4] and is an integer; the preset distribution range calculation formula is as follows:

[0020] Distribution range parameter = floor(preset reference parameter value / signal-to-noise ratio); where floor is the result of dividing the preset reference parameter value by the signal-to-noise ratio and rounding it down.

[0021] In one possible implementation, obtaining the center frequency distribution range based on the distribution range parameter, the first center frequency, and the frequency interval, and obtaining the bandwidth distribution range based on the distribution range parameter, the first bandwidth, and the frequency interval, includes:

[0022] Based on the distribution range parameter and the frequency interval, a range deviation value is obtained; wherein, the range deviation value is the product of the distribution range parameter and the frequency interval;

[0023] Subtracting the range deviation value from the first center frequency yields the lower limit of the center frequency distribution range, and adding the range deviation value to the first center frequency yields the upper limit of the center frequency distribution range, thus obtaining the center frequency distribution range.

[0024] Subtracting the range deviation value from the first bandwidth yields the lower limit of the bandwidth distribution range, and adding the range deviation value to the first bandwidth yields the upper limit of the bandwidth distribution range, thus obtaining the bandwidth distribution range.

[0025] In one possible implementation, the parameters of the Fourier transform further include: a sampling rate; and the bandpass filtering of the original signal based on the bandwidth distribution range to obtain the filtered signal includes:

[0026] The original signal is bandpass filtered using a preset bandpass finite impulse response filter to obtain a filtered signal; wherein the window coefficient of the bandpass finite impulse response filter is obtained based on the bandwidth distribution range and the sampling rate, and the window coefficient is the ratio of the total width of the bandwidth distribution range to half of the sampling rate.

[0027] In one possible implementation, determining the frequency point with the largest power difference value within the center frequency distribution range based on the target frequency domain signal to obtain the second center frequency includes:

[0028] Based on the center frequency distribution range, the corresponding first starting frequency point and first ending frequency point are determined, and the power values ​​of all frequency points between the first starting frequency point and the first ending frequency point are extracted from the target frequency domain signal.

[0029] Calculate the first absolute value difference sequence of the power values ​​of all the extracted frequency points; wherein each element in the first absolute value difference sequence is the absolute value of the difference between the power values ​​of two adjacent frequency points;

[0030] The frequency point corresponding to the element with the largest value in the first absolute value difference sequence is determined to obtain the second center frequency.

[0031] In one possible implementation, determining the corresponding lower bandwidth distribution range and upper bandwidth distribution range based on the second center frequency, the first bandwidth, the signal-to-noise ratio, and the frequency interval includes:

[0032] Based on the signal-to-noise ratio, determine the corresponding distribution range parameters;

[0033] Based on the second center frequency, the first bandwidth, the distribution range parameter, and the frequency interval, the corresponding lower bandwidth distribution range and upper bandwidth distribution range are determined.

[0034] In one possible implementation, determining the corresponding lower bandwidth distribution range and upper bandwidth distribution range based on the second center frequency, the first bandwidth, the distribution range parameter, and the frequency interval includes:

[0035] Based on the distribution range parameter and the frequency interval, a range deviation value is obtained; wherein, the range deviation value is the product of the distribution range parameter and the frequency interval;

[0036] Based on the second center frequency and the first bandwidth, a first reference center value for the lower bound distribution range of the bandwidth and a second reference center value for the upper bound distribution range of the bandwidth are obtained; wherein, the first reference center value is the difference between the second center frequency and half of the first bandwidth, and the second reference center value is the sum between the second center frequency and half of the first bandwidth.

[0037] Based on the first reference center value and the range deviation value, the bandwidth lower bound distribution range is determined; wherein, the lower limit of the bandwidth lower bound distribution range is the first reference center value minus the range deviation value, and the upper limit of the bandwidth lower bound distribution range is the first reference center value plus the range deviation value.

[0038] Based on the second reference center value and the range deviation value, the upper limit distribution range of the bandwidth is determined; wherein, the lower limit of the upper limit distribution range of the bandwidth is the second reference center value minus the reference deviation value, and the upper limit of the upper limit distribution range of the bandwidth is the second reference center value plus the reference deviation value.

[0039] In one possible implementation, the step of determining the frequency point with the largest power difference value based on the target frequency domain signal, respectively within the lower bound distribution range and the upper bound distribution range of the bandwidth, to obtain the bandwidth start frequency and the bandwidth end frequency, and obtaining the second bandwidth based on the bandwidth start frequency and the bandwidth end frequency, includes:

[0040] Based on the bandwidth lower bound distribution range, the corresponding second starting frequency point and second ending frequency point are determined, and the power values ​​of all frequency points between the second starting frequency point and the second ending frequency point are extracted in the target frequency domain signal to obtain the power values ​​of all frequency points within the bandwidth lower bound distribution range.

[0041] Calculate the second absolute value difference sequence of power values ​​for all frequency points within the extracted lower bound of the bandwidth distribution range; wherein each element in the second absolute value difference sequence is the absolute value of the difference between the power values ​​of two adjacent frequency points;

[0042] Determine the frequency point corresponding to the element with the largest value in the second absolute value difference sequence to obtain the bandwidth starting frequency;

[0043] Based on the bandwidth upper bound distribution range, the corresponding third starting frequency point and third ending frequency point are determined, and the power values ​​of all frequency points between the third starting frequency point and the third ending frequency point are extracted in the target frequency domain signal to obtain the power values ​​of all frequency points within the bandwidth upper bound distribution range.

[0044] Calculate the third absolute value difference sequence of power values ​​for all frequency points within the extracted upper bound of the bandwidth distribution range; wherein each element in the third absolute value difference sequence is the absolute value of the difference between the power values ​​of two adjacent frequency points;

[0045] Determine the frequency point corresponding to the element with the largest value in the third absolute value difference sequence to obtain the bandwidth end frequency;

[0046] The second bandwidth is obtained by subtracting the bandwidth start frequency from the bandwidth end frequency.

[0047] Secondly, an embodiment of this application provides a signal frequency parameter estimation device, comprising:

[0048] An initial data acquisition module is used to acquire the original signal, and the first center frequency, first bandwidth, and signal-to-noise ratio corresponding to the original signal; wherein, the first center frequency and the first bandwidth are determined based on the time-frequency diagram obtained by Fourier transform of the original signal, and the parameters of the Fourier transform include: frequency interval;

[0049] The distribution range determination module is used to determine the center frequency distribution range and the bandwidth distribution range based on the first center frequency, the first bandwidth, the signal-to-noise ratio, and the frequency interval;

[0050] The original signal processing module is used to perform bandpass filtering on the original signal based on the bandwidth distribution range to obtain a filtered signal, and to perform autocorrelation processing on the filtered signal to obtain a target signal.

[0051] The signal frequency domain conversion module is used to perform Fourier transform and frequency domain correction processing on the target signal to obtain the target frequency domain signal; wherein, the target frequency domain signal includes multiple frequency points and the power value corresponding to each frequency point;

[0052] The precise center determination module is used to determine the frequency point with the largest power difference value within the center frequency distribution range based on the target frequency domain signal, and obtain the second center frequency;

[0053] The bandwidth range determination module is used to determine the corresponding lower bandwidth distribution range and upper bandwidth distribution range based on the second center frequency, the first bandwidth, the signal-to-noise ratio, and the frequency interval;

[0054] The precise bandwidth determination module is used to determine the frequency point with the largest power difference value based on the target frequency domain signal, respectively, within the lower bandwidth distribution range and the upper bandwidth distribution range, to obtain the bandwidth start frequency and the bandwidth end frequency, and to obtain the second bandwidth based on the bandwidth start frequency and the bandwidth end frequency.

[0055] Compared to existing technologies, this application offers the following advantages: Based on initially acquired parameters such as the first center frequency and first bandwidth, the distribution range of the center frequency and bandwidth is adaptively defined by combining the signal-to-noise ratio and frequency interval, achieving precise constraints on the effective signal range. Then, through dual noise reduction via bandpass filtering and autocorrelation processing, noise interference in the original signal is significantly reduced, resulting in a target signal with higher purity. Subsequently, after obtaining a precise target frequency domain signal through Fourier transform and frequency domain correction, the signal energy abrupt change characteristic of the maximum power difference is utilized to locate the high-precision second center frequency, bandwidth start frequency, and bandwidth end frequency within a preset distribution range, ultimately calculating the second bandwidth. The entire process, through a logical closed loop of adaptive range constraint, noise reduction optimization, and precise feature localization, effectively avoids problems such as coarse estimation of initial parameters and noise interference. The estimation accuracy of both the center frequency and bandwidth is significantly better than traditional methods, making it particularly suitable for complex signal environments such as those with low signal-to-noise ratios, providing reliable support for signal parameter identification in fields such as communication and radar. Attached Figure Description

[0056] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0057] Figure 1A flowchart illustrating a signal frequency parameter estimation method provided in an embodiment of this application;

[0058] Figure 2 A flowchart illustrating the process of determining the center frequency distribution range and bandwidth distribution range provided in this application embodiment;

[0059] Figure 3 A flowchart illustrating the determination of the second center frequency provided in an embodiment of this application;

[0060] Figure 4 A flowchart illustrating the process of determining the lower and upper bounds of bandwidth distribution ranges provided in this application embodiment;

[0061] Figure 5 This is a schematic diagram of a signal frequency parameter estimation device provided in an embodiment of this application. Detailed Implementation

[0062] As described earlier, the center frequency and bandwidth of a signal are core parameters describing its frequency domain characteristics. The center frequency is the reference frequency used in communication systems to describe the center position of a signal's frequency band; it is the center point of the signal's spectral energy distribution. In wireless communication and signal processing, the center frequency determines the channel location, bandwidth calculation, and system tuning reference. Bandwidth describes the range a signal occupies in the frequency domain; it refers to the frequency range covered by a signal or spectrum. Bandwidth determines the signal's transmission rate and information capacity. The accuracy of center frequency and bandwidth estimation directly affects the performance of subsequent stages such as signal identification, demodulation, anti-interference, and spectrum resource allocation. For example, in spectrum detection, the center frequency is a key basis for identifying the frequency band to which a signal belongs and distinguishing different signal sources, while bandwidth can determine whether it exceeds the approved frequency band range, thereby identifying illegal spectrum occupation.

[0063] In current technology, the estimation of the center frequency and bandwidth of a signal is usually done by time-frequency graph observation. The specific process is as follows: the target signal is captured by the signal acquisition device, the acquired target signal is subjected to fast Fourier transform to generate a time-frequency graph that can intuitively reflect the distribution of signal energy in the two-dimensional time-frequency plane; finally, the region where the signal energy is concentrated in the time-frequency graph is identified by manual observation or simple thresholding algorithm, and thus the estimated values ​​of the center frequency and bandwidth of the signal are obtained.

[0064] However, current signal frequency parameter estimation methods based on time-frequency plot observations are affected by the length of the Fast Fourier Transform (FFT) and noise. From the perspective of FFT length, the frequency domain resolution of the FFT is determined by the sampling rate and the FFT length. If the FFT length is short, the frequency domain resolution decreases, the spacing between adjacent frequency points on the frequency axis increases, and the true center frequency of the signal is very likely to fall between two discrete frequency points and cannot be accurately captured. At the same time, the wider frequency point spacing also makes it difficult to accurately define the boundary of the signal bandwidth, requiring only a rough estimate based on the energy distribution of adjacent frequency points, inevitably leading to a large estimation error. Furthermore, with a fixed FFT length, the frequency domain resolution cannot be adaptively adjusted, failing to adapt to signal estimation scenarios with different bandwidths and varying levels of precision, further limiting the estimation accuracy. From the perspective of the impact of noise, noise is superimposed on the original signal in the form of random energy, which is confused with the actual energy distribution of the signal in the time-frequency graph. Under normal signal-to-noise ratio scenarios, the interference of noise on signal characteristics can be partially ignored. However, under low signal-to-noise ratio scenarios, the actual energy of the signal will be masked by strong noise, causing the boundary between signal and noise in the time-frequency graph to become blurred. It is impossible to accurately identify the core area where the signal energy is concentrated, and it is also difficult to distinguish the true start and end positions of the signal bandwidth. It may even be possible to misjudge the false energy peaks of noise as part of the signal, resulting in large estimation errors of the signal frequency parameters.

[0065] In summary, current signal frequency parameter estimation methods based on time-frequency plot observation suffer from low resolution due to the FFT length, resulting in fundamental biases in the initial estimation. Strong noise further amplifies these biases, severely interfering with the true characteristics of the signal in the time-frequency plot. Existing technologies, relying solely on observation of the time-frequency plot or simple threshold judgments, cannot effectively remove noise interference or accurately locate the true frequency domain parameters of the signal, ultimately leading to a significant increase in estimation errors and failing to meet the requirements for high-precision estimation.

[0066] This application provides a signal frequency parameter estimation method, comprising: acquiring an original signal, and a first center frequency, a first bandwidth, and a signal-to-noise ratio (SNR) corresponding to the original signal; determining a center frequency distribution range and a bandwidth distribution range based on the first center frequency, the first bandwidth, the SNR, and a frequency interval; performing bandpass filtering on the original signal based on the bandwidth distribution range to obtain a filtered signal, and performing autocorrelation processing on the filtered signal to obtain a target signal; performing Fourier transform and frequency domain correction processing on the target signal to obtain a target frequency domain signal; wherein, the target frequency domain signal includes multiple frequency points and the power value corresponding to each frequency point; determining the frequency point with the largest power difference within the center frequency distribution range based on the target frequency domain signal to obtain a second center frequency; determining the corresponding lower bandwidth distribution range and upper bandwidth distribution range based on the second center frequency, the first bandwidth, the SNR, and the frequency interval; determining the frequency point with the largest power difference within the lower bandwidth distribution range and the upper bandwidth distribution range based on the target frequency domain signal to obtain a bandwidth start frequency and a bandwidth end frequency, and obtaining a second bandwidth based on the bandwidth start frequency and the bandwidth end frequency. The embodiments of this application effectively reduce the impact of noise and interference through filtering and autocorrelation processing, and avoid the resolution limitation caused by the Fourier transform length by adaptively limiting the distribution range. Then, the frequency parameters are accurately located by using the difference method, which reduces the estimation error of center frequency and bandwidth and significantly improves the accuracy of center frequency and bandwidth estimation, thereby meeting the need for high-precision estimation of frequency parameters.

[0067] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0068] Example 1:

[0069] The following is combined Figures 1-4 This application provides a detailed description of a signal frequency parameter estimation method based on its embodiments.

[0070] like Figure 1 As shown in the figure, a signal frequency parameter estimation method provided in this application includes the following steps:

[0071] S101. Obtain the original signal, and the first center frequency, first bandwidth and signal-to-noise ratio corresponding to the original signal.

[0072] The first center frequency and the first bandwidth are determined based on the time-frequency diagram obtained by Fourier transform of the original signal. The parameters of the Fourier transform include the frequency interval.

[0073] The first center frequency refers to the center frequency value of the concentrated signal energy region in the time-frequency graph generated by the Fourier transform of the original signal, and is initially determined by manual observation or a simple thresholding algorithm. It is a rough estimate of the signal's center frequency. Similarly, the first bandwidth refers to the frequency band covered by the signal energy from start to finish in the time-frequency graph generated by the Fourier transform of the original signal, and is initially determined by manual observation or a simple thresholding algorithm. It is a rough estimate of the signal's bandwidth.

[0074] The Fourier transform is a process that converts a discrete-time signal into a frequency-domain signal (time-frequency diagram). The parameters of the Fourier transform include the frequency interval, which refers to the difference between two adjacent frequency points in the frequency-domain signal generated after the Fourier transform.

[0075] Furthermore, the parameters of the Fourier transform also include: Fourier transform length and sampling rate. The Fourier transform length refers to the number of signal sampling points selected when performing the Fourier transform; the longer the length, the smaller the interval between adjacent frequency points in the frequency domain (i.e., the smaller the frequency domain interval). The sampling rate refers to the number of times the signal acquisition device samples the original signal per unit time, and the unit is usually Hertz.

[0076] To make it easier to understand, the frequency interval of the Fourier transform will be introduced below in conjunction with formula (1).

[0077] df=fs / N (1)

[0078] Where df is the frequency interval of the Fourier transform, fs is the sampling rate of the Fourier transform, and N is the length of the Fourier transform.

[0079] The original signal refers to the unprocessed raw electrical signal directly acquired by signal acquisition equipment (such as antennas and sensors), which exists in the form of a discrete sampling point sequence (i.e., discrete-domain signal).

[0080] The signal-to-noise ratio (SNR or S / N) refers to the ratio of the energy of the useful (effective) signal to the energy of random noise superimposed on the signal, usually expressed in decibels (dB). A higher SNR means less noise interference with the signal, while a lower SNR means greater noise interference.

[0081] S102. Based on the first center frequency, the first bandwidth, the signal-to-noise ratio, and the frequency interval, determine the center frequency distribution range and the bandwidth distribution range.

[0082] The center frequency distribution range refers to the narrow frequency band range adaptively defined by taking the first center frequency as a reference and combining the signal-to-noise ratio and frequency interval. It is the distribution range where the true center frequency of the signal is most likely to exist.

[0083] The bandwidth distribution range refers to the frequency domain interval adaptively defined by taking the frequency band corresponding to the first bandwidth as a benchmark and combining the signal-to-noise ratio and frequency interval. It is the distribution range in which the actual bandwidth of the signal is most likely to exist.

[0084] To make it easier to understand, the following will be combined with... Figure 2 This application provides examples of how the center frequency distribution range and bandwidth distribution range are determined in its embodiments.

[0085] S210. Based on the signal-to-noise ratio, determine the corresponding distribution range parameters.

[0086] The distribution range parameter refers to the key parameter used to adaptively divide the center frequency distribution range and the bandwidth distribution range. In this embodiment, the distribution range parameter is used to quantify the width of the frequency domain parameter distribution range, and can also be understood as: a deviation adjustment index parameter dynamically generated in conjunction with the signal-to-noise ratio, used to define the boundary deviation of the center frequency distribution range and the bandwidth distribution range.

[0087] In one possible implementation, the corresponding distribution range parameter is calculated based on the signal-to-noise ratio using a preset distribution range calculation formula. The distribution range parameter belongs to the range [1, 4] and is an integer.

[0088] The formula for calculating the preset distribution range is shown in formula (2):

[0089] a = floor(M / SNR) (2)

[0090] Where a is the distribution range parameter; M is the preset reference parameter value, which is a preset fixed positive integer; SNR is the signal-to-noise ratio of the signal; floor is the floor function that rounds down the ratio of the preset reference parameter value to the signal-to-noise ratio.

[0091] As can be seen from the above description, in the embodiments of this application, the distribution range parameter a = rounded down (preset reference parameter value / signal-to-noise ratio), and a is an integer in [1,4], that is, the distribution range parameter a is 1, 2, 3 or 4.

[0092] As shown in formula (2), it can be seen that the lower the signal-to-noise ratio (SNR), the larger the distribution range parameter 'a'; the higher the SNR, the smaller the distribution range parameter 'a'. Specifically, a lower SNR indicates a weaker signal and stronger noise. Since noise interferes with the judgment of the spectral peak, the error in the rough estimation of the first center frequency and the first bandwidth based on the time-frequency diagram is larger, requiring a larger distribution range to ensure that the true center frequency and bandwidth are within the distribution range. Therefore, a lower SNR results in a larger distribution range parameter 'a'. Conversely, a higher SNR indicates a stronger signal and weaker noise, resulting in a smaller error in the rough estimation of the first center frequency and the first bandwidth. The distribution range can be appropriately reduced to improve efficiency. Therefore, a higher SNR results in a larger distribution range parameter 'a'.

[0093] Furthermore, the lower limit of the distribution range parameter is set to 1 to ensure that even with a high signal-to-noise ratio, the center frequency and bandwidth roughly estimated from the time-frequency plot may still contain errors due to the picket fence effect of the Fourier transform. Therefore, a distribution range parameter value of 1 preserves a relatively small distribution range for the center frequency and bandwidth, preventing subsequent omissions of true values. The upper limit of the distribution range parameter is 4. If the distribution range parameter exceeds 4, the distribution range of subsequent deviations will be too large, potentially introducing more noise frequency bands and reducing the accuracy of subsequent frequency parameter estimations. An upper limit of 4 for the distribution range parameter represents the optimal upper limit for balancing coverage and noise interference in actual testing.

[0094] In one possible implementation, the preset reference parameter value M is 6, then the distribution range parameter a = floor(6 / SNR), and a ∈ [1,4].

[0095] For example, if SNR=6dB, then the distribution range parameter a=floor(6 / 6)=1; if SNR=4dB, then the distribution range parameter a=floor(6 / 4=1.5)=1; if SNR=1dB, then the distribution range parameter =floor(6 / 1)=6, and if a∈[1,4], then the distribution range parameter is 4.

[0096] S220. The center frequency distribution range is obtained based on the distribution range parameters, the first center frequency, and the frequency interval, and the bandwidth distribution range is obtained based on the distribution range parameters, the first bandwidth, and the frequency interval.

[0097] Specifically, based on the roughly estimated first center frequency and first bandwidth, combined with the distribution range parameters adaptively calculated based on the signal-to-noise ratio and the frequency interval of the Fourier transform, the distribution range that the actual center frequency and bandwidth may fall into (i.e., the center frequency distribution range and bandwidth distribution range) is adaptively determined. This defines the focusing boundary for subsequent accurate estimation of the center frequency band and bandwidth, avoiding the inefficiency caused by blindly searching the entire frequency band, while ensuring that the true parameters (center frequency, bandwidth) of the signal are fully included in low signal-to-noise ratio scenarios, thus balancing search efficiency and estimation accuracy.

[0098] The specific implementation method of S220 in the embodiments of this application is described in detail below with reference to S221-S223:

[0099] S221. Based on the distribution range parameter and frequency interval, the range deviation value is obtained.

[0100] The range deviation is the product of the distribution range parameter and the frequency interval. That is, range deviation = distribution range parameter (a) × frequency interval (df). Specifically, the range deviation reflects the maximum reasonable error range of the rough estimate.

[0101] Specifically, the distribution range parameter is adaptively determined by the signal-to-noise ratio (SNR), reflecting the error tolerance; that is, the lower the SNR, the greater the error. The frequency interval is the smallest scale of the Fourier transform frequency domain analysis, ensuring that the range deviation value matches the frequency domain resolution and avoiding the error range from exceeding the actually resolvable frequency domain interval. The range deviation value (i.e., the product of the distribution range parameter and the frequency interval) ensures both the rationality of the range deviation value and its adaptability.

[0102] For example, assuming the distribution range parameter a=1 and the frequency interval df=30kHz, the range deviation value = 1×30=30kHz; assuming the distribution range parameter a=2 and the frequency interval df=30kHz, the range deviation value = 2×30=60kHz.

[0103] S222. Subtract the range deviation value from the first center frequency to obtain the lower limit of the center frequency distribution range, and add the range deviation value to the first center frequency to obtain the upper limit of the center frequency distribution range, so as to obtain the center frequency distribution range.

[0104] Specifically, the center frequency distribution range is a frequency interval consisting of "first center frequency ± range deviation value," used to limit the search range for subsequent precise estimation of the center frequency. The center frequency distribution range is [first center frequency - range deviation value, first center frequency + range deviation value]. Wherein, the lower limit of the center frequency distribution range = first center frequency - range deviation value, and the upper limit of the center frequency distribution range = first center frequency + range deviation value.

[0105] The center frequency distribution range fc_range = nb_fc + [-a×df, +a×df], where nb_fc is the first center frequency, a is the distribution range parameter, and df is the frequency interval.

[0106] In this embodiment, the first center frequency is a rough estimate. It is affected by the Fourier transform length and noise, and has an error with the true center frequency. By adding or subtracting the range deviation value from the first center frequency, a minimum reasonable frequency range containing the true center frequency is constructed. This not only narrows the subsequent search range (avoiding invalid calculations across the entire frequency band), but also ensures that the true center frequency of the low signal-to-noise ratio signal is within the center frequency distribution range because the signal-to-noise ratio is adaptively adjusted.

[0107] For example, assuming the first center frequency = 433.28MHz and the range deviation value = 30KHz, then the center frequency distribution range = [433.28MHz-30KHz, 433.28MHz+30KHz] = [433.25MHz, 433.31MHz].

[0108] S223. Subtract the range deviation value from the first bandwidth to obtain the lower limit of the bandwidth distribution range, and add the range deviation value to the first bandwidth to obtain the upper limit of the bandwidth distribution range, so as to obtain the bandwidth distribution range.

[0109] Specifically, the bandwidth distribution range is a bandwidth interval consisting of "first bandwidth ± range deviation value", used to limit the search range for subsequent accurate bandwidth estimation. The bandwidth distribution range is [first bandwidth - range deviation value, first bandwidth + range deviation value], where the lower limit of the bandwidth distribution range = first bandwidth - range deviation value, and the upper limit of the bandwidth distribution range = first bandwidth + range deviation value.

[0110] The bandwidth distribution range bw_range = nb_bw + [-a×df, +a×df], where nb_bw is the first bandwidth, a is the distribution range parameter, and df is the frequency interval.

[0111] In this embodiment, the first bandwidth is a rough estimate, which is affected by the Fourier transform length and noise, and deviates from the true bandwidth. By adding or subtracting the range deviation value from the first bandwidth, a minimum reasonable bandwidth range that includes the true bandwidth is constructed, ensuring that the true bandwidth is included within the bandwidth distribution range, while avoiding noise interference caused by an excessively wide range.

[0112] For example, assuming the first bandwidth = 256KHz and the range deviation value = 30KHz, then the bandwidth distribution range = [256KHz-30KHz, 256KHz+30KHz] = [226KHz, 286KHz].

[0113] like Figure 2As shown, this embodiment dynamically adjusts the range deviation value based on the signal-to-noise ratio (SNR) to adapt the distribution range (center frequency distribution range and bandwidth distribution range) to signals with different SNRs, thereby accommodating both strong and weak signal scenarios. The range deviation value is obtained through the range deviation value and frequency interval, and based on the first center frequency and the first bandwidth, it is converted into a distribution range containing the true center frequency and the true bandwidth, providing a clear distribution range for subsequent accurate estimation.

[0114] Furthermore, the center frequency distribution range and bandwidth distribution range are bound to the frequency domain interval, ensuring the frequency domain compatibility of the interval and avoiding the problem that the real parameters cannot be detected because the interval scale does not match the discrete frequency points of the Fourier transform.

[0115] The above combination Figure 2 This application details how the center frequency distribution range and bandwidth distribution range are obtained in its embodiments. The following section continues with further details... Figure 1 This application introduces a signal frequency parameter estimation method provided by an embodiment.

[0116] S103. Based on the bandwidth distribution range, the original signal is subjected to bandpass filtering to obtain the filtered signal, and the filtered signal is subjected to autocorrelation processing to obtain the target signal.

[0117] Bandpass filtering is a frequency-selective signal processing technique that allows signal components within a preset frequency range ("passband") to pass through with almost no attenuation, while significantly attenuating signal components outside this range ("stopband").

[0118] Specifically, the original signal is bandpass filtered based on the bandwidth distribution range to ensure that all possible frequency components related to the frequency parameters in the original signal are completely preserved, while noise unrelated to the frequency parameters is attenuated, providing a clean signal base for subsequent processing.

[0119] In one possible implementation, the original signal is bandpass filtered using a pre-defined bandpass finite impulse response filter to obtain the filtered signal.

[0120] The window coefficient of the bandpass finite impulse response filter is obtained based on the bandwidth distribution range and the sampling rate of the Fourier transform. Specifically, the window coefficient is the ratio of the total width of the bandpass distribution range to half of the sampling rate.

[0121] A bandpass finite impulse response filter (FIR filter) is a frequency selector with a finite impulse response that allows signals within a specific frequency range ("passband") to pass through with almost no attenuation, while significantly attenuating signals outside that range ("stopband").

[0122] The window coefficient is a key parameter used in the design of bandpass finite impulse response filters to correct the impulse response of an ideal filter. Essentially, it modulates the infinitely long ideal impulse response using a finite-length window coefficient.

[0123] Specifically, the window coefficient of the bandpass finite impulse response filter = (total width of the bandwidth distribution range) / (sampling rate / 2). Here, the total width of the bandwidth distribution range reflects the width of the complete frequency domain interval that subsequent signals may occupy. The total width of the bandwidth distribution range = first bandwidth + (upper limit of the bandwidth distribution range - lower limit of the bandwidth distribution range) = first bandwidth + 2 × range deviation value; "2 × range deviation value" is the calculated two-sided error redundancy. Thus, the total width of the bandwidth distribution range covers the maximum frequency range width, ensuring that the true bandwidth is fully included regardless of whether there is a deviation. Here, sampling rate / 2 is also the Nyquist frequency, which is the highest frequency that can be processed without distortion in digital signal processing (according to the Nyquist sampling theorem, the sampling rate fs must be ≥ 2 × the highest frequency of the signal, therefore the highest processable frequency is fs / 2).

[0124] Furthermore, the bandpass range corresponding to the bandpass finite impulse response filter based on the window coefficient "(total bandwidth of the bandwidth distribution range) / (sampling rate / 2)" is: [first center frequency - total bandwidth of the bandwidth distribution range / 2, first center frequency + total bandwidth of the bandwidth distribution range / 2]. In the original signal, the frequency components in the bandpass range can pass through this bandpass finite impulse response filter, while the frequency components outside the bandpass range (such as low-frequency interference and noise) are attenuated to near zero.

[0125] For example, assuming a first bandwidth of 256kHz, a distribution range parameter a of 2, a frequency interval of 30kHz, and a sampling rate of 61.44MHz, the window coefficient of the bandpass finite impulse response filter is approximately 0.01224 (256kHz + 2 × 2 × 30kHz) / (61.44MHz / 2). The window coefficient of 0.01224 indicates that the passband width of the bandpass finite impulse response filter is 1.224% of the Nyquist frequency (30.72MHz), corresponding to an actual passband width of 376kHz (the numerator value). Combined with the first center frequency (assuming a center frequency of 433.28MHz), the passband range is [433.28MHz - 376kHz / 2, 433.28MHz + 376kHz / 2] = [433.092MHz, 433.468MHz]. This passband range ensures that all possible frequency components of the real signal (even with deviations) can pass through the filter without being misfiltered.

[0126] Autocorrelation refers to the dependency between the instantaneous values ​​of a signal at one time and at another. Autocorrelation processing involves calculating the correlation between the signal and itself at different time delays to analyze the periodicity and repetition patterns of the signal. Specifically, autocorrelation processing, also known as autocorrelation operation, is a processing method for discrete-time signals. It analyzes the self-correlation of the signal and suppresses noise interference by calculating the sum of the products of the signal and its delayed complex conjugate.

[0127] To facilitate understanding, the autocorrelation processing (autocorrelation operation) will be introduced below in conjunction with formula (3):

[0128] (3)

[0129] Among them, R x [k] represents the output of the autocorrelation sequence, R x Let [k] represent the autocorrelation of signal x, where [k] represents the autocorrelation value corresponding to the delay k, i.e., the autocorrelation result of signal x after a delay of k sampling points; n is the index of the sampling point, taking values ​​of 0, 1, 2, ..., M-1; M is the length of signal x, i.e., the total number of sampling points contained in signal x; x[n] is the complex value of signal x at the sampling point with index n, i.e., the complex value of signal x at the (n+1)th sampling point; x * To take the complex conjugate of signal x, [nk] represents the sequence of signal x after a delay of k sampling bands, where nk is the index of the delayed sampling point; k is the delay amount, used to measure the degree of delay of the signal itself, and its value ranges from 0, 1, 2, ..., M-1. Different k correspond to different autocorrelation results of Yan Chixia.

[0130] As shown in formula (3), the autocorrelation operation accumulates pointwise by multiplying the signal with its own delay, and utilizes the strong correlation of the useful signal and the weak correlation of the noise to achieve noise reduction and output a clean autocorrelation sequence.

[0131] The target signal is the output sequence after autocorrelation processing of the filtered signal, which is a signal optimized by both "bandpass filtering and autocorrelation processing".

[0132] In one possible implementation, the autocorrelation function xcorr is called to perform a biased autocorrelation estimate on the filtered signal and output the autocorrelation sequence, i.e., the target signal.

[0133] For example, the target signal is shown in formula (4):

[0134] Sig=xcorr(signal_filtered,"biased") (4)

[0135] Where Sig is the target signal, i.e. the output autocorrelation sequence; xcorr is the autocorrelation operation function commonly used in the field of signal processing, as shown in formula (3); signal_filtered is the filtered signal, i.e. the signal after bandpass filtering; biased is the type parameter of autocorrelation encouragement, indicating that biased autocorrelation estimation is used. Biased autocorrelation estimation means that when calculating the autocorrelation sequence, the summation result is not normalized by "dividing by the delay" (or only simple normalization is performed). Therefore, theoretically, the estimation result has statistical bias, but it can retain more stable signal energy characteristics and has higher computational efficiency.

[0136] In this embodiment, the original signal is bandpass filtered based on the bandwidth distribution range to remove out-of-band noise and clutter. However, some random noise remains in the signal, especially in low signal-to-noise ratio scenarios. Therefore, autocorrelation processing is performed on the filtered signal. Through the mechanism of useful signal accumulation enhancement and noise cancellation reduction of autocorrelation processing, the interference of residual noise is significantly reduced, and the frequency domain characteristics of the signal are more prominent, which helps to improve the accuracy of signal frequency parameter estimation.

[0137] S104. Perform Fourier transform and frequency correction on the target signal to obtain the target frequency domain signal.

[0138] The target frequency domain signal includes multiple frequency points and the power value corresponding to each frequency point.

[0139] Specifically, the target signal is converted from the time domain (sampling point sequence) to the frequency domain (complex number sequence) through Fourier transform, revealing the frequency domain component distribution of the target signal. The amplitude value of the Fourier transform result is then normalized by the Fourier transform length to obtain the frequency domain amplitude sequence. Subsequently, frequency correction processing is performed on the frequency domain amplitude sequence, moving the zero-frequency component (DC component) to the center of the array so that the frequency distribution conforms to the intuitive logic of "positive and negative frequency symmetry", resulting in a power value array. Based on the reference frequency, the sampling rate of the Fourier transform, and the Fourier transform length, frequency points corresponding one-to-one with the power value array are generated, and the specific frequency value corresponding to each frequency point is determined. Finally, the frequency point sequence and the corresponding power value array together constitute the target frequency domain signal.

[0140] For ease of understanding, the following uses formulas (5) and (6) as examples, taking the Fast Fourier Transform as an example, to illustrate how the target frequency domain signal is obtained in the embodiments of this application.

[0141] The power value array is obtained through formula (5):

[0142] Psd_signal=fftshift(abs(fft(Sig,M)) / M) (5)

[0143] In this array, Psd_signal is the power value array, representing the core frequency domain data of the output. It has a length of M, with each element corresponding to the energy intensity value at a specific frequency point. Sig is the target signal; M is the length of the Fast Fourier Transform (FFT), typically a power of 2, such as 1024 or 2048; fft is the Fast Fourier Transform, where fft(Sig, M) converts the target signal Sig into a complex frequency domain sequence of length M; abs is the absolute value (modulo) operation, used to take the modulus of the complex output of the FFT, stripping phase information and retaining only the amplitude values ​​of each frequency component; / M is the normalization operation, dividing the amplitude value by the FFT length M to eliminate the influence of the transform length on the energy level, making the power values ​​under different FFT lengths comparable; and fftshift is the frequency correction operation.

[0144] The frequency point sequence is calculated using formula (6) and aligned with the power value array:

[0145] Sig_freqs=wb_fc-fs / 2+(0:M-1)*fs / M (6)

[0146] Here, Sig_freqs is a frequency point sequence, an array of length M, where each element is the specific frequency point corresponding to the power value of Psd_signal, completely covering the frequency domain distribution range of the signal; wb_fc is the reference frequency, the center frequency (reference value) of the bandwidth distribution range, used to align the frequency axis with the actual frequency domain position of the signal, ensuring the accuracy of frequency point calculation; fs is the sampling rate of the Fast Fourier Transform, -fs / 2 is used to cooperate with the frequency correction logic of fftshift, adjusting the starting point of the frequency axis from wb_fc to wb_fc-fs / 2, matching the "symmetric positive and negative frequency" distribution of Psd_signal; (0:M-1) is the index sequence, a continuous integer sequence from 0 to M-1 (a total of M frequency points), ensuring that the frequency points correspond one-to-one with Psd_signal; fs / M is the frequency interval, that is, the difference between two adjacent frequency points.

[0147] As can be seen from formulas (5) and (6), Sig_freqs specifies the specific frequency corresponding to each frequency point in the target signal, that is, each power value; Psd_signal specifies the power value corresponding to each frequency point in the target signal.

[0148] S105. Based on the target frequency domain signal, determine the frequency point with the largest power difference value within the center frequency distribution range to obtain the second center frequency.

[0149] The power difference value refers to the absolute value of the difference between the power values ​​of two adjacent frequency points in the target frequency domain signal, and is used to quantify the power change amplitude between adjacent frequency points. Specifically, the larger the power difference value, the more drastic the power change between the two adjacent frequency points, which usually corresponds to the abrupt boundary of signal energy from the non-peak region to the peak region (or from the peak region to the non-peak region); conversely, the smaller the power difference value, the more gradual the power change between the two adjacent frequency points.

[0150] The second center frequency is a high-precision estimate of the true center frequency of the signal obtained by locating the frequency point with the largest power difference within the center frequency distribution range. Compared with the first center frequency, the second center frequency further eliminates noise interference through range constraints and differential extreme value location, resulting in higher accuracy.

[0151] To make it easier to understand, the following will be combined with... Figure 3 This application provides a detailed description of how the second center frequency is obtained in the embodiments of this application.

[0152] S301. Based on the center frequency distribution range, determine the corresponding first starting frequency point and first ending frequency point, and extract the power values ​​of all frequency points between the first starting frequency point and the first ending frequency point in the target frequency domain signal.

[0153] The first starting frequency point (start_frame) is the first frequency point in the target frequency domain signal corresponding to the center frequency distribution range, that is, the frequency point in the target frequency domain signal Sig_freqs that satisfies Sig_freqs[i]≥ the lower limit of the center frequency distribution range.

[0154] The second end frequency point (end_frame) is the last frequency point in the target frequency domain signal within the center frequency distribution range, that is, the frequency point in the target frequency domain signal Sig_freqs that satisfies Sig_freqs[j]≥the upper limit of the center frequency distribution range.

[0155] Specifically, based on the first starting frequency point and the first ending frequency point, the corresponding sub-frequency point sequence is extracted from the target frequency domain signal. The sub-frequency point sequence includes all frequency points from the first starting frequency point to the first ending frequency point, as well as their respective power values.

[0156] S302. Calculate the first absolute value difference sequence of the power values ​​at all extracted frequency points.

[0157] In this sequence, each element of the first absolute value difference sequence is the absolute value of the difference between the power values ​​of two adjacent frequency points.

[0158] Specifically, the absolute difference between two adjacent power values ​​in the sub-frequency point sequence formed by all extracted frequency points is calculated to generate the first absolute value difference sequence. In this embodiment, the power difference value is used to quantify the power change amplitude of adjacent frequency points, highlighting the abrupt change region of signal energy (i.e., near the center frequency).

[0159] S303. Determine the frequency point corresponding to the element with the largest value in the first absolute value sequence to obtain the second center frequency.

[0160] Specifically, find the element with the largest value (D_max) in the first absolute value difference sequence D_range; the element with the largest value corresponds to the pair of adjacent frequency points in the sub-frequency point sequence where the power change is most drastic, indicating that this position is the critical point where the signal energy enters the peak region from the non-peak region (or from the peak region to the non-peak region), that is, the location of the true center frequency; determine the frequency corresponding to the previous frequency point in the pair of adjacent frequency points corresponding to the element with the largest value as the second center frequency.

[0161] In this embodiment, the center frequency of the signal corresponds to the core position of the energy amplitude region. The adjacent frequency point pair with the largest power difference is the critical point where the signal energy rapidly rises from the non-peak region to the peak region (or the critical point where the signal energy rapidly falls from the peak region to the non-peak region). The frequency point preceding this critical point is exactly the beginning or end boundary of the peak region, which is closer to the physical characteristics of the true center frequency of the signal.

[0162] The above combination Figure 3 This application details how the second center frequency is obtained in its embodiments. The following section continues with further examples. Figure 1 This paper introduces a signal frequency parameter estimation method provided in the embodiments of this application.

[0163] S106. Based on the second center frequency, the first bandwidth, the signal-to-noise ratio, and the frequency interval, determine the corresponding lower bandwidth distribution range and upper bandwidth distribution range.

[0164] The bandwidth lower bound distribution range is a frequency domain interval defined to accurately locate the lower bound of the signal's true bandwidth. It is based on the lower bound center derived from the second center frequency and the first bandwidth, and the boundary is adjusted by the deviation amount generated adaptively by the signal-to-noise ratio and frequency interval. This forms a reasonable search range that includes the lower bound of the true bandwidth. It can avoid missing the lower bound of the true bandwidth due to noise interference and prevent the introduction of irrelevant noise due to an excessively wide range. It serves as a constraint boundary for subsequent location of the lower bound of the true bandwidth.

[0165] Among them, the bandwidth upper bound distribution range is the frequency domain interval defined to accurately locate the true upper bound of the signal bandwidth. It is based on the bandwidth upper bound center derived from the second center frequency and the first bandwidth, and the boundary is adjusted by the deviation amount generated adaptively by the signal-to-noise ratio and frequency interval. It forms a reasonable search range that includes the true upper bound of the bandwidth. It can avoid missing the true upper bound of the bandwidth due to noise interference, and prevent the introduction of irrelevant noise due to an excessively wide range. It is the constraint boundary for subsequent location of the true upper bound of the bandwidth.

[0166] Specifically, based on the signal-to-noise ratio, the corresponding distribution range parameters are determined; based on the second center frequency, the first bandwidth, the distribution range parameters, and the frequency interval, the corresponding lower bandwidth distribution range and upper bandwidth distribution range are determined.

[0167] To make it easier to understand, the following will be combined with... Figure 4 This application provides a detailed description of how the lower and upper bounds of bandwidth are obtained in the embodiments of this application.

[0168] S401. Based on the distribution range parameter and frequency interval, obtain the range deviation value.

[0169] The range deviation is the product of the distribution range parameter and the frequency interval. That is, range deviation = distribution range parameter (a) × frequency interval (df). Specifically, the range deviation reflects the maximum reasonable error range of the rough estimate.

[0170] S402. Based on the second center frequency and the first bandwidth, obtain the first reference center value of the lower bandwidth distribution range and the second reference center value of the upper bandwidth distribution range.

[0171] The first reference center value is the difference between the second center frequency and half of the first bandwidth. It can be understood that the first reference center value is the core anchor point for defining the lower bound of the bandwidth distribution range. All deviation adjustments to the lower bound of the bandwidth distribution range revolve around this first reference center value, ensuring that the lower bound of the bandwidth distribution range is always anchored to the approximate location of the lower bound of the signal's true bandwidth.

[0172] Specifically, the first reference center value = the second center frequency - the first bandwidth / 2.

[0173] The second reference center value is the sum of the second center frequency and half of the first bandwidth. It can be understood that the second reference center value is the core anchor point for defining the upper bound of the bandwidth distribution range. All deviation adjustments for the lower bound of the bandwidth distribution range revolve around this second reference center value, ensuring that the upper bound of the bandwidth distribution range is always anchored to the approximate location of the true upper bound of the signal bandwidth.

[0174] Specifically, the second reference center value = the second center frequency + the first bandwidth / 2.

[0175] For example, assuming the second center frequency is 800Hz and the first bandwidth is 120Hz, then the first reference center value = 800 - 120 / 2 = 740Hz, and the second reference center value = 800 + 120 / 2 = 860Hz. This can be simply interpreted as: using the high-precision center frequency of 800Hz as the axis of symmetry, shifting 60Hz (half the bandwidth) towards the lower frequency side yields the first reference center value of 740Hz for the lower bandwidth distribution range, and shifting 60Hz towards the higher frequency side yields the second reference center value of 860Hz for the upper bandwidth distribution range.

[0176] S403. Determine the lower bound distribution range of the bandwidth based on the first reference center value and the range deviation value.

[0177] The lower limit of the bandwidth distribution range is the first reference center value minus the range deviation value, and the upper limit of the bandwidth distribution range is the first reference center value plus the range deviation value.

[0178] Specifically, the bandwidth lower bound distribution range is a frequency domain interval formed by extending the range deviation value to the low-frequency side and the high-frequency side with the first reference center value as the core. It is the constrained search range of the true bandwidth lower bound of the positioning signal. Its upper and lower limits are strictly symmetrically distributed around the first reference center value to ensure coverage of all possible positions of the true bandwidth lower bound under noise interference.

[0179] For example, assuming the first reference center value is fc1 and the range deviation value is a×df, then the lower limit of the bandwidth lower bound distribution range is fc1-a×df, and the upper limit of the bandwidth lower bound distribution range is fc1+a×df. Therefore, the bandwidth lower bound distribution range is [fc1-a×df, fc1+a×df].

[0180] S404. Determine the upper limit distribution range of the bandwidth based on the second reference center value and the range deviation value.

[0181] The lower limit of the bandwidth upper bound distribution range is the second reference center value minus the range deviation value, and the upper limit of the bandwidth upper bound distribution range is the second reference center value plus the range deviation value.

[0182] Specifically, the upper limit of the bandwidth distribution range is the frequency domain interval formed by extending the range deviation value to the low-frequency side and the high-frequency side with the second reference center value as the core. It is the constrained search range of the upper limit of the true bandwidth of the positioning signal. Its upper and lower limits are strictly symmetrically distributed around the second reference center value to ensure coverage of all possible positions of the upper limit of the true bandwidth under noise interference.

[0183] For example, assuming the first reference center value is fc2 and the range deviation value is a×df, then the lower limit of the bandwidth upper bound distribution range is fc2-a×df, and the upper limit of the bandwidth upper bound distribution range is fc2+a×df. Therefore, the bandwidth upper bound distribution range is [fc2-a×df, fc2+a×df].

[0184] The above combination Figure 4 This application details how the lower and upper bounds of bandwidth are obtained in its embodiments. The following section continues with further details... Figure 1 This paper introduces a signal frequency parameter estimation method provided in the embodiments of this application.

[0185] S107. Based on the target frequency domain signal, determine the frequency point with the largest power difference value in the lower and upper bound distribution ranges of the bandwidth respectively, obtain the bandwidth start frequency and bandwidth end frequency, and obtain the second bandwidth based on the bandwidth start frequency and bandwidth end frequency.

[0186] The bandwidth start frequency is the low-frequency side boundary of the predicted signal's true bandwidth, obtained by locating the frequency point with the largest power difference within the bandwidth lower boundary distribution range, and is the low-frequency boundary for calculating the second bandwidth; the bandwidth end frequency is the high-frequency side boundary of the predicted signal's true bandwidth, obtained by locating the frequency point with the largest power difference within the bandwidth upper boundary distribution range, and is the high-frequency boundary for the second bandwidth.

[0187] The second bandwidth is a high-precision estimate of the signal bandwidth, obtained by subtracting the starting frequency of the bandwidth from the ending frequency of the bandwidth.

[0188] Specifically, for the bandwidth starting frequency, based on the bandwidth lower bound distribution range, the corresponding second starting frequency point and second ending frequency point are determined, and the power values ​​of all frequency points between the second starting frequency point and the second ending frequency are extracted in the target frequency domain signal to obtain the power of all frequency points within the bandwidth lower bound distribution range; the second absolute value difference sequence of the extracted power values ​​of all frequency points within the bandwidth lower bound distribution range is calculated; the frequency point corresponding to the element with the largest value in the second absolute value difference sequence is determined to obtain the bandwidth starting frequency.

[0189] The second starting frequency point is the first frequency point in the target frequency domain signal corresponding to the lower bound of the bandwidth distribution range. The second ending frequency point is the last frequency point in the target frequency domain signal corresponding to the lower bound of the bandwidth distribution range.

[0190] The second absolute value difference sequence is the absolute difference sequence of power values ​​between adjacent frequency points within the lower bound of the bandwidth distribution range. Each element in the second absolute value difference sequence is the absolute value between the power values ​​of two adjacent frequency points, which is used to characterize the power variation amplitude within the lower bound of the bandwidth distribution range.

[0191] In one possible implementation, the frequency value of the preceding frequency point in the frequency point pair corresponding to the maximum value in the second absolute value difference sequence is determined as the bandwidth starting frequency.

[0192] Specifically, for the bandwidth end frequency, based on the bandwidth upper bound distribution range, the corresponding third starting frequency point and third ending frequency point are determined, and the power values ​​of all frequency points between the third starting frequency point and the third ending frequency point are extracted in the target frequency domain signal to obtain the power values ​​of all frequency points within the bandwidth upper bound distribution range; the third absolute value difference sequence of the extracted power values ​​of all frequency points within the bandwidth upper bound distribution range is calculated; the frequency point corresponding to the element with the largest value in the third absolute value difference sequence is determined to obtain the bandwidth end frequency.

[0193] The third starting frequency point is the first frequency point in the target frequency domain signal corresponding to the upper bound of the bandwidth distribution range. The third ending frequency point is the last frequency point in the target frequency domain signal corresponding to the upper bound of the bandwidth distribution range.

[0194] Among them, the third absolute value difference sequence is the difference sequence of power values ​​between adjacent frequency points within the upper bound of the bandwidth distribution range. Each element in the third absolute value difference sequence is the difference between the power values ​​of two adjacent frequency points, which is used to characterize the power change amplitude within the upper bound of the bandwidth distribution range.

[0195] In one possible implementation, the frequency value of the preceding frequency point in the frequency point pair corresponding to the maximum value in the third absolute value difference sequence is determined as the bandwidth end frequency.

[0196] In this embodiment, the bandwidth boundary is the abrupt change point where signal energy transitions from in-band (high power) to out-of-band (low power) and from out-of-band to in-band. The power difference between adjacent signals at this location is the largest, thus the bandwidth boundary can be accurately located using the differential extrema. The distribution range of the lower / upper bandwidth limits the search interval, avoiding the introduction of irrelevant noise during full-frequency domain search, while also covering the fluctuation range of the bandwidth boundary under noise interference. The bandwidth boundary is located using a high-precision center frequency (second center frequency) and the differential extrema to obtain the second bandwidth. Compared to the first bandwidth, this effectively avoids the influence of noise and time-frequency map resolution, significantly improving accuracy.

[0197] This application provides a signal frequency parameter estimation method, comprising: acquiring an original signal, and a first center frequency, a first bandwidth, and a signal-to-noise ratio (SNR) corresponding to the original signal; determining a center frequency distribution range and a bandwidth distribution range based on the first center frequency, the first bandwidth, the SNR, and a frequency interval; performing bandpass filtering on the original signal based on the bandwidth distribution range to obtain a filtered signal, and performing autocorrelation processing on the filtered signal to obtain a target signal; performing Fourier transform and frequency domain correction processing on the target signal to obtain a target frequency domain signal; wherein the target frequency domain signal includes multiple frequency points and the power value corresponding to each frequency point; determining the frequency point with the largest power difference within the center frequency distribution range based on the target frequency domain signal to obtain a second center frequency; determining the corresponding lower bandwidth distribution range and upper bandwidth distribution range based on the second center frequency, the first bandwidth, the SNR, and the frequency interval; determining the frequency point with the largest power difference within the lower bandwidth distribution range and the upper bandwidth distribution range based on the target frequency domain signal to obtain a bandwidth start frequency and a bandwidth end frequency, and obtaining a second bandwidth based on the bandwidth start frequency and the bandwidth end frequency. This application's embodiments, based on initially acquired parameters such as the first center frequency and first bandwidth, adaptively define the center frequency and bandwidth distribution range by combining signal-to-noise ratio and frequency interval, achieving precise constraints on the effective signal range. Then, through dual noise reduction via bandpass filtering and autocorrelation processing, noise interference in the original signal is significantly reduced, resulting in a target signal with higher purity. Subsequently, after obtaining a precise target frequency domain signal through Fourier transform and frequency domain correction, the signal energy abrupt change characteristic of the maximum power difference is utilized to locate the high-precision second center frequency, bandwidth start frequency, and bandwidth end frequency within a preset distribution range, ultimately calculating the second bandwidth. The entire process, through a logical closed loop of adaptive range constraint, noise reduction optimization, and precise feature localization, effectively avoids problems such as coarse estimation of initial parameters and noise interference. The estimation accuracy of both the center frequency and bandwidth is significantly better than traditional methods, making it particularly suitable for complex signal environments such as low signal-to-noise ratio, providing reliable support for signal parameter identification in fields such as communication and radar.

[0198] Furthermore, by adaptively calculating the distribution range parameter using the signal-to-noise ratio (SNR), a precise and interference-resistant interval constraint is provided for subsequent signal processing. First, based on the SNR and preset reference parameter values, a floor-rounded operation is performed to obtain the distribution range parameter, limited to the integer interval [1, 4]. This ensures that the parameter can adapt to different noise intensities: the lower the SNR (the stronger the noise), the larger the distribution range parameter, and vice versa. Then, combining this distribution range parameter with the frequency interval, the first center frequency and the first bandwidth are extended to obtain the center frequency distribution range and bandwidth distribution range, respectively, achieving dynamic matching between noise intensity and the width of the distribution range. This avoids the problem of missing effective signals due to an excessively narrow fixed range, and also prevents irrelevant noise interference introduced by an excessively wide range. This lays a reliable interval foundation for subsequent bandpass filtering, autocorrelation processing, and precise frequency parameter positioning, significantly improving the stability and accuracy of signal frequency parameter estimation in complex scenarios such as low SNR.

[0199] Furthermore, the range deviation value is obtained by multiplying the distribution range parameter by the frequency interval. The distribution range is then symmetrically expanded based on the first center frequency and the first bandwidth to achieve the quantification and precision of the interval constraint, ensuring that the range is adapted to the frequency domain resolution and avoiding the omission of effective signals or the introduction of noise.

[0200] Furthermore, the window coefficients of the bandpass finite impulse response filter are determined based on the bandwidth distribution range and sampling rate, so that the filter parameters are precisely matched with the signal bandwidth, improving the targeting of the bandpass filter and effectively filtering out out-of-band interference.

[0201] Furthermore, by extracting power values ​​within the center frequency distribution range, calculating the absolute value difference sequence, and locating the frequency point corresponding to the maximum value, the location of signal energy abrupt changes is accurately captured, achieving high-precision estimation of the second center frequency. Simultaneously, the signal-to-noise ratio (SNR) adaptation distribution range parameters are reused, combined with the high-precision second center frequency, to determine the upper and lower bounds of the bandwidth distribution range. This makes the bandwidth search interval more closely match the actual frequency domain location of the signal, improving the accuracy of subsequent bandwidth estimation. Specifically, the reference center values ​​of the upper and lower bounds of the bandwidth are derived from the second center frequency, and the distribution range is symmetrically expanded using the range deviation value. This both anchors the approximate location of the bandwidth boundary and covers fluctuations caused by noise through adaptive deviation, ensuring the rationality of the search range.

[0202] Example 2:

[0203] The following is combined Figure 5 This application provides a detailed description of a signal frequency parameter estimation device according to its embodiments.

[0204] like Figure 5 As shown in the figure, the signal frequency parameter estimation device provided in this application includes the following modules:

[0205] The initial data acquisition module 501 is used to acquire the original signal, the first center frequency, the first bandwidth, and the signal-to-noise ratio corresponding to the original signal; wherein, the first center frequency and the first bandwidth are determined based on the time-frequency diagram obtained by Fourier transform of the original signal, and the parameters of the Fourier transform include: frequency interval;

[0206] The distribution range determination module 502 is used to determine the center frequency distribution range and the bandwidth distribution range based on the first center frequency, the first bandwidth, the signal-to-noise ratio, and the frequency interval;

[0207] The original signal processing module 503 is used to perform bandpass filtering on the original signal based on the bandwidth distribution range to obtain the filtered signal, and to perform autocorrelation processing on the filtered signal to obtain the target signal.

[0208] The signal frequency domain conversion module 504 is used to perform Fourier transform and frequency domain correction processing on the target signal to obtain the target frequency domain signal; wherein, the target frequency domain signal includes multiple frequency points and the power value corresponding to each frequency point;

[0209] The precise center determination module 505 is used to determine the frequency point with the largest power difference value within the center frequency distribution range based on the target frequency domain signal, and obtain the second center frequency.

[0210] The bandwidth range determination module 506 is used to determine the corresponding lower bandwidth distribution range and upper bandwidth distribution range based on the second center frequency, the first bandwidth, the signal-to-noise ratio and the frequency interval.

[0211] The precise bandwidth determination module 507 is used to determine the frequency point with the largest power difference value based on the target frequency domain signal, respectively within the lower and upper bound distribution ranges of the bandwidth, to obtain the bandwidth start frequency and bandwidth end frequency, and to obtain the second bandwidth based on the bandwidth start frequency and bandwidth end frequency.

[0212] In one possible implementation, the distribution range determination module 502 is specifically used to determine the corresponding distribution range parameters based on the signal-to-noise ratio; obtain the center frequency distribution range based on the distribution range parameters, the first center frequency and the frequency interval; and obtain the bandwidth distribution range based on the distribution range parameters, the first bandwidth and the frequency interval.

[0213] In one possible implementation, the distribution range determination module 502 is specifically used to calculate the corresponding distribution range parameter based on the signal-to-noise ratio using a preset distribution range calculation formula; wherein the distribution range parameter belongs to [1,4] and is an integer; and the preset distribution range calculation formula is as follows:

[0214] Distribution range parameter = floor(preset reference parameter value / signal-to-noise ratio); where floor is the result of dividing the preset reference parameter value by the signal-to-noise ratio and rounding it down.

[0215] In one possible implementation, the distribution range determination module 502 is specifically used to obtain a range deviation value based on the distribution range parameter and the frequency interval; wherein, the range deviation value is the product of the distribution range parameter and the frequency interval; subtracting the range deviation value from the first center frequency to obtain the lower limit of the center frequency distribution range, and adding the range deviation value to the first center frequency to obtain the upper limit of the center frequency distribution range, so as to obtain the center frequency distribution range; subtracting the range deviation value from the first bandwidth to obtain the lower limit of the bandwidth distribution range, and adding the range deviation value to the first bandwidth to obtain the upper limit of the bandwidth distribution range, so as to obtain the bandwidth distribution range.

[0216] In one possible implementation, the original signal processing module is specifically used to perform bandpass filtering on the original signal through a preset bandpass finite impulse response filter to obtain the filtered signal; wherein, the window coefficient of the bandpass finite impulse response filter is obtained based on the bandwidth distribution range and the sampling rate, and the window coefficient is the ratio of the total width of the bandwidth distribution range to half of the sampling rate.

[0217] In one possible implementation, the precise center determination module 505 is specifically used to determine the corresponding first starting frequency point and first ending frequency point based on the center frequency distribution range, and extract the power values ​​of all frequency points between the first starting frequency point and the first ending frequency point in the target frequency domain signal; calculate the first absolute value difference sequence of the power values ​​of all extracted frequency points; wherein each element in the first absolute value difference sequence is the absolute value of the difference between the power values ​​of two adjacent frequency points; determine the frequency point corresponding to the element with the largest value in the first absolute value difference sequence to obtain the second center frequency.

[0218] In one possible implementation, the bandwidth range determination module 506 is specifically used to determine the corresponding distribution range parameters based on the signal-to-noise ratio; and to determine the corresponding lower bandwidth distribution range and upper bandwidth distribution range based on the second center frequency, the first bandwidth, the distribution range parameters, and the frequency interval.

[0219] In one possible implementation, the bandwidth range determination module 506 specifically includes: a range deviation determination module, a reference center determination module, a lower limit range determination module, and an upper limit range determination module.

[0220] The range deviation determination module is used to obtain the range deviation value based on the distribution range parameter and the frequency interval; wherein, the range deviation value is the product of the distribution range parameter and the frequency interval;

[0221] The reference center determination module is used to obtain a first reference center value for the lower bound distribution range of the bandwidth and a second reference center value for the upper bound distribution range of the bandwidth based on the second center frequency and the first bandwidth; wherein, the first reference center value is the difference between the second center frequency and half of the first bandwidth, and the second reference center value is the sum between the second center frequency and half of the first bandwidth.

[0222] The lower bound range determination module is used to determine the lower bound distribution range of bandwidth based on the first reference center value and the range deviation value; wherein, the lower limit of the lower bound distribution range of bandwidth is the first reference center value minus the range deviation value, and the upper limit of the lower bound distribution range of bandwidth is the first reference center value plus the range deviation value.

[0223] The upper bound range determination module is used to determine the upper bound distribution range of bandwidth based on the second reference center value and the range deviation value; wherein, the lower limit of the upper bound distribution range of bandwidth is the second reference center value minus the reference deviation value, and the upper limit of the upper bound distribution range of bandwidth is the second reference center value plus the reference deviation value.

[0224] In one possible implementation, the precise bandwidth determination module 507 includes: a lower bound frequency point extraction module, a second absolute value difference module, a bandwidth start frequency determination module, an upper bound frequency point extraction module, a third absolute value difference module, a bandwidth end frequency determination module, and a second bandwidth determination module.

[0225] The lower bound frequency point extraction module is used to determine the corresponding second starting frequency point and second ending frequency point based on the bandwidth lower bound distribution range, and extract the power values ​​of all frequency points between the second starting frequency point and the second ending frequency point in the target frequency domain signal to obtain the power values ​​of all frequency points within the bandwidth lower bound distribution range.

[0226] The second absolute value difference module is used to calculate the second absolute value difference sequence of power values ​​at all frequency points within the extracted lower bound of the bandwidth distribution range; wherein each element in the second absolute value difference sequence is the absolute value of the difference between the power values ​​of two adjacent frequency points;

[0227] The bandwidth starting frequency determination module is used to determine the frequency point corresponding to the element with the largest value in the second absolute value difference sequence, and thus obtain the bandwidth starting frequency.

[0228] The upper bound frequency point extraction module is used to determine the corresponding third starting frequency point and third ending frequency point based on the upper bound distribution range of the bandwidth, and extract the power values ​​of all frequency points between the third starting frequency point and the third ending frequency point in the target frequency domain signal to obtain the power values ​​of all frequency points within the upper bound distribution range of the bandwidth.

[0229] The third absolute value difference module is used to calculate the third absolute value difference sequence of power values ​​at all frequency points within the extracted upper bound of the bandwidth distribution range; where each element in the third absolute value difference sequence is the absolute value of the difference between the power values ​​of two adjacent frequency points;

[0230] The bandwidth end frequency determination module is used to determine the frequency point corresponding to the element with the largest value in the third absolute value difference sequence, and obtain the bandwidth end frequency.

[0231] The second bandwidth determination module is used to subtract the bandwidth start frequency from the bandwidth end frequency to obtain the second bandwidth.

[0232] This application provides a signal frequency parameter estimation device, including: an initial data acquisition module 501, used to acquire an original signal, a first center frequency, a first bandwidth, and a signal-to-noise ratio corresponding to the original signal; a distribution range determination module 502, used to determine the center frequency distribution range and the bandwidth distribution range based on the first center frequency, the first bandwidth, the signal-to-noise ratio, and the frequency interval; an original signal processing module 503, used to perform bandpass filtering on the original signal based on the bandwidth distribution range to obtain a filtered signal, and to perform autocorrelation processing on the filtered signal to obtain a target signal; and a signal frequency domain conversion module 504, used to perform Fourier transform and frequency domain correction on the target signal. The system processes the signal to obtain the target frequency domain signal. A precise center determination module 505 determines the frequency point with the largest power difference within the center frequency distribution range based on the target frequency domain signal, thus obtaining the second center frequency. A bandwidth range determination module 506 determines the corresponding lower and upper bandwidth distribution ranges based on the second center frequency, the first bandwidth, the signal-to-noise ratio, and the frequency interval. A precise bandwidth determination module 507 determines the frequency point with the largest power difference within the lower and upper bandwidth distribution ranges respectively based on the target frequency domain signal, thus obtaining the bandwidth start frequency and bandwidth end frequency, and obtains the second bandwidth based on the bandwidth start frequency and bandwidth end frequency. This application's embodiments, based on initially acquired parameters such as the first center frequency and first bandwidth, adaptively define the center frequency and bandwidth distribution range by combining signal-to-noise ratio and frequency interval, achieving precise constraints on the effective signal range. Then, through dual noise reduction via bandpass filtering and autocorrelation processing, noise interference in the original signal is significantly reduced, resulting in a target signal with higher purity. Subsequently, after obtaining a precise target frequency domain signal through Fourier transform and frequency domain correction, the signal energy abrupt change characteristic of the maximum power difference is utilized to locate the high-precision second center frequency, bandwidth start frequency, and bandwidth end frequency within a preset distribution range, ultimately calculating the second bandwidth. The entire process, through a logical closed loop of adaptive range constraint, noise reduction optimization, and precise feature localization, effectively avoids problems such as coarse estimation of initial parameters and noise interference. The estimation accuracy of both the center frequency and bandwidth is significantly better than traditional methods, making it particularly suitable for complex signal environments such as low signal-to-noise ratio, providing reliable support for signal parameter identification in fields such as communication and radar.

[0233] It should be noted that the various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, for the device embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the description of the method embodiments. The device embodiments described above are merely illustrative, and the units described as separate components may or may not be physically separate. The components indicated as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment solution according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0234] The above description is merely one specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for estimating signal frequency parameters, characterized in that, include: Acquire the original signal, and the first center frequency, first bandwidth, and signal-to-noise ratio corresponding to the original signal; wherein the first center frequency and the first bandwidth are determined based on the time-frequency diagram obtained by Fourier transform of the original signal, and the parameters of the Fourier transform include: frequency interval; Based on the first center frequency, the first bandwidth, the signal-to-noise ratio, and the frequency interval, the center frequency distribution range and the bandwidth distribution range are determined. Based on the bandwidth distribution range, the original signal is subjected to bandpass filtering to obtain a filtered signal, and the filtered signal is subjected to autocorrelation processing to obtain the target signal. The target signal is subjected to Fourier transform and frequency domain correction to obtain the target frequency domain signal; wherein, the target frequency domain signal includes multiple frequency points and the power value corresponding to each frequency point; Based on the target frequency domain signal, the frequency point with the largest power difference value is determined within the center frequency distribution range to obtain the second center frequency; Based on the second center frequency, the first bandwidth, the signal-to-noise ratio, and the frequency interval, the corresponding lower bandwidth distribution range and upper bandwidth distribution range are determined; Based on the target frequency domain signal, the frequency point with the largest power difference value is determined in the lower boundary distribution range and the upper boundary distribution range of the bandwidth, respectively, to obtain the bandwidth start frequency and the bandwidth end frequency, and based on the bandwidth start frequency and the bandwidth end frequency, the second bandwidth is obtained.

2. The method according to claim 1, characterized in that, Determining the center frequency distribution range and bandwidth distribution range based on the first center frequency, the first bandwidth, the signal-to-noise ratio, and the frequency interval includes: Based on the signal-to-noise ratio, determine the corresponding distribution range parameters; Based on the distribution range parameters, the first center frequency, and the frequency interval, the center frequency distribution range is obtained, and based on the distribution range parameters, the first bandwidth, and the frequency interval, the bandwidth distribution range is obtained.

3. The method according to claim 2, characterized in that, The step of determining the corresponding distribution range parameters based on the signal-to-noise ratio includes: Based on the signal-to-noise ratio, the corresponding distribution range parameter is calculated using a preset distribution range calculation formula; wherein the distribution range parameter belongs to [1,4] and is an integer; the preset distribution range calculation formula is as follows: Distribution range parameter = floor(preset reference parameter value / signal-to-noise ratio); where floor is the result of dividing the preset reference parameter value by the signal-to-noise ratio and rounding it down.

4. The method according to claim 2, characterized in that, The process of obtaining the center frequency distribution range based on the distribution range parameters, the first center frequency, and the frequency interval, and obtaining the bandwidth distribution range based on the distribution range parameters, the first bandwidth, and the frequency interval, includes: Based on the distribution range parameter and the frequency interval, a range deviation value is obtained; wherein, the range deviation value is the product of the distribution range parameter and the frequency interval; Subtracting the range deviation value from the first center frequency yields the lower limit of the center frequency distribution range, and adding the range deviation value to the first center frequency yields the upper limit of the center frequency distribution range, thus obtaining the center frequency distribution range. Subtracting the range deviation value from the first bandwidth yields the lower limit of the bandwidth distribution range, and adding the range deviation value to the first bandwidth yields the upper limit of the bandwidth distribution range, thus obtaining the bandwidth distribution range.

5. The method according to claim 1, characterized in that, The parameters of the Fourier transform also include: sampling rate; the bandpass filtering process performed on the original signal based on the bandwidth distribution range to obtain the filtered signal includes: The original signal is bandpass filtered using a preset bandpass finite impulse response filter to obtain a filtered signal; wherein the window coefficient of the bandpass finite impulse response filter is obtained based on the bandwidth distribution range and the sampling rate, and the window coefficient is the ratio of the total width of the bandwidth distribution range to half of the sampling rate.

6. The method according to claim 1, characterized in that, The step of determining the frequency point with the largest power difference within the center frequency distribution range based on the target frequency domain signal to obtain the second center frequency includes: Based on the center frequency distribution range, the corresponding first starting frequency point and first ending frequency point are determined, and the power values ​​of all frequency points between the first starting frequency point and the first ending frequency point are extracted from the target frequency domain signal. Calculate the first absolute value difference sequence of the power values ​​of all the extracted frequency points; wherein each element in the first absolute value difference sequence is the absolute value of the difference between the power values ​​of two adjacent frequency points; The frequency point corresponding to the element with the largest value in the first absolute value difference sequence is determined to obtain the second center frequency.

7. The method according to claim 1, characterized in that, The step of determining the corresponding lower bound distribution range and upper bound distribution range of bandwidth based on the second center frequency, the first bandwidth, the signal-to-noise ratio, and the frequency interval includes: Based on the signal-to-noise ratio, determine the corresponding distribution range parameters; Based on the second center frequency, the first bandwidth, the distribution range parameter, and the frequency interval, the corresponding lower bandwidth distribution range and upper bandwidth distribution range are determined.

8. The method according to claim 7, characterized in that, The step of determining the corresponding lower bound distribution range and upper bound distribution range of bandwidth based on the second center frequency, the first bandwidth, the distribution range parameter, and the frequency interval includes: Based on the distribution range parameter and the frequency interval, a range deviation value is obtained; wherein, the range deviation value is the product of the distribution range parameter and the frequency interval; Based on the second center frequency and the first bandwidth, a first reference center value for the lower bound distribution range of the bandwidth and a second reference center value for the upper bound distribution range of the bandwidth are obtained; wherein, the first reference center value is the difference between the second center frequency and half of the first bandwidth, and the second reference center value is the sum between the second center frequency and half of the first bandwidth. Based on the first reference center value and the range deviation value, the bandwidth lower bound distribution range is determined; wherein, the lower limit of the bandwidth lower bound distribution range is the first reference center value minus the range deviation value, and the upper limit of the bandwidth lower bound distribution range is the first reference center value plus the range deviation value. Based on the second reference center value and the range deviation value, the upper limit distribution range of the bandwidth is determined; wherein, the lower limit of the upper limit distribution range of the bandwidth is the second reference center value minus the range deviation value, and the upper limit of the upper limit distribution range of the bandwidth is the second reference center value plus the range deviation value.

9. The method according to claim 1, characterized in that, Based on the target frequency domain signal, the frequency point with the largest power difference value is determined within the lower and upper bound distribution ranges of the bandwidth, respectively, to obtain the bandwidth start frequency and bandwidth end frequency. Based on the bandwidth start frequency and bandwidth end frequency, a second bandwidth is obtained, including: Based on the bandwidth lower bound distribution range, the corresponding second starting frequency point and second ending frequency point are determined, and the power values ​​of all frequency points between the second starting frequency point and the second ending frequency point are extracted in the target frequency domain signal to obtain the power values ​​of all frequency points within the bandwidth lower bound distribution range. Calculate the second absolute value difference sequence of power values ​​for all frequency points within the extracted lower bound of the bandwidth distribution range; wherein each element in the second absolute value difference sequence is the absolute value of the difference between the power values ​​of two adjacent frequency points; Determine the frequency point corresponding to the element with the largest value in the second absolute value difference sequence to obtain the bandwidth starting frequency; Based on the bandwidth upper bound distribution range, the corresponding third starting frequency point and third ending frequency point are determined, and the power values ​​of all frequency points between the third starting frequency point and the third ending frequency point are extracted in the target frequency domain signal to obtain the power values ​​of all frequency points within the bandwidth upper bound distribution range. Calculate the third absolute value difference sequence of power values ​​for all frequency points within the extracted upper bound of the bandwidth distribution range; wherein each element in the third absolute value difference sequence is the absolute value of the difference between the power values ​​of two adjacent frequency points; Determine the frequency point corresponding to the element with the largest value in the third absolute value difference sequence to obtain the bandwidth end frequency; The second bandwidth is obtained by subtracting the bandwidth start frequency from the bandwidth end frequency.

10. A signal frequency parameter estimation device, characterized in that, include: An initial data acquisition module is used to acquire the original signal, and the first center frequency, first bandwidth, and signal-to-noise ratio corresponding to the original signal; wherein, the first center frequency and the first bandwidth are determined based on the time-frequency diagram obtained by Fourier transform of the original signal, and the parameters of the Fourier transform include: frequency interval; The distribution range determination module is used to determine the center frequency distribution range and the bandwidth distribution range based on the first center frequency, the first bandwidth, the signal-to-noise ratio, and the frequency interval; The original signal processing module is used to perform bandpass filtering on the original signal based on the bandwidth distribution range to obtain a filtered signal, and to perform autocorrelation processing on the filtered signal to obtain a target signal. The signal frequency domain conversion module is used to perform Fourier transform and frequency domain correction processing on the target signal to obtain the target frequency domain signal; wherein, the target frequency domain signal includes multiple frequency points and the power value corresponding to each frequency point; The precise center determination module is used to determine the frequency point with the largest power difference value within the center frequency distribution range based on the target frequency domain signal, and obtain the second center frequency; The bandwidth range determination module is used to determine the corresponding lower bandwidth distribution range and upper bandwidth distribution range based on the second center frequency, the first bandwidth, the signal-to-noise ratio, and the frequency interval; The precise bandwidth determination module is used to determine the frequency point with the largest power difference value based on the target frequency domain signal, respectively, within the lower bandwidth distribution range and the upper bandwidth distribution range, to obtain the bandwidth start frequency and the bandwidth end frequency, and to obtain the second bandwidth based on the bandwidth start frequency and the bandwidth end frequency.

Images