A method for detecting time-frequency fingerprint characteristics of a frequency source

Abstract

This invention relates to a method for detecting the time-frequency fingerprint characteristics of a frequency source, belonging to the field of frequency detection technology. The method acquires the clock signal of a frequency standard source, extracts a time-frequency fingerprint feature region from the curve of the clock signal's frequency stability changing with sampling time, and obtains the time-frequency fingerprint characteristics of the frequency source by detecting the time-frequency characteristics of this region. This fingerprint is then compared with the original, inherent time-frequency fingerprint characteristics to identify the type of the frequency standard source and the individual frequency source. This invention creatively proposes using the detection of a specific curve characteristic within the frequency source's stability changing with sampling time as a unique time-frequency fingerprint feature region to identify the type and individual characteristics of the frequency source. This achieves the application objective of uniquely identifying and determining the health status of a specific electronic device by detecting the time-frequency fingerprint characteristics of the frequency source. The method is unique and has good market application prospects.

Description

Technical Field

[0001] This invention belongs to the field of frequency detection technology, specifically relating to a method for detecting the time-frequency fingerprint characteristics of a frequency source. Background Technology

[0002] Currently, existing technologies lack methods for detecting the time-frequency fingerprint characteristics of frequency sources. Traditional methods and indicators for detecting frequency source characteristics only consider the accuracy and stability of the frequency source, neglecting the detection of time-frequency fingerprint characteristics. While accuracy and stability indicators ensure the main characteristics of the frequency source in application, they cannot determine the specific type and individual differences of the frequency source. The time-frequency fingerprint characteristics of a frequency source are influenced by numerous complex factors such as the materials, processes, circuitry, and debugging techniques used in its manufacture, resulting in characteristics that vary from person to person for each individual frequency source. This provides an important application basis for detecting the time-frequency fingerprint characteristics of frequency sources. This patent application is an application of a detection method based on this foundation. Summary of the Invention

[0003] The purpose of this invention is to provide a method for detecting the time-frequency fingerprint characteristics of a frequency source, which solves the problem of detecting the time-frequency fingerprint features of a frequency source that has not been addressed in the prior art. Furthermore, this method provides another way to solve the identification of electronic devices or the monitoring of their operational health status.

[0004] Based on the above objectives, the technical solution of a method for detecting the time-frequency fingerprint characteristics of a frequency source is as follows:

[0005] Step 1: Obtain the clock signal of the frequency source; or, obtain the signal emitted by the electronic device and extract the clock signal of the frequency source in the electronic device from it.

[0006] Step 2: Determine the frequency stability trend curve of the clock signal as a function of sampling time, including the monotonically decreasing region and the non-monotonic decreasing region. Take the last segment of the monotonically decreasing region and the first segment of the non-monotonic decreasing region in the curve. The curve area from the last segment of the monotonically decreasing region to the first segment of the non-monotonic decreasing region is taken as the time-frequency fingerprint feature area.

[0007] Step 3: Using the acquired time-frequency fingerprint feature area, analyze and extract the individual characteristics of the time-frequency fingerprint features of the frequency source, which can be used to identify the type of frequency source or monitor the working health status of the frequency source, or to identify electronic devices containing frequency sources or to monitor the working health status of electronic devices.

[0008] The beneficial effects of the above technical solution are:

[0009] This invention creatively proposes a method for detecting the time-frequency fingerprint characteristics of frequency sources. In the field of frequency source technology, in addition to focusing on accuracy and stability, it introduces the technical characteristics of time-frequency fingerprint features. The detection method uses a specific, inherent region in the curve of the frequency stability of the frequency source changing with sampling time—specifically, the transition region from a monotonically decreasing region to a non-monotonic region—as the time-frequency fingerprint feature region. By detecting the characteristics of this region, more information can be provided for describing, detecting, understanding, and researching the technical performance indicators of the frequency source. For example, it can be used to identify frequency sources or electronic devices containing frequency sources, or to monitor the operating status of frequency sources or electronic devices containing frequency sources.

[0010] Furthermore, in order to determine the time-frequency fingerprint feature region, the steps for dividing the monotonically decreasing change region and the non-monotonic change region are as follows: select the turning point from monotonically decreasing change to non-monotonic change on the frequency stability change curve with sampling time as the dividing point. The curve before the dividing point is the monotonically decreasing change region, and the curve after the dividing point is the non-monotonic change region.

[0011] In the monotonically decreasing region, the curve segment before the dividing point is taken as the final segment; in the non-monotonically decreasing region, the curve segment after the dividing point is taken as the first segment. The length of the final and first segments depends on the analytical characteristics of the frequency source.

[0012] Furthermore, the identification of electronic devices containing frequency sources specifically includes the following steps:

[0013] Step 1): After detecting the time-frequency signal of a frequency source, extract the time-frequency fingerprint features within its time-frequency fingerprint feature region;

[0014] Step 2) Compare the time-frequency fingerprint feature profiles to compare the frequency source type and individual characteristics;

[0015] Step 3) After determining the frequency source type, identify and confirm individual electronic devices containing that type of frequency source.

[0016] Furthermore, it also includes monitoring the operational status of the frequency source or electronic device containing the frequency source to achieve fault diagnosis of the frequency source or electronic device, specifically including the following steps:

[0017] Step 1): After determining the time-frequency fingerprint features within the time-frequency fingerprint feature region of the frequency source, detect the time-frequency fingerprint features of the frequency source at regular intervals.

[0018] Step 2) Compare the time-frequency fingerprint feature region detected and calculated at the current moment with the inherent time-frequency fingerprint feature region of this type of frequency source, and calculate the curve error;

[0019] Step 3): When the curve error is greater than the preset limit, it is determined that the frequency source is faulty and a fault prompt is given; when the curve error is not greater than the preset limit, it is determined that the working status of the frequency source is normal, or the working status of the electronic device containing the frequency source is normal.

[0020] Furthermore, to ensure the accuracy of the calculated time-frequency characteristics, a digital linear phase comparison method is used to determine the frequency stability of the clock signal. The specific steps are as follows:

[0021] Step 1) The clock signal is sent to the signal preprocessor (1a) under test for processing to maintain the original signal characteristics;

[0022] Step 2), the reference signal (3a) is multiplied by a frequency multiplier to n times the nominal value of the frequency of the signal under test, where n is a positive number. The multiplied signal is used as the clock standard signal for sampling by the high-speed phase processor (2a).

[0023] Step 3) Use the first processor to control the high-speed phase processor (2a) to sample the signal under test, and send the collected voltage data of the linear region to the second processor.

[0024] Step 4) The received voltage data is converted into the phase difference between the measured signal and the reference signal source (3a) by the second processor, and the frequency stability of the measured signal is calculated by the phase difference change, and the time-frequency fingerprint feature area is extracted.

[0025] Furthermore, in step 3, the inherent time-frequency fingerprint features of each type of specific frequency source are stored in the information resource library (6a). The second processor is used to compare the time-frequency fingerprint features in the currently extracted time-frequency fingerprint feature area with each inherent time-frequency fingerprint feature to find the type that matches the comparison result.

[0026] Furthermore, the first processor is FGPA(4a).

[0027] Furthermore, the second processor is a single-chip microcomputer (MCU) (5a). Attached Figure Description

[0028] Figure 1 This is a flowchart of the frequency source time-frequency fingerprint characteristic detection method in Embodiment 1 of the present invention;

[0029] Figure 2-1 This is a characteristic curve of the frequency stability of the crystal frequency source in Embodiment 1 of the present invention changing with sampling time;

[0030] Figure 2-2 This is a characteristic curve of the frequency stability of the passive atomic frequency source in Embodiment 1 of the present invention as a function of sampling time;

[0031] Figure 3This is a schematic diagram of the digital linear phase comparison method in Embodiment 1 of the present invention;

[0032] Figure 3 The symbols in the text are explained below:

[0033] 1a, Signal under test preprocessor; 2a, High-speed phase processor; 3a, Reference signal source; 4a, FPGA; 5a, Microcontroller (MCU); 6a, Information resource library;

[0034] Figure 4-1 This is a schematic diagram of the time-frequency fingerprint feature region exhibited by the stability transition region of different individual crystal frequency sources in Embodiment 2 of the present invention;

[0035] Figure 4-2 This is a schematic diagram showing the differences in time-frequency fingerprint feature regions exhibited by the stability transition regions of different cesium atom frequency sources in Embodiment 2 of the present invention;

[0036] Figure 4-3 This is a schematic diagram of the time-frequency fingerprint feature region exhibited by the stability transition region of different crystal frequency source individuals showing health instability characteristics in Embodiment 3 of the present invention. Detailed Implementation

[0037] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0038] Example 1:

[0039] This embodiment proposes a method for detecting the time-frequency fingerprint characteristics of a frequency source. The overall process is as follows: Figure 1 As shown, it includes the following steps:

[0040] Step 1: Acquire the signal emitted by the electronic device and extract the clock signal of the frequency standard source in the electronic device.

[0041] Step 2: Determine the trend curve of the frequency stability of the clock signal as a function of the sampling time interval. Take the last segment of the monotonically decreasing region and the first segment of the non-monotonic region in the curve. The curve region from the last segment of the monotonically decreasing region to the first segment of the non-monotonic region, which shows the change of stability with sampling time, is taken as the time-frequency fingerprint feature region.

[0042] The division steps are as follows: Select the turning point on the frequency stability curve that changes from monotonically decreasing to non-monotonic decreasing as the dividing point. With the dividing point as the boundary, the curve before the dividing point is the monotonically decreasing region, and the curve after the dividing point is the non-monotonic decreasing region.

[0043] In this step, the last segment before the dividing point is taken as the final segment, and the first segment after the dividing point is taken as the preceding segment. The dividing point, that is, the area before and after the turning point, is called the turning zone. The length of the turning zone, that is, the length of the final segment and the preceding segment, is determined according to the analytical characteristics required. The reason for this division is that through experimental studies on a large number of frequency standard sources, it has been confirmed that the frequency stability of commonly used frequency standard sources exhibits two distinct trends with varying sampling time: a monotonically decreasing trend and a non-monotonic decreasing trend.

[0044] Furthermore, this change process varies depending on the physical device of the frequency source. Therefore, regardless of whether it is a passive atomic frequency source or a crystal frequency source, on the one hand, before the transition point, the stability changes monotonically with sampling time, and after the transition point, the stability changes non-monotonically with sampling time. The range of the transition region before and after this transition point will also vary depending on the materials, processes, electronic circuits, processing methods, equipment structure, etc. used to manufacture the frequency source.

[0045] In summary, the frequency stability of the frequency source exhibits two distinct curve characteristics with varying sampling time, such as... Figure 2-1 and Figure 2-2 As shown, this reflects two different characteristics in the frequency stability trends of crystal frequency sources and passive atomic frequency sources as a function of sampling time. It can be seen that the transition regions defined before and after the dividing point exhibit different characteristics depending on the type and individual frequency source, and the length of the transition regions also varies. The selection of the length of the transition region in the detection of time-frequency fingerprint features is based on the ability to accurately identify the time-frequency fingerprint features. There is no definite range, but it requires a certain level of high stability resolution and time resolution.

[0046] In this step, with sufficiently high stability resolution and sampling time resolution, the time-frequency fingerprint characteristics of the frequency source can be accurately detected. In-depth research has revealed that these time-frequency fingerprint characteristics are stable for individual frequency sources and cannot be copied or repeated by other individuals, providing a basis for effectively identifying individual frequency source devices.

[0047] However, detecting this time-frequency fingerprint feature requires a high-resolution measurement of the frequency stability of the frequency source, covering the effective range of both feature segments. Therefore, the detection device needs to have sufficiently high stability resolution and sufficiently high sampling time resolution. Besides existing high-performance testing instruments, the existing digital linear phase comparison method can also be used to detect the frequency stability and sampling time of the clock signal. The principle of this method is as follows: Figure 3 As shown, it includes the following steps:

[0048] (1) Compare the signal f x1 f x2 f x3 (i.e., clock signal) is sent as the signal under test to the signal under test preprocessor (1a) for preprocessing to maintain its original characteristics;

[0049] (2) The reference signal source 3a is multiplied by a frequency multiplier to n times the nominal value of the frequency of the signal under test (n is a positive number, such as 10). The multiplied signal is used as the clock standard signal for sampling by the high-speed phase processor (2a).

[0050] (3) Use FGPA (4a) to control A / D converter (2a) to sample the measured signal and send the collected voltage data of the linear region to the microcontroller MCU (5a);

[0051] (4) The microcontroller (MCU) (5a) converts the received voltage data into the phase difference between the measured signal 1a and the reference signal 3a, and calculates the frequency stability of the measured signal by the phase difference change, and extracts the time-frequency fingerprint feature area.

[0052] In this step, the formula for calculating frequency stability is as follows:

[0053]

[0054] In the formula, σ y (τ) represents frequency stability, τ is the average time of the measurement (sampling time interval), m is the number of samples, and ΔT i+1 ΔT i This represents the phase difference change. Wherein, the phase ΔT i+1 ΔT i The calculation formula is as follows:

[0055]

[0056] In the formula, ΔT n For phase, All angles are described in units of time.

[0057] Unlike traditional frequency stability measurement methods, this reference clock signal does not have the same frequency as the measured signal, but rather a multiple of it. Thus, multiple clock cycles correspond to the period value of a measured signal, with only one, and always one, falling within the linear region near 0 degrees of the measured signal's sinusoidal waveform. This provides a broader range of frequency stability measurement capabilities through digital direct linear phase comparison, ensuring the measurement of the overall stability parameters for crystal oscillators and atomic clocks. For determining the two-segment characteristics of frequency stability in crystal frequency sources and passive atomic frequency sources, this method possesses good ability to precisely measure the stability transition region.

[0058] Step 3: Using the acquired time-frequency fingerprint feature area, analyze and extract the individual characteristics of the time-frequency fingerprint features of the frequency source, and use the time-frequency fingerprint features of the frequency source to identify a specific electronic device containing the frequency source.

[0059] Example 2:

[0060] This embodiment proposes a method for detecting the time-frequency fingerprint characteristics of a frequency source. The difference between this method and the method in Embodiment 1 is that the data information obtained in Embodiment 1 is used to specifically identify the frequency source.

[0061] Specifically, such as Figure 3 The information resource library 6a shown stores the original time-frequency fingerprint feature regions of various types of frequency sources. The MCU is used to compare the current time-frequency fingerprint feature region with the original time-frequency fingerprint feature region and find the type that matches the comparison result, which is the type of frequency source used in the electronic device and its individual.

[0062] The reason for using time-frequency fingerprint feature regions to identify the type of frequency source is based on the research on various characteristics of frequency standard sources. It was found that due to the influence of many inherent physical factors of the frequency standard source itself (such as materials, structure, circuit, devices, working principle and control method, etc.), the time-frequency characteristics of the frequency standard source exhibit unique and unrepeatable (or uncopyable) individual differences in their individual performance.

[0063] Just as the characteristics of human fingerprints can uniquely identify the differences between individuals, the fingerprint feature regions of time and frequency characteristics can also be used to determine the individual differences of electronic devices containing frequency standard sources. This achieves the goal of uniquely identifying a certain frequency standard source through time and frequency fingerprint features, and further achieves the goal of uniquely identifying and determining the application of a specific electronic device through the time and frequency fingerprint feature regions of the frequency source.

[0064] like Figure 4-1 The stability of the different crystal frequency sources shown in the figure varies with sampling time (i.e., Figure 4-1 The horizontal axis in the graph represents the sampling time interval (not the instant) and the transition regions that exhibit time-frequency fingerprint characteristics, as well as... Figure 4-2 The graph shows the time-frequency fingerprint characteristics of the transition regions where the stability of individual cesium atomic frequency sources changes with sampling time (also known as the time interval). The vertical axis represents frequency stability, and the horizontal axis represents the sampling time interval. Figure 4-1 The numbers 1 to 9 in the graph represent the frequency stability curves of the nine frequency sources as a function of time intervals. See the curve information below. Figure 4-1 The lower middle section includes input frequency, amplitude, variance, sampling, time, and the instruments used. Figure 4-2The numbers 1 to 6 in the text represent the frequency stability curves of the other six frequency sources as a function of time intervals. The meanings of the other symbols are similar and will not be repeated here.

[0065] according to Figure 4-1 and Figure 4-2 It can be concluded that, regardless of the frequency source, its stability exhibits a characteristic change pattern with two distinct segments: a monotonically decreasing trend followed by a non-monotonic trend. This transition from monotonically decreasing to non-monotonic change is the turning point—the time-frequency fingerprint feature region—that can be used to identify stability changes in a frequency source device. Furthermore, this feature region often occurs within a sampling time interval ranging from milliseconds to seconds. Unlike the two segments of the stability-sampling time interval curve, which can be represented solely by either the monotonically decreasing or non-monotonic region, this time-frequency fingerprint feature region, composed of both monotonically and non-monotonic regions, can meticulously reflect the individual differences of a frequency source.

[0066] The method of the present invention, by detecting the time-frequency fingerprint characteristics of frequency sources, can better identify and determine a large number of navigation, positioning, radar, communication systems and other electronic devices that use various types of frequency standard sources as clock frequencies, and also facilitates better management of time-frequency information.

[0067] Example 3:

[0068] This embodiment of the method for detecting the time-frequency fingerprint characteristics of a frequency source differs from the methods in Embodiments 1 and 2 in that, after identifying the type of the frequency standard source, this method further includes monitoring the operating status of the frequency source (or an electronic device containing a frequency source) to achieve fault diagnosis of the frequency source (or electronic device). Specifically, it includes the following steps:

[0069] Step 1): After determining the time-frequency fingerprint characteristics of the frequency source, the time-frequency fingerprint characteristics of the frequency source are detected at regular intervals.

[0070] Step 2) Compare the time-frequency fingerprint feature region calculated at the current time with the inherent time-frequency fingerprint feature region of this type of frequency source, and calculate the curve error;

[0071] Step 3): When the curve error is greater than the preset limit, the frequency source is determined to be faulty and a fault prompt is issued; when the curve error is not greater than the preset limit, the frequency source is determined to be in normal working condition.

[0072] The above steps can be utilized Figure 3The process is implemented using an MCU, which is connected to both a frequency standard source and an information resource library. The MCU obtains the signal to be processed from the frequency source, determines the frequency source type and its individual components according to the method in Example 2, periodically calculates the time-frequency fingerprint feature area according to the method in this example, and then obtains the pre-stored standard time-frequency fingerprint feature area of ​​the frequency source of this type from the information resource library. Finally, the MCU provides the judgment result.

[0073] The time-frequency fingerprint feature regions exhibited by the stability transition regions of different crystal frequency sources showing health instability characteristics are as follows: Figure 4-3 As shown, it can be verified that the time-frequency fingerprint features displayed in the figure no longer satisfy the characteristics of the time-frequency fingerprint feature region formed by the two segments that change from monotonically decreasing to non-monotonic decreasing when the frequency source is working normally. Figure 4-3 The numbers 1 to 9 represent the frequency stability curves of the nine frequency sources under unhealthy operating conditions as a function of time intervals; the other symbols have similar meanings. Figure 4-1 I will not go into details.

[0074] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for detecting the time-frequency fingerprint characteristics of a frequency source, characterized in that, Includes the following steps: Step 1: Obtain the clock signal of the frequency source; or, obtain the signal emitted by the electronic device and extract the clock signal of the frequency source in the electronic device from it. Step 2: Determine the frequency stability trend curve of the clock signal as a function of sampling time. This trend curve shows two distinct trends, called the monotonically decreasing trend and the non-monotonic decreasing trend. Select the turning point from the monotonically decreasing trend to the non-monotonic decreasing trend on the frequency stability curve as the dividing point. The curve before the dividing point is called the monotonically decreasing region, and the curve after the dividing point is called the non-monotonic decreasing region. Take the last segment of the monotonically decreasing region and the first segment of the non-monotonic decreasing region in the curve. The region of the curve from the last segment of the monotonically decreasing region to the first segment of the non-monotonic decreasing region, which represents the stability as a function of sampling time, is taken as the time-frequency fingerprint feature region. Step 3: Using the acquired time-frequency fingerprint feature area, analyze and extract the individual characteristics of the time-frequency fingerprint features of the frequency source, which can be used to identify the type of frequency source or monitor the working health status of the frequency source, or to identify electronic devices containing frequency sources or to monitor the working health status of electronic devices.

2. The method for detecting the time-frequency fingerprint characteristics of a frequency source according to claim 1, characterized in that, In step 2, In the monotonically decreasing region, the curve segment before the dividing point is taken as the final segment; in the non-monotonically decreasing region, the curve segment after the dividing point is taken as the first segment. The length of the final and first segments depends on the analytical characteristics of the frequency source.

3. The method for detecting the time-frequency fingerprint characteristics of a frequency source according to claim 1, characterized in that, The identification of electronic devices containing frequency sources specifically includes the following steps: Step 1) After detecting the time-frequency signal of a frequency source, extract the time-frequency fingerprint features within its time-frequency fingerprint feature region; Step 2) Compare the time-frequency fingerprint feature profiles to compare the frequency source type and individual characteristics; Step 3) After determining the frequency source type, identify and confirm individual electronic devices containing that type of frequency source.

4. The method for detecting the time-frequency fingerprint characteristics of a frequency source according to claim 1, characterized in that, It also includes monitoring the operational status of the frequency source or electronic device containing the frequency source, and diagnosing faults in the frequency source or electronic device, specifically including the following steps: Step 1): After determining the time-frequency fingerprint features within the time-frequency fingerprint feature region of the frequency source, the time-frequency fingerprint features of the frequency source are detected at regular intervals. Step 2) Compare the time-frequency fingerprint feature region detected and calculated at the current moment with the inherent time-frequency fingerprint feature region of this type of frequency source, and calculate the curve error; Step 3): When the curve error is greater than the preset limit, the frequency source is determined to be faulty and a fault prompt is issued; when the curve error is not greater than the preset limit, the frequency source is determined to be in normal working condition, or the electronic device containing the frequency source is determined to be in normal working condition.

5. The method for detecting the time-frequency fingerprint characteristics of a frequency source according to claim 1, characterized in that, In step 2, the frequency stability of the clock signal is determined using the digital linear phase comparison method. The specific steps are as follows: Step 1), the clock signal is sent to the signal preprocessor (1a) under test for processing, while maintaining the original signal characteristics; Step 2), the reference signal source (3a) is multiplied by a frequency multiplier to n times the nominal value of the frequency of the signal under test, where n is a positive number. The multiplied signal is used as the clock standard signal for sampling by the high-speed phase processor (2a). Step 3), the first processor controls the high-speed phase processor (2a) to sample the signal under test, and sends the collected voltage data of the linear region to the second processor; Step 4) The second processor converts the received voltage data into the phase difference between the measured signal and the reference signal source (3a), and calculates the frequency stability of the measured signal by the phase difference change, and extracts the time-frequency fingerprint feature area.

6. The method for detecting the time-frequency fingerprint characteristics of a frequency source according to claim 5, characterized in that, In step 3, the inherent time-frequency fingerprint features of each type of specific frequency source are stored in the information resource library (6a). The second processor is used to compare the time-frequency fingerprint features in the currently extracted time-frequency fingerprint feature area with each inherent time-frequency fingerprint feature to find the type that matches the comparison result.

7. The method for detecting the time-frequency fingerprint characteristics of a frequency source according to claim 5, characterized in that, The first processor is an FPGA (4a).

8. The method for detecting the time-frequency fingerprint characteristics of a frequency source according to claim 5 or 6, characterized in that, The second processor is a single-chip microcomputer (MCU) (5a).

Images