A method for detecting frequency-hopping signals for a filter

Abstract

The present invention discloses a method for detecting a frequency-hopping signal for a filter, which relates to the technical field of filters. The method includes: Step S1: Set the expected frequency-hopping range of the target frequency-hopping signal, and divide the expected frequency-hopping range of the target frequency-hopping signal into several sub-bands; Step S2: Use PSK phase modulation to convert the target frequency-hopping signal to the frequency domain and then perform phase modulation to generate a modulation signal; Step S3: Use a phase-locked loop frequency tracking module to perform VCO frequency modulation on the modulation signals of each sub-band; Step S4: Transfer all the sub-bands output after VCO frequency modulation to the time domain through PSK phase demodulation and label them as corrected frequency-hopping signals. Compared with the prior art, the present invention optimizes the modulation by converting the target frequency-hopping signal to the frequency domain through PSK phase modulation, and synchronizes the frequencies of the modulation signals of each sub-band using phase-locked loop technology, having the advantages of avoiding phase drift caused by frequency mismatch and obtaining the beneficial effect of higher-precision signal detection.

Description

Technical Field

[0001] The present invention relates to the technical field of filters, and particularly to a method for detecting hopping signals for filters. Background Art

[0002] In a filter hopping communication system, the filter needs to respond to frequency changes and quickly adjust between multiple frequency bands. During high-speed hopping or complex multi-band hopping of the filter, there is a high probability of random time response delay problems. The time response delay is the time required for the filter to adapt to a new frequency after receiving a signal. The response delay causes the filter to take longer to identify and lock the frequency range of the signal. When the signal frequency jumps, the filter cannot adjust its operating frequency in time, possibly missing the key part of the signal, resulting in signal capture failure or incomplete capture.

[0003] The time response delay of the filter directly affects the capture rate of the system. The capture rate refers to the ability of the system to accurately identify and lock a signal within a sufficiently short time. When the response delay of the filter is large, the time required to capture the signal increases, resulting in an extended adaptation process of the system to the signal, increasing the probability of signal loss or demodulation errors. For example, the filter needs to adjust its frequency band range to adapt to signal jumps or frequency changes, but the delay causes the filter adjustment process to lag behind the signal changes. Over time, this lag effect may cause the captured signal to become inaccurate, resulting in phase errors in the signal, and the filter may not be able to accurately lock the signal within the correct time-frequency range. When the filter cannot accurately lock the signal, the signal quality and demodulation accuracy will both decrease, thereby significantly reducing the signal detection accuracy of the filter during the hopping process. Summary of the Invention

[0004] The present invention provides a method for detecting hopping signals for filters, which solves the problem that existing hopping filters are prone to signal phase errors due to time response delay during high-speed hopping or complex multi-band hopping, resulting in reduced signal detection accuracy.

[0005] The present invention is achieved through the following technical solutions:

[0006] A method for detecting hopping signals for filters, the method comprising:

[0007] Step S1: Select a target filter and a target hopping signal for hopping, collect historical hopping data of the target hopping signal and use it to set an expected hopping range, and divide the expected hopping range of the target hopping signal into several sub-bands;

[0008] Step S2: Start the frequency hopping process of the target filter. After converting the target frequency hopping signal to the frequency domain using PSK phase modulation, perform phase modulation along each sub-band to generate a modulated signal, and maintain smooth transition phase control between adjacent phase modulations;

[0009] Step S3: Construct a phase-locked loop frequency tracking module, and use the phase-locked loop frequency tracking module to perform VCO frequency modulation on the modulated signal of each sub-band to complete the frequency synchronization of the modulated signal during the phase modulation process of each sub-band;

[0010] Step S4: Transfer all sub-bands output after VCO frequency modulation to the time domain through PSK phase demodulation to convert the modulated signal and label it as the corrected frequency hopping signal, and apply the corrected frequency hopping signal to the target filter and re-perform the detection of the frequency hopping signal.

[0011] When the response delay of the filter is large, the time required to capture the signal increases, resulting in an extended adaptation process of the system to the signal, increasing the probability of signal loss or demodulation error. Over time, this lag effect may cause the captured signal to be inaccurate, resulting in a phase error in the signal, and the filter may not be able to accurately lock the signal within the correct time-frequency range. When the filter cannot accurately lock the signal, the quality and demodulation accuracy of the signal will both decrease, thereby significantly reducing the signal detection accuracy of the filter during the frequency hopping process. Based on this, the present invention provides a frequency hopping signal detection method for a filter to solve the problem that existing frequency hopping filters are prone to signal phase errors due to time response delays during high-speed frequency hopping or complex multi-band frequency hopping processes, resulting in a decrease in signal detection accuracy.

[0012] Further, the process of performing modulation on each sub-band using PSK phase modulation includes: The historical frequency hopping data includes the mean signal-to-noise ratio and a frequency change curve representing time-frequency corresponding to the x-y coordinate axes. In the frequency change curve, collect the total amount of absolute frequency changes for each sub-band, and obtain the total frequency hopping duration based on the time axis on the abscissa of the frequency change curve; Calculate the average frequency hopping rate of the target frequency hopping signal for each sub-band according to the total frequency hopping duration and the total amount of absolute frequency changes; Set a rate threshold and an SNR threshold for the average frequency hopping rate and the mean signal-to-noise ratio of all sub-bands respectively; The modulation methods of the PSK phase modulation are represented as BPSK modulation, QPSK modulation, and 8-PSK modulation according to 2-phase, 4-phase, and 8-phase respectively;

[0013] When the average frequency hopping rate of the frequency hopping signal within the sub-band is higher than the rate threshold and the mean signal-to-noise ratio is lower than the SNR threshold, the PSK phase modulation uses BPSK modulation;

[0014] When the average hopping rate of the hopping signal within the sub - band is higher than the rate threshold and the average SNR is higher than the SNR threshold, or when the average hopping rate of the hopping signal within the sub - band is lower than the rate threshold and the average SNR is lower than the SNR threshold, the PSK phase modulation uses QPSK modulation;

[0015] When the average hopping rate of the hopping signal within the sub - band is lower than the rate threshold and the average SNR is higher than the SNR threshold, the PSK phase modulation uses 8 - PSK modulation.

[0016] Further, the process of setting the rate threshold includes: setting a number of rate sequence points along the time axis of the frequency change curve within each sub - band, marking the rate values at each rate sequence point based on the frequency change curve; using the K - means clustering analysis method to divide all rate sequence points into k clusters and performing K - means clustering to generate a centroid within each cluster, where the centroid represents the average of the rate values of all rate sequence points within the corresponding cluster, and setting the average of the two centroids with the largest interval as the rate threshold.

[0017] Further, based on the elbow method, the best k - value correction is performed on the k - value. The process includes: marking the k - value used in the initial execution of K - means clustering as the initial k - value, and marking the generated centroids as the initial centroids; selecting a number of values on both sides of the number axis around the initial k - value as k - value elements, calculating the sum of squared errors of the rate values of the rate sequence points corresponding to each k - value element and the initial centroids, marking all the calculated sum of squared error values on the plane coordinate system with the k - value on the abscissa and connecting the discrete sum of squared error value numerical points into a curve, setting the k - value corresponding to the sum of squared error value closest to the inflection point of the curve as the corrected best k - value, and using the best k - value in the K - means clustering analysis method to regenerate the centroids for calculating the rate threshold.

[0018] Further, the calculation process of the sum of squared errors is set as: representing the sum of squared errors as SSE, representing the rate sequence points as x, and representing the total number and ordinal number of the rate sequence points as n and i respectively, and setting the centroid as μ,

[0019] Then the calculation formula of the sum of squared errors is expressed as: ,

[0020] where the k - value is selected from the k - value elements, x i represents the i - th rate sequence point, and μ k represents the centroid under the current k - value.

[0021] Further, the phase - locked loop frequency tracking module includes:

[0022] A phase comparator for calculating the phase difference between the input signal and the local oscillation signal;

[0023] A low-pass filter, which is used to filter high-frequency noise output by a phase comparator to generate a smooth control signal;

[0024] A local oscillator, which receives a control voltage output by the phase comparator to adjust the output frequency of an input signal;

[0025] Wherein, the phase comparator, the low-pass filter and the local oscillator are sequentially connected by signals.

[0026] Further, the phase comparator adopts an edge detection type phase comparator, which is used to detect the zero-crossing position of an input signal and a VCO output signal, and obtain a phase error value after calculating the time difference between the two.

[0027] Further, the process of obtaining the zero-crossing position is set as follows: interpolation fitting is performed between three consecutive sampling points using the least squares method to obtain a quadratic polynomial, and the zero-crossing position is obtained based on the quadratic polynomial.

[0028] Further, the local oscillator is provided with a VCO reference frequency for dynamically adjusting the output frequency of a modulation signal;

[0029] When the output frequency of the modulation signal is lower than the VCO reference frequency, the phase comparator generates a positive control signal;

[0030] When the output frequency of the modulation signal is higher than the VCO reference frequency, the phase comparator generates a negative control signal;

[0031] When the output frequency of the modulation signal is equal to the VCO reference frequency, the phase comparator generates a zero control signal.

[0032] Further, offset redundancies are provided for both the upper boundary value and the lower boundary value of each sub-band.

[0033] Compared with the prior art, the present invention transfers a target frequency hopping signal to the frequency domain for modulation optimization through PSK phase shift keying modulation, and performs frequency synchronization on the modulation signals of each sub-band through a phase-locked loop technology, so that the phase-locked loop can dynamically adjust and correct the frequency of a VCO voltage controlled oscillator, ensuring that the frequencies of each sub-band are consistent, and having the advantages of being able to avoid phase drift caused by frequency mismatch and the beneficial effect of obtaining higher-precision signal detection in a target filter. Description of the Drawings

[0034] The drawings described herein are used to provide a further understanding of the embodiments of the present invention, form a part of this application, and do not limit the embodiments of the present invention. In the drawings:

[0035] Figure 1 It is a flowchart of the present invention. Detailed Implementation Modes

[0036] To make the objectives, technical solutions, and advantages of the present invention more clearly understood, the present invention will be further described in detail below in conjunction with embodiments and the accompanying drawings. The illustrative implementation modes of the present invention and their descriptions are only used to explain the present invention and are not intended to limit the present invention. Embodiment

[0037] As Figure 1 shown, this embodiment is a method for detecting a frequency-hopping signal for a filter, and the method includes:

[0038] Step S1: Select a target filter and a target frequency-hopping signal for frequency hopping, collect historical frequency-hopping data of the target frequency-hopping signal and use it to set an expected frequency-hopping range, and divide the expected frequency-hopping range of the target frequency-hopping signal into several sub-bands;

[0039] Step S2: Start the frequency-hopping process of the target filter, use PSK phase modulation to transfer the target frequency-hopping signal to the frequency domain and then perform phase modulation along each sub-band to generate a modulation signal, and maintain smooth transition phase control between adjacent phase modulations;

[0040] Step S3: Construct a phase-locked loop frequency tracking module, and use the phase-locked loop frequency tracking module to perform VCO frequency modulation on the modulation signal of each sub-band to complete the frequency synchronization of the modulation signal during the phase modulation process of each sub-band;

[0041] Step S4: Transfer all sub-bands output after VCO frequency modulation to the time domain through PSK phase demodulation and label them as corrected frequency-hopping signals, and apply the corrected frequency-hopping signals to the target filter and re-perform the detection of the frequency-hopping signal.

[0042] The target filter refers to a filter device that is currently greatly affected by time response delay, resulting in a decrease in the system capture rate, a large number of phase errors in the signal, and thus a significant reduction in the signal detection accuracy during the frequency-hopping process. The target frequency-hopping signal refers to the signal to be frequency-hopped in the system. The frequency-hopping signal is usually a signal containing multiple frequency points, which changes its frequency by "hopping" to avoid interference, improve confidentiality, or improve signal reliability. In practical applications, the frequency-hopping signal may be a carrier signal in wireless communication, or a signal used in fields such as radar, navigation, and radio spectrum detection. "Signal frequency hopping" means that the signal hops within a predetermined frequency range according to a specific rule. For example, in a frequency-hopping communication system, the carrier signal will quickly switch between multiple frequency points. This frequency-hopping operation is usually to prevent spectrum congestion, improve anti-interference ability, or increase signal security. Here, the target frequency-hopping signal for frequency hopping refers to the signal that undergoes frequency changes under the processing of the filter.

[0043] The meaning of the expected frequency hopping range refers to the frequency range of the target frequency hopping signal during the frequency hopping process, which is set in advance according to the application requirements and historical data of the target system. The expected frequency hopping range includes the lowest frequency and the highest frequency of frequency hopping, and is divided into multiple sub-bands. The frequency range of each sub-band can be regarded as an independent working interval during the frequency hopping process. The goal is to ensure smooth frequency changes and minimum phase error within each sub-band. The PSK phase modulation is a digital modulation technique that represents different digital information by changing the phase of the signal carrier; PSK modulation maps digital data to different phases of the carrier by discretely changing the phase of the carrier, thereby realizing data transmission. Using a phase-locked loop for VCO frequency modulation means controlling and adjusting the VCO frequency through the phase-locked loop PLL to ensure frequency synchronization of the frequency hopping signal between different frequency bands; the role of the phase-locked loop is to provide stable and accurate frequency tracking during the frequency hopping process and achieve frequency synchronization to reduce the phase error caused by frequency inconsistency.

[0044] Start the process of the frequency hopping signal and begin frequency hopping. Use PSK phase modulation to transfer the frequency hopping signal to the frequency domain and perform phase modulation on each sub-band. The purpose is to map the phase information of the original signal to the frequency domain. The frequency hopping signal within each sub-band will be modulated into a signal with a specific phase. Among them, it is necessary to perform smooth control processing on the phase transition between adjacent frequency bands to avoid sudden changes and phase jumps caused by frequency band switching. In the commonly used technologies in this field, exponential smoothing transition or sine transition function can be used for optimization processing. The phase-locked loop frequency tracking module is used to synchronize the frequency of the modulation signal in each sub-band and modulate the VCO frequency to maintain frequency consistency. The phase-locked loop ensures that the frequency of each sub-band is consistent with the reference signal of the phase-locked loop by adjusting the frequency of the VCO in real time. The working principle of the PLL is to compare the phase of the input signal with the phase of the local reference signal and adjust the output frequency of the VCO to keep it synchronized with the frequency of the input signal. Then use PSK phase demodulation to perform PSK phase demodulation on the modulation signal, transfer the signal from the frequency domain back to the time domain, and obtain the corrected frequency hopping signal. The previous PSK modulation and PLL synchronization processes have corrected the phase error, and the demodulated frequency hopping signal will have less phase error than the original signal, so as to achieve the beneficial effect of obtaining a higher-precision signal detection in the target filter.

[0045] As a feasible implementation, an offset redundancy is provided for both the upper and lower boundary values of each sub - band. The offset redundancy refers to an additional frequency range added to the upper and lower boundary values of each sub - band. This redundancy is a dynamic and adjustable value that can help compensate for possible frequency errors, drifts, or response delays in the system. The system may be affected by factors such as noise, frequency errors, and device precision. The offset redundancy provides a buffer to tolerate these uncertainties. During high - speed frequency hopping or multi - band hopping, the frequency of the signal may fluctuate or slightly exceed the predetermined boundaries due to various reasons. Adding the offset redundancy ensures that within these fluctuation ranges, the frequency band can still be effectively detected and processed.

[0046] Furthermore, as a feasible implementation, the process of modulating each sub - band using PSK phase modulation includes: The historical frequency - hopping data includes the mean signal - to - noise ratio and a frequency change curve representing time - frequency corresponding to the x - y coordinate axes. For each sub - band in the frequency change curve, the total absolute value of the frequency change is collected, and the total frequency - hopping duration is obtained based on the time axis on the abscissa of the frequency change curve; The average frequency - hopping rate of the target frequency - hopping signal is calculated for each sub - band according to the total frequency - hopping duration and the total absolute value of the frequency change; Rate thresholds and SNR thresholds are respectively set for the average frequency - hopping rates and the mean signal - to - noise ratios of all sub - bands; The modulation methods of the PSK phase modulation are represented as BPSK modulation, QPSK modulation, and 8 - PSK modulation according to 2 - phase, 4 - phase, and 8 - phase respectively;

[0047] When the average frequency - hopping rate of the frequency - hopping signal within the sub - band is higher than the rate threshold and the mean signal - to - noise ratio is lower than the SNR threshold, the PSK phase modulation uses BPSK modulation;

[0048] When the average frequency - hopping rate of the frequency - hopping signal within the sub - band is higher than the rate threshold and the mean signal - to - noise ratio is higher than the SNR threshold, or when the average frequency - hopping rate of the frequency - hopping signal within the sub - band is lower than the rate threshold and the mean signal - to - noise ratio is lower than the SNR threshold, the PSK phase modulation uses QPSK modulation;

[0049] When the average frequency - hopping rate of the frequency - hopping signal within the sub - band is lower than the rate threshold and the mean signal - to - noise ratio is higher than the SNR threshold, the PSK phase modulation uses 8 - PSK modulation.

[0050] The average signal-to-noise ratio refers to the average value of the ratio of signal strength to noise strength within a historical operating time period. In a frequency-hopping system, the signal-to-noise ratio is a key indicator for measuring signal quality, which reflects the strength of the received signal relative to the background noise. The average signal-to-noise ratio can be obtained by averaging the SNR values over a certain time period. For example, for the signals in each frequency band or time window, the signal-to-noise ratio at each moment can be calculated, and then these values are averaged to obtain the average signal-to-noise ratio for that time period or frequency band. The frequency change curve is a trajectory in the time-frequency two-dimensional coordinate system that represents how the frequency of a frequency-hopping signal changes over time. Its abscissa is time and its ordinate is frequency; the curve shows the change of the signal frequency over time during the frequency-hopping process. The total amount of the absolute value of the frequency change refers to the total change amount of the amplitude of the signal frequency change within their respective corresponding sub-bands; the absolute value of the frequency change is a scalar, which only represents taking the absolute value of the "amplitude difference between two adjacent frequency values", regardless of positive or negative cancellation, and the meaning of its total amount is the cumulative change amount of the signal frequency. For example, in one frequency hop, from +50 KHz to -50 KHz and then to +20 KHz, the absolute value of the frequency change between +50 KHz and -50 KHz is |50 - (-50)| = 100 KHz, and the absolute value of the frequency change between -50 KHz and +20 KHz is |-50 - 20| = 70 KHz, so the total amount of the absolute value of the frequency change in this process is 100 + 70 = 170 KHz. For the frequency change curve in the time-frequency two-dimensional coordinate system, the total frequency-hopping duration of this frequency hop of the target filter is obtained through the abscissa, and then by combining the total amount of the absolute value of the frequency change with the total frequency-hopping duration, the average frequency-hopping rate of the frequency-hopping signal is calculated = total amount of the absolute value of the frequency change / total frequency-hopping duration. The average frequency-hopping rate reflects the speed of frequency change within a certain sub-band.

[0051] When the high frequency-hopping rate, that is, when the frequency of the frequency-hopping signal changes relatively fast, it may lead to a relatively large phase change, thereby increasing the demodulation difficulty of the system. When the average frequency-hopping rate of the frequency-hopping signal within the sub-band is relatively high (that is, the frequency of the signal changes very rapidly) and the average signal-to-noise ratio is relatively low (that is, the signal quality is poor and the noise is strong), the system selects BPSK modulation to ensure that the signal can still be stably transmitted under the conditions of high-speed frequency hopping and low signal-to-noise ratio, and to avoid excessive phase errors and bit errors; at high frequency-hopping rates, rapid frequency hopping may lead to relatively large phase errors. As the most basic and noise-resistant modulation method, BPSK modulation can provide better fault tolerance under such conditions; under low signal-to-noise ratio conditions, higher-order modulation modes are more vulnerable to the influence of noise, resulting in a higher bit error rate. Therefore, BPSK has only two possible phases, 0° and 180°, which makes the influence of phase errors on the demodulation process relatively small. Its lower order enables it to better adapt to the environment with low signal-to-noise ratio, reduce the risk of bit errors, and ensure the reliable reception of the signal.

[0052] When the average hopping rate is higher than the rate threshold and the average SNR is higher than the SNR threshold, it indicates that the hopping signal switches rapidly within the current sub-band, the hopping rate is high, the instability of the signal is large, and the rapid frequency change is likely to introduce phase errors and interference. However, the signal quality is good, and the signal is relatively strong and stable. A low average rate means better signal stability because the frequency changes slowly and the signal does not change rapidly, so it is relatively easy to process. When the average hopping rate is lower than the rate threshold and the average SNR is lower than the SNR threshold, it indicates that the frequency change of the hopping signal within the sub-band is relatively slow, and the frequency switching speed is low. Because the frequency changes slowly, the signal does not change rapidly and has good stability. At the same time, due to the low average SNR, it indicates that the signal quality is poor, the background noise is strong, and the signal may be greatly interfered at the receiving end. Therefore, the signal is vulnerable to noise during transmission, which may lead to a high bit error rate. Based on this, when the hopping rate is low and the SNR is high, or the hopping rate is high but the SNR is low, QPSK modulation is selected in this embodiment. The QPSK modulation can provide a higher data transmission rate compared to BPSK modulation and is more stable than modulation modes with higher phases. In the case of a relatively high hopping rate and good SNR, QPSK modulation can provide a balanced transmission rate and anti-interference ability. On the other hand, even when the hopping rate is low and the SNR is poor, QPSK modulation can still provide higher transmission efficiency and anti-interference ability. Compared with BPSK or higher-phase modulation, QPSK still has good performance under low SNR conditions. The QPSK modulation scheme under this condition aims to provide a compromise solution, that is, when the hopping is fast and the signal quality is good, make full use of the high transmission efficiency of QPSK; when the hopping is slow and the signal quality is poor, choosing QPSK can still provide good anti-interference performance while ensuring an increase in transmission efficiency.

[0053] When the average hopping rate of the hopping signal within the sub-band is lower than the rate threshold and the average SNR is higher than the SNR threshold, the phase change of the signal is relatively stable at this time and is not easily affected by frequent frequency hopping. At the same time, a high SNR indicates good signal quality and small noise interference. The system can effectively identify more phase changes. Therefore, it is suitable to use a higher-order modulation method. The 8-PSK modulation uses 8 different phases to represent data. Compared with the 2 phases of BPSK and the 4 phases of QPSK, 8-PSK can transmit more data under the same bandwidth condition, and can improve the efficiency and capacity of data transmission while ensuring signal stability.

[0054] Further, as a feasible implementation manner, the process of setting the rate threshold includes: setting a number of rate sequence points along the time axis of the frequency variation curve in each sub-band, marking the rate values at each rate sequence point based on the frequency variation curve; using the K-means clustering analysis method to divide all the rate sequence points into k clusters and performing K-means clustering to generate a centroid in each cluster, where the centroid represents the average value of the rate values of all the rate sequence points in the corresponding cluster, and setting the average value of the two centroids with the largest interval as the rate threshold.

[0055] On the time axis of the frequency variation curve, a number of rate sequence points are selected. These rate sequence points represent the hopping rates at different time moments, that is, each rate sequence point corresponds to the speed of signal frequency change at a specific time point. K-means clustering analysis is a common unsupervised learning algorithm used to divide a data set into several clusters. In this implementation manner, the goal of K-means clustering analysis is to divide all the rate sequence points according to the similarity of their rate values and generate multiple clusters. The centroid represents the average rate of the rate values in each cluster. The two centroids with the largest interval are selected because these two centroids represent the two extreme values of the hopping rate, namely high-speed hopping and low-speed hopping. These two extreme values can effectively define different intervals of the hopping rate. The centroids with the largest interval can fully reflect the variation range of the hopping signal in terms of rate. When setting the rate threshold, using the average value of these two centroids helps to find a suitable demarcation point between high-speed and low-speed hopping.

[0056] Further, as a feasible implementation manner, the optimal k value is corrected based on the elbow method. The process includes: marking the k value used for the first execution of K-means clustering as the initial k value and marking the generated centroids as the initial centroids; selecting a number of values as k value elements on both sides of the number axis around the initial k value, calculating the sum of squared errors of the rate values of the rate sequence points corresponding to each k value element and the initial centroids, marking all the calculated sums of squared errors on a plane coordinate system with the k value as the abscissa and connecting the discrete sum of squared error value points into a curve, setting the k value corresponding to the sum of squared errors closest to the inflection point position of the curve as the corrected optimal k value, and using the optimal k value for the K-means clustering analysis method to regenerate the centroids for calculating the rate threshold.

[0057] Perform K-means clustering using an initial value of k to obtain the clustering results and the corresponding centroids. The initial value of k can be estimated based on domain knowledge or previous experience to obtain a suitable initial value of k. The k-value elements refer to selecting multiple different k-values (i.e., the number of clusters) within a certain range as candidate cluster numbers based on the initial value of k. The purpose is to explore the changes in the clustering effect and the sum of squared error values under different numbers of clusters. For a series of different k-values in the k-value elements, calculate the sum of squared error values corresponding to each k-value. These sum of squared error values will form a set of discrete numerical points; taking these sum of squared error values and the corresponding k-values as coordinates, plot a curve in the plane coordinate system. The abscissa represents the number of clusters k, and the ordinate represents the sum of squared error values. Each k-value corresponds to a sum of squared error value, and the discrete numerical points will be connected into a curve. As the k-value increases, the sum of squared error values will gradually decrease, showing a decreasing trend; but after the k-value increases to a certain extent, the rate of decrease of the sum of squared error values will significantly slow down, and the curve will form an inflection point, that is, the position of the elbow in the curve. By observing the curve, the position of the elbow can be found, that is, the place where the rate of decrease of the sum of squared error values significantly slows down. The elbow usually represents the optimal point of clustering because after this, increasing the number of clusters will not significantly improve the clustering effect. According to the elbow method, select the k-value at the inflection point of the curve as the optimal number of clusters. Specifically, the sum of squared error value corresponding to the k-value closest to the inflection point can be selected by measuring the distances between each point on the curve and the position of the elbow.

[0058] Further, as a feasible implementation manner, the calculation process of the sum of squared error values is set as follows: Represent the sum of squared error values as SSE, represent the rate sequence points as x, and represent the total number and ordinal number of the rate sequence points as n and i respectively, and set the centroid as μ.

[0059] Then the calculation formula of the sum of squared error values is expressed as: ,

[0060] where the k-value is selected from the k-value elements, x i represents the i-th rate sequence point, and μ k represents the centroid under the current k-value.

[0061] The sum of squared error value SSE is an important indicator in clustering analysis, which is used to measure the average deviation of the data points within the cluster from their cluster centroids; the smaller the value, the closer the data points are to their cluster centroids, and the better the clustering effect. In this embodiment, the sum of squared error value SSE represents the sum of the squared differences between all the rate sequence points within the k-th cluster and its cluster centroid μ k therebetween. The represents the squared difference between each rate sequence point and its cluster centroid. Adding up all these squared differences gives the sum of squared errors of the cluster. The centroid μ kDenote the centroid of the k-th cluster, which is the average of all rate sequence points within the cluster. The centroid is determined by calculating the average position of all data points within the cluster. In a specific implementation, the centroid μ k can be expressed by the calculation formula: , where n k represents the number of rate sequence points in the k-th cluster.

[0062] Furthermore, the phase-locked loop frequency tracking module includes:

[0063] A phase comparator for calculating the phase difference between the input signal and the local oscillation signal;

[0064] A low-pass filter for filtering out high-frequency noise output by the phase comparator to generate a smooth control signal;

[0065] A local oscillator that receives the control voltage output from the phase comparator to adjust the output frequency of the input signal;

[0066] Among them, the phase comparator, the low-pass filter, and the local oscillator are sequentially connected by signals.

[0067] The phase comparator is used to calculate the phase difference between the input signal and the local oscillation signal. It compares the input modulation signal with the local reference oscillation signal and outputs a voltage signal proportional to their phase difference. If the two signals are in phase, the output of the phase comparator is zero. If there is a phase difference, the output voltage will increase as the phase difference increases. The output of the phase comparator is a varying signal indicating the phase change of the input signal relative to the reference signal. The main task of the low-pass filter is to remove high-frequency noise from the output of the phase comparator and generate a smooth control signal; the output signal of the phase comparator usually contains some high-frequency components or noise, and the low-pass filter can filter out these high-frequency components and only retain the low-frequency part, thereby ensuring that the output control signal is stable. By smoothing the output, the low-pass filter helps to stabilize the frequency control, making the control signal more suitable for adjusting the local oscillator, and this smooth control signal can directly affect the frequency adjustment of the oscillator. The local oscillator receives the smooth control voltage output by the low-pass filter and adjusts its output frequency according to this control voltage. The local oscillator is the core part of the frequency adjustment in the system, and it tracks the frequency change of the input signal by changing its output frequency; when the control signal changes, the frequency of the local oscillator also changes accordingly, thereby ensuring the frequency synchronization between the input signal and the local oscillation signal. The frequency of the local oscillator is adjusted by the control voltage provided by the low-pass filter, and the change of the control voltage directly affects the operating frequency of the oscillator; if the frequency of the input signal changes, the low-pass filter will adjust the control signal, thereby changing the output frequency of the local oscillator, and finally achieving the frequency synchronization between the input signal and the local oscillation signal.

[0068] Further, as a feasible implementation, the phase comparator adopts an edge detection type phase comparator, which is used to detect the zero crossing positions of the input signal and the VCO output signal, and obtain the phase error value after calculating the time difference between the two; the process of obtaining the zero crossing positions is set as follows: interpolation fitting is performed between three consecutive sampling points using the least squares method to obtain a quadratic polynomial, and the zero crossing positions are calculated based on the quadratic polynomial.

[0069] The edge detection type phase comparator is used to detect the zero crossings of the signal, that is, the moments when the signal changes from positive to negative or from negative to positive. The detection of zero crossings is the basis for determining the phase difference of the signal because the zero crossing positions are directly related to the periodic changes and phase characteristics of the signal. The input signal and the VCO output signal detect their zero crossings through the phase comparator, and calculate the time difference of the zero crossings; the time difference of the zero crossings is the phase difference between the input signal and the VCO output signal, and this phase difference can be calculated through the relationship between the time difference and the signal period. In order to improve the accuracy of the zero crossing positions, the edge detection type phase comparator uses the least squares method to perform interpolation fitting between three consecutive sampling points. The least squares method is an optimization algorithm aimed at minimizing the sum of squared errors and finding the most suitable quadratic polynomial to approximate the change trend of these three points. The three sampling points refer to three consecutive data points of the input signal or the VCO output signal within the sampling period. Based on these three sampling points, a quadratic polynomial is fitted using the least squares method, and this polynomial can accurately represent the change of the signal between these three points. In specific applications, the form of the quadratic polynomial can be set as f(t)=λt 2 +βt+φ, where λ, β, and φ are fitting parameters obtained through the least squares method, which describe the change trend of the signal between these three points. Using the obtained quadratic polynomial, the zero crossing positions can be solved. The zero crossing positions refer to the solutions of the polynomial, and its form can be set as: λt 2 +βt+φ = 0. By solving the above equation, the time t0 of the zero crossing is obtained, that is, the zero crossing time of the input signal and the VCO output signal. By using the least squares method to fit a quadratic polynomial between three sampling points, a more accurate zero crossing position can be obtained than a single sampling point. This method can reduce the errors caused by uneven sampling points or noise and improve the accuracy of phase comparison. Performing interpolation fitting between consecutive sampling points can effectively reduce the influence of sampling noise on the zero crossing positions and avoid large phase errors caused by errors in individual sampling points.

[0070] Further, as a feasible implementation, the local oscillator is provided with a VCO reference frequency for dynamically adjusting the output frequency of the modulation signal;

[0071] When the output frequency of the modulation signal is lower than the VCO reference frequency, the phase comparator generates a positive control signal;

[0072] When the output frequency of the modulation signal is higher than the VCO reference frequency, the phase comparator generates a negative control signal;

[0073] When the output frequency of the modulation signal is equal to the VCO reference frequency, the phase comparator generates a zero control signal.

[0074] The VCO reference frequency is the reference frequency of the local oscillator and is used to set the target frequency. When the frequency of the modulation signal needs to be synchronized with this reference frequency, the local oscillator adjusts its output frequency according to the feedback of the phase comparator. The output frequency of the modulation signal is the frequency output by the local oscillator and is usually frequency synchronized or adjusted with the input signal. The frequency change of the modulation signal is controlled by the phase-locked loop and adjusted according to the output of the phase comparator to achieve frequency synchronization. When the output frequency of the modulation signal is lower than the VCO reference frequency, it indicates that the frequency of the modulation signal is low and the output frequency needs to be increased to match the reference frequency. The phase comparator will generate a positive control signal, which will be fed back to the local oscillator to prompt it to increase the output frequency. When the output frequency of the modulation signal is higher than the VCO reference frequency, the output frequency needs to be decreased to achieve synchronization. The phase comparator generates a negative control signal, which will be fed back to the local oscillator to prompt it to decrease the output frequency. When the output frequency of the modulation signal is equal to the VCO reference frequency, the frequency of the modulation signal is exactly the same as the reference frequency and no adjustment is required. The phase comparator generates a zero control signal, meaning that the output frequency of the VCO has exactly matched the target frequency.

[0075] The specific embodiments described above further elaborate on the purpose, technical solutions, and beneficial effects of the present invention. It should be understood that the above description is only the specific embodiments of the present invention and is not used to limit the protection scope of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection scope of the present invention.

Claims

1. A frequency hopping signal detection method for a filter, characterized in that: The method includes: Step S1: Select a target filter and a target frequency hopping signal for frequency hopping, collect historical frequency hopping data of the target frequency hopping signal and use it to set an expected frequency hopping range, and divide the expected frequency hopping range of the target frequency hopping signal into a plurality of sub-frequency bands; Step S2: starting the frequency hopping process of the target filter, using PSK phase modulation to convert the target frequency hopping signal to the frequency domain and then performing phase modulation along each sub-band to generate a modulated signal, and maintaining smooth transition phase control between adjacent phase modulations; Step S3: construct a phase-locked loop frequency tracking module, and use the phase-locked loop frequency tracking module to perform VCO frequency modulation on the modulation signal of each sub-band to complete the frequency synchronization of the modulation signal in the phase modulation process of each sub-band; Step S4: convert all sub-bands output after VCO frequency modulation into time domain through PSK phase demodulation and mark them as modified frequency hopping signals, apply the modified frequency hopping signals to the target filter and re-implement the detection of the frequency hopping signals.

2. A frequency hopping signal detection method for a filter according to claim 1, characterized in that: The process of modulating each sub-frequency band using PSK phase modulation includes: the historical frequency hopping data includes a signal-to-noise ratio mean value and a frequency change curve corresponding to the xy coordinate axis representing time-frequency, the total amount of the absolute value of the frequency change is collected for each sub-frequency band in the frequency change curve, and the total frequency hopping duration is obtained based on the time axis on the abscissa of the frequency change curve; the average frequency hopping rate of the target frequency hopping signal is calculated in each sub-frequency band according to the total frequency hopping duration and the total amount of the absolute value of the rate change; the rate threshold and the SNR threshold are set for the average frequency hopping rate and the signal-to-noise ratio mean value of all sub-frequency bands; the modulation modes of the PSK phase modulation are respectively represented as BPSK modulation, QPSK modulation and 8-PSK modulation according to 2 phases, 4 phases and 8 phases; When the frequency hopping average rate of the frequency hopping signal in the sub-frequency band is higher than the rate threshold and the signal-to-noise ratio average is lower than the SNR threshold, the PSK phase modulation uses BPSK modulation; When the average frequency hopping rate of the frequency hopping signal in the sub-frequency band is higher than the rate threshold and the average signal-to-noise ratio is higher than the SNR threshold, or when the average frequency hopping rate of the frequency hopping signal in the sub-frequency band is lower than the rate threshold and the average signal-to-noise ratio is lower than the SNR threshold, the PSK phase modulation uses QPSK modulation; When the average frequency hopping rate of the frequency hopping signal in the sub-frequency band is lower than the rate threshold and the average signal-to-noise ratio is higher than the SNR threshold, the PSK phase modulation uses 8-PSK modulation.

3. A frequency hopping signal detection method for a filter according to claim 2, characterized in that: The process of setting the rate threshold includes: setting a number of rate sequence points along the time axis of the frequency change curve in each sub-frequency band, marking the rate value at each rate sequence point based on the frequency change curve; using the K-means clustering analysis method to divide all rate sequence points into k clusters and perform K-means clustering to generate a centroid in each cluster, the centroid represents the average value of the rate values ​​of all rate sequence points in the corresponding cluster, and setting the mean of the two centroids with the largest interval as the rate threshold.

4. A frequency hopping signal detection method for a filter according to claim 3, characterized in that: The k value is corrected to the best k value based on the elbow rule, and the process includes: marking the k value used in the initial execution of K-means clustering as the initial k value, and marking the generated centroid as the initial centroid; selecting a number of values ​​on both sides of the number axis around the initial k value as k value elements, calculating the sum of squared errors between the rate values ​​of the rate sequence points corresponding to each k value element and the initial centroid, marking all the calculated sum of squared errors on a plane coordinate system with the abscissa being the k value, and connecting the discrete sum of squared error value points into a curve, setting the k value corresponding to the sum of squared errors closest to the inflection point of the curve as the best k value for correction, and using the best k value in the K-means clustering analysis method to regenerate the centroid for calculating the rate threshold.

5. A frequency hopping signal detection method for a filter according to claim 4, characterized in that: The calculation process of the square sum of errors is as follows: the square sum of errors is represented as SSE, the rate sequence point is represented as x, the total number and ordinal number of the rate sequence points are represented as n and i respectively, and the centroid is represented as μ. The calculation formula of the error sum of squares is expressed as: , The k value is selected from the k value element, x i represents the i-th rate sequence point, μ k Represents the center of mass at the current k value.

6. A frequency hopping signal detection method for a filter according to claim 1, characterized in that: The phase-locked loop frequency tracking module comprises: A phase comparator, used to calculate the phase difference between the input signal and the local oscillation signal; A low-pass filter is used to filter the high-frequency noise output by the phase comparator to generate a smooth control signal; A local oscillator receives a control voltage from the phase comparator output to adjust an output frequency of an input signal; Wherein, the phase comparator, the low-pass filter and the local oscillator are signal-connected in sequence.

7. A frequency hopping signal detection method for a filter according to claim 6, characterized in that: The phase comparator is an edge detection type phase comparator, which is used to detect the zero crossing position of the input signal and the VCO output signal, and calculate the time difference between the two to obtain the phase error value.

8. A frequency hopping signal detection method for a filter according to claim 7, characterized in that: The process of acquiring the zero-crossing position is as follows: using the least square method to perform interpolation fitting between three consecutive sampling points to obtain a quadratic polynomial, and calculating and acquiring the zero-crossing position based on the quadratic polynomial.

9. A frequency hopping signal detection method for a filter according to claim 6, characterized in that: The local oscillator is provided with a VCO reference frequency for dynamically adjusting the output frequency of the modulation signal; When the output frequency of the modulation signal is lower than the VCO reference frequency, the phase comparator generates a positive control signal; When the output frequency of the modulation signal is higher than the VCO reference frequency, the phase comparator generates a negative control signal; When the output frequency of the modulation signal is equal to the VCO reference frequency, the phase comparator generates a zero-value control signal.

10. A frequency hopping signal detection method for a filter according to any one of claims 1 to 9, characterized in that: An upper boundary value and a lower boundary value of each sub-frequency band are both provided with an offset redundancy.

Images