A DAC high-speed cable fault detection method and system

Abstract

The present invention relates to the technical field of cable fault detection, and more particularly to a fault detection method and system for a DAC high-speed cable. The method comprises: decomposing test data to obtain multiple component signals, and obtaining a degree of change of the component signals based on fluctuations of the component signals; setting sliding windows, and obtaining directional anisotropy of adjacent sliding windows based on a change trend of data within the sliding windows; thereby obtaining the complexity of the change of each component signal; further obtaining the attenuation probability of time domain data; obtaining the variability of the time domain data based on the attenuation probability of the time domain data and the opening height and opening width of eye diagram data; thereby obtaining the degree of attenuation of the time domain data; screening abnormal data based on the attenuation degree of the time domain data, and detecting faults in the DAC high-speed cable. The method can accurately analyze signal attenuation caused by electromagnetic interference and other conditions during data transmission of the DAC high-speed cable, provide timely warnings, and avoid signal loss.

Description

Technical Field

[0001] The present invention relates to the technical field of cable fault detection, and in particular to a fault detection method and system for a DAC high-speed cable. Background Art

[0002] DAC high-speed cable, or Direct Attach Cable, is a fixed-length cable assembly equipped with fixed connectors at both ends. The interface cannot be replaced, and the modular connector cannot be separated from the copper cable. DAC high-speed cable does not contain optical lasers and electronic components, so it can only transmit electrical signals rather than optical signals. This feature allows DAC high-speed cable to significantly reduce costs and power consumption in short-distance applications. With the increasing demand for bandwidth in data center networks, DAC high-speed cable is widely used for connections between servers and switches due to its high efficiency, low latency, and economic characteristics. However, DAC cables are susceptible to physical damage, signal attenuation, and electromagnetic interference in long-term use, resulting in degraded network performance and link interruption. Testing the signal attenuation of DAC high-speed cables can help confirm the integrity and reliability of the signal during transmission. By detecting the degree of attenuation, problems that may lead to increased bit error rates can be identified in a timely manner to ensure the correctness of data transmission.

[0003] In existing technologies, fully automated SI high-frequency parameter testing is a highly efficient and accurate method for evaluating signal integrity in high-speed circuits. Its primary purpose is to detect and analyze issues such as attenuation, reflection, crosstalk, and distortion that high-speed signals may experience during transmission. However, when analyzing signals obtained from SI high-frequency parameter testing, the insertion loss of DAC cables increases nonlinearly with frequency due to the skin effect and dielectric loss when transmitting high-frequency signals. This nonlinear characteristic distorts the signal waveform, leading to significant errors in the analysis of test data. Summary of the Invention

[0004] In order to solve the above problems, the present invention proposes a fault detection method and system for a DAC high-speed cable.

[0005] An embodiment of the present invention provides a method for detecting a fault in a DAC high-speed cable, the method comprising the following steps:

[0006] Acquire test data of a DAC high-speed cable, wherein the test data includes time domain data and eye diagram data;

[0007] Decompose the time domain data to obtain multiple component signals. The degree of change of the component signals is obtained based on the fluctuation of each component signal. The sliding window is adjusted based on the degree of change of the component signals. The directional anisotropy of adjacent sliding windows is obtained based on the change trend of the data in the sliding window. The change complexity of each component signal is obtained based on the numerical value of the data in the sliding window. The attenuation possibility of the time domain data is obtained based on the change complexity and frequency of each component signal.

[0008] According to the attenuation possibility of the time domain data and the opening height and opening width of the eye diagram data, the variability of the time domain data is obtained; according to the variability of the time domain data and the distribution characteristics of the eye diagram data, the attenuation degree of the time domain data is obtained;

[0009] Abnormal data is obtained by filtering out the attenuation degree of the time domain data, and faults of the DAC high-speed cable are detected.

[0010] Optionally, the step of decomposing the time domain data to obtain multiple component signals and obtaining the degree of change of the component signals according to the fluctuation of each component signal includes the following specific steps:

[0011] Perform EMD decomposition on the time domain data to obtain multiple component signals; the degree of change of the component signals is obtained, and the calculation formula is as follows:

[0012]

[0013] Where p represents the degree of change of the component signal, N represents the number of component signals, and s l represents the variance of all amplitudes in the lth component signal, and sigmoid() represents the sigmoid function.

[0014] Optionally, the step of adjusting the sliding window according to the degree of change of the component signal includes the following specific steps:

[0015] The initial length of the sliding window is preset and adjusted according to the degree of change of the component signal. The calculation formula is as follows:

[0016]

[0017] Where T represents the length of the sliding window, p represents the degree of change of the component signal, A represents the preset length, Represents the ceiling function.

[0018] Optionally, obtaining the directional anisotropy of adjacent sliding windows according to the change trend of data in the sliding window includes the following specific steps:

[0019] Based on the length of the sliding window, several sliding windows are obtained for each component signal. The step size and length of the sliding window are equal, so there is no duplicate data between different sliding windows. Least squares fitting is performed on the data in any sliding window to obtain a fitting curve. The principal component direction of the fitting curve is obtained by the PCA algorithm and used as the data distribution direction of the sliding window.

[0020] The calculation formula for obtaining the directional anisotropy of adjacent sliding windows is as follows:

[0021] q l(i,i+1) =θ li ×θ l(i+1)

[0022] Where q l(i,i+1) Indicates the directional anisotropy between the ith sliding window and the i+1th sliding window of the lth component signal, θ li represents the data distribution direction of the i-th sliding window of the l-th component signal, θ l(i+1) Indicates the data distribution direction of the i+1th sliding window of the lth component signal.

[0023] Optionally, obtaining the change complexity of each component signal includes the following specific steps:

[0024]

[0025] Where V l represents the complexity of the change of the lth component signal, represents the maximum value of the data mean in all sliding windows of the lth component signal, represents the minimum value of the data mean in all sliding windows of the lth component signal, n represents the number of sliding windows in the component signal, q l(i,i+1) It represents the directional anisotropy between the i-th sliding window and the i+1-th sliding window of the l-th component signal, and exp[] represents an exponential function with a natural constant as the base.

[0026] Optionally, obtaining the attenuation possibility of the time domain data according to the change complexity and frequency of each component signal includes the following specific steps:

[0027]

[0028] Where w represents the attenuation possibility of time domain data, f l Represents the frequency of the lth component signal, V l represents the complexity of the change of the lth component signal, N represents the number of component signals, and sigmoid() represents the sigmoid function.

[0029] Optionally, obtaining the variability of the time domain data according to the attenuation possibility of the time domain data and the opening height and opening width of the eye diagram data includes the following specific steps:

[0030]

[0031] Where K represents the variability of the time domain data, w represents the attenuation possibility of the time domain data, H represents the opening height of the eye diagram, and E represents the opening width of the eye diagram.

[0032] Optionally, obtaining the attenuation degree of the time domain data according to the variability of the time domain data and the distribution characteristics of the eye diagram data includes the following specific steps:

[0033] Obtain the upper and lower vertices of the eye diagram, as well as the two middle intersection points. Use the line connecting the upper and lower vertices as the symmetry axis, fold the eye diagram in half, and obtain the spatial distance between the two middle intersection points. Use the template matching algorithm to match the eye diagrams on both sides of the symmetry axis to obtain the matching degree of the eye diagrams on both sides of the symmetry axis.

[0034] The calculation method for obtaining the attenuation degree of time domain data is as follows:

[0035]

[0036] Where γ represents the attenuation degree of time domain data, K represents the variability of time domain data, d represents the spatial distance between the two intermediate intersection points, g represents the matching degree of the eye diagrams on both sides of the symmetry axis, and sigmoid() represents the sigmoid function.

[0037] Optionally, the step of filtering out abnormal data according to the attenuation degree of the time domain data includes the following specific steps:

[0038] A fault threshold is preset. If the attenuation of the time domain data is greater than the fault threshold, the test data is abnormal and the DAC high-speed cable is faulty.

[0039] Another embodiment of the present invention provides a DAC high-speed cable fault detection system, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above method when executing the computer program.

[0040] The beneficial effects of the technical solution of the present invention are as follows: in the process of analyzing the attenuation characteristics of the test data, the time domain data of the test data is first decomposed to obtain multiple decomposed signals; then the change complexity of the component signals is obtained according to the changes of the component signals; further the attenuation possibility of the time domain data is obtained; because the time domain data is subject to greater interference and the degree of data change is greater, the long-term time series changes of the signal are analyzed through the eye diagram data; the variability of the time domain data is obtained according to the opening height, opening width and attenuation degree of the time domain data in the eye diagram data; then the attenuation degree of the time domain data is obtained according to the variability of the time domain data; and the fault condition of the tested DAC high-speed cable is obtained according to the attenuation degree of the time domain data. This method can accurately analyze the signal attenuation of the DAC high-speed cable during data transmission due to electromagnetic interference and other conditions, thereby performing timely manual intervention to avoid signal loss during transmission. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0042] Figure 1 This is a flowchart of the steps of the fault detection method of the DAC high-speed cable of the present invention;

[0043] Figure 2 This is an example diagram of time domain data in the test data;

[0044] Figure 3 This is an example of the eye diagram data in the test data;

[0045] Figure 4 Schematic diagram of the distribution characteristics of eye diagram data. DETAILED DESCRIPTION

[0046] To further illustrate the technical means and effectiveness of the present invention in achieving its intended objectives, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail the specific implementation, structure, features, and effectiveness of the DAC high-speed cable fault detection method proposed by the present invention. In the following description, different references to "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0047] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0048] The specific scheme of the fault detection method of the DAC high-speed cable provided by the present invention is described in detail below with reference to the accompanying drawings.

[0049] See also Figure 1 , which shows a flowchart of a method for detecting a fault of a DAC high-speed cable provided by one embodiment of the present invention, the method comprising the following steps:

[0050] Step S001: Acquire test data of the DAC high-speed cable.

[0051] 1. Test system architecture design:

[0052] 1. Core equipment configuration:

[0053] 1) High-speed oscilloscope: bandwidth ≥100 GHz, used for eye diagram capture and dynamic signal analysis.

[0054] 2. Auxiliary modules:

[0055] 1) Temperature-controlled environmental chamber: Maintains constant temperature (25 ± 0.5°C) and humidity (≤ 60% RH) to reduce the impact of material Dk / Df fluctuations on the test;

[0056] 2) Automatic fixture system: uses low-loss SMA / 2.92mm connector, with impedance matching accuracy of ±2%.

[0057] 2. Key steps of experimental arrangement:

[0058] 1. Fixture calibration and de-embedding processing:

[0059] 1) Use the SOLT (Short-Open-Load-Thru) calibration kit to eliminate fixture parasitic parameters (such as SMA connector capacitance > 1pF);

[0060] 2) Compensate for fixture loss using the VNA's built-in de-embedding algorithm (error < 0.05dB / 40GHz).

[0061] 2.DAC cable fixing and connection:

[0062] 1) Connect both ends of the DAC cable to the automatic clamping system to ensure uniform distribution of mechanical stress (bending radius ≥ 5 times the wire diameter);

[0063] 2) Allow the device to stand for ≥12 hours before testing to eliminate impedance fluctuations caused by temperature drift (impedance tolerance ±5%).

[0064] 3. Test parameter configuration:

[0065] 1) Frequency domain test: Set the VNA sweep frequency range to 0.1-40 GHz and measure S11 (return loss), S21 (insertion loss), and crosstalk (S31 / S41).

[0066] 2) Time domain test: TDR sends a 10ps rising edge pulse to locate the impedance mutation point (resolution ±1mm);

[0067] 3) Dynamic test: The BERT transmits a PRBS31 pattern, and the oscilloscope captures the eye diagram and calculates the eye height (≥50mV@112Gbps) and total jitter (≤0.1UI).

[0068] 3. Data acquisition process:

[0069] 1. First, initialize the system. Start the temperature-controlled environmental chamber and load a preset test script (such as a Python or LabVIEW program). Simultaneously connect core equipment such as a vector network analyzer (VNA), a high-speed oscilloscope, and a bit error rate tester (BERT). During this stage, strictly monitor the ambient temperature stability (fluctuation ≤ ±0.5°C) to eliminate the impact of external temperature drift on the test.

[0070] 2. Next, perform calibration and compensation. Use the SOLT (Short-Open-Load-Thru) calibration kit to calibrate the test fixture and eliminate fixture parasitic parameters (such as connector capacitance and inductance).

[0071] 3. After calibration, the S parameter scanning phase begins. The VNA performs multi-port synchronous measurements on the DAC cable to obtain full frequency domain parameters including return loss (S11), insertion loss (S21), and crosstalk (S31 / S41), as well as time domain data and eye diagram data. Time domain data is as follows: Figure 2 As shown, the eye diagram data is as follows Figure 3 shown.

[0072] Step S002: Decompose the time domain data to obtain multiple component signals, and obtain the degree of change of the component signals based on the fluctuation of each component signal; adjust the sliding window according to the degree of change of the component signal; obtain the directional anisotropy of adjacent sliding windows based on the change trend of the data in the sliding window; obtain the change complexity of each component signal based on the numerical size of the data in the sliding window; and obtain the attenuation possibility of the time domain data based on the change complexity and frequency of each component signal.

[0073] It's important to note that DAC cables face stringent performance requirements in high-frequency scenarios. Faults can lead to signal attenuation, excessive bit error rates, or system communication interruption, triggering cascading risks, particularly in data centers and 5G base stations. Fully automated signal integrity (SI) high-frequency parameter testing, combined with frequency, time, and dynamic performance analysis, accurately locates physical defects and electrical performance deviations in cables.

[0074] It's also worth noting that when testing DAC high-speed cables, both time-domain waveforms and eye diagrams play a crucial role. The time-domain waveform provides detailed information about signal amplitude variations over time, enabling analysis of the signal's rise and fall times, as well as its stability. The eye diagram, on the other hand, overlays multiple cycles of time-domain waveforms to visually display signal quality characteristics, such as the aperture height and width. The aperture height reflects the signal's noise tolerance, while the aperture width indicates the signal's timing jitter.

[0075] It's important to note that the rise and fall times of a time-domain waveform reflect the time it takes for a signal to rise from a low level to a high level and fall back from a high level to a low level. Signal attenuation typically manifests as a gradual decrease in the amplitude of the time-domain waveform. As the signal attenuates, the time-domain waveform not only loses amplitude but may also become distorted, leading to increased jitter. An eye diagram, formed by superimposing multiple repeated signal waveforms, can intuitively reflect signal quality, particularly characteristics such as attenuation and jitter during transmission. However, electromagnetic coupling between wire pairs causes near-end crosstalk (NEXT) and far-end crosstalk (FEXT), which can cause random fluctuations in signal amplitude and increase the bit error rate (BER). Fluctuating signals act as noise, and when the signal attenuates, the impact of noise increases, manifesting as a rise in the noise floor in the eye diagram. As the signal amplitude decreases, the relative strength of noise and interference increases, blurring the eye opening in the eye diagram. This indicates a decrease in signal quality and an increase in the bit error rate (BER). Therefore, by analyzing the relationship between time-domain data and eye diagrams, faults in DAC high-speed cables can be identified and their quality assessed.

[0076] It should be further explained that in the time domain signal, the greater the fluctuation of the received signal, the greater the interference generated by the signal during transmission by the DAC high-speed cable. Therefore, the time domain signal is first described to obtain the degree of interference of the signal; and by performing EMD decomposition on the time domain data, the component signals obtained can represent the complexity of the changes of the original signal. If the complexity of the change of the original signal is higher, the number of component signals obtained by decomposition will be greater, and the degree of change of the signal will be higher.

[0077] Specifically, the time domain data is decomposed by EMD to obtain multiple component signals; the degree of change of the component signals is obtained, and the calculation formula is as follows:

[0078]

[0079] Where p represents the degree of change of the component signal, N represents the number of component signals, and s l represents the variance of all amplitudes in the lth component signal, sigmoid() represents the sigmoid function, and this embodiment is used for normalization processing.

[0080] What needs to be explained is that l The variance of the component signal amplitude reflects the fluctuation and then characterizes the degree of change. The larger the value, the greater the degree of change of the component signal. Because in the EMD algorithm, the decomposition of the signal is based on frequency, but if the degree of change of a single component signal is greater, then the original signal is more affected, and the corresponding degree of change of the original signal will also be greater. The degree of change p of the component signal here reflects the overall degree of change of all component signals.

[0081] It's important to note that the variations in the component signals described above describe overall characteristics. When analyzing a single component signal, cable quality is assessed by transmitting an electrical signal to the cable and then monitoring the resulting fluctuations in the echo signal. Signal attenuation, manifested as cable conductor and dielectric loss, is more pronounced in high-frequency components, resulting in an overall decrease in signal amplitude. Therefore, analyzing the variations in a single component signal is crucial.

[0082] It should be further explained that a single component signal represents the detailed changes of the original time domain signal. If the original signal is attenuated, the high-frequency component signal will be attenuated. In order to capture the changes in the signal at different times, a sliding window needs to be set for analysis.

[0083] Specifically, the initial length of the sliding window is preset. In this embodiment, the initial length is described as A=5. The length of the sliding window is adjusted according to the degree of change of the component signal. The calculation formula is as follows:

[0084]

[0085] Where T represents the length of the sliding window, p represents the degree of change of the component signal, A represents the preset length, Represents the ceiling function.

[0086] Furthermore, based on the length of the sliding window, several sliding windows are obtained for each component signal. The step size of the sliding window is equal to the length, so there is no duplicate data between different sliding windows. Least squares fitting is performed on the data in any sliding window to obtain a fitting curve. Least squares fitting is a well-known technology and will not be described in detail here. The principal component direction of the fitting curve is obtained by the PCA algorithm and used as the data distribution direction of the sliding window. The calculation formula for obtaining the directional anisotropy of adjacent sliding windows is as follows:

[0087] q l(i,i+1) =θ li ×θ l(i+1)

[0088] Where q l(i,i+1) Indicates the directional anisotropy between the ith sliding window and the i+1th sliding window of the lth component signal, θ li represents the data distribution direction of the i-th sliding window of the l-th component signal, θ l(i+1) Indicates the data distribution direction of the i+1th sliding window of the lth component signal.

[0089] Furthermore, based on the directional anisotropy of adjacent sliding windows and the numerical value of the data in the sliding window, the complexity of the change of each component signal is obtained, and the calculation formula is as follows:

[0090]

[0091] Where V l represents the complexity of the change of the lth component signal, represents the maximum value of the data mean in all sliding windows of the lth component signal, represents the minimum value of the data mean in all sliding windows of the lth component signal, n represents the number of sliding windows in the component signal, q l(i,i+1) represents the directional anisotropy between the i-th sliding window and the i+1-th sliding window of the l-th component signal, exp[] represents an exponential function with a natural constant as the base. This embodiment adopts the exp[-x] model to present the inverse proportional relationship and normalization processing, x is the input of the model, and the implementer can set the inverse proportional function and normalization function according to actual conditions.

[0092] What needs to be explained is that It represents the amplitude change of the lth component signal. The greater the difference, the greater the change of the time domain signal during the transmission process, and therefore the more serious the signal attenuation may be. The directional anisotropy of the sliding window represents the change of the signal amplitude. The greater the directional anisotropy, the closer the data distribution direction trend of the continuous sliding window is, indicating that the signal is showing a state of continuous attenuation or growth. Here, q l(i,i+1)The value of can be positive or negative. When it is negative, it means that the signal is showing a decaying trend.

[0093] Furthermore, the attenuation probability of the time domain data is obtained according to the complexity and frequency of the changes of each component signal. The calculation formula is as follows:

[0094]

[0095] Where w represents the attenuation possibility of time domain data, f l Represents the frequency of the lth component signal, V l represents the complexity of the change of the lth component signal, N represents the number of component signals, and sigmoid() represents the sigmoid function, which is used for normalization processing in this embodiment.

[0096] It should be noted that because the high-frequency component attenuates more significantly when signal attenuation occurs, the frequency of the component signal is used as the reference weight here. If the signal change complexity of the high-frequency signal is greater, then the possibility of high-frequency signal attenuation is greater, and therefore the possibility of signal attenuation in the original test data is greater.

[0097] So far, the attenuation probability of the time domain data has been obtained.

[0098] Step S003: Obtain the variability of the time domain data according to the attenuation possibility of the time domain data and the opening height and opening width of the eye diagram data; obtain the attenuation degree of the time domain data according to the variability of the time domain data and the distribution characteristics of the eye diagram data.

[0099] It should be noted that, based on the above analysis, it is possible to obtain signal attenuation. However, when testing DAC high-speed cables, environmental noise, ground interference, crosstalk between signal lines, etc. will cause the noise level to rise, resulting in reduced signal time stability, which is manifested as irregularities in the signal edges in the time domain waveform, with random fluctuations or spikes, affecting signal clarity. Therefore, it is inaccurate to judge signal attenuation based solely on changes in the time domain waveform. Eye diagrams are a very effective tool for analyzing the integrity and quality of high-speed digital signals, especially in assessing whether signals have attenuated. By superimposing multiple cycles of signals, eye diagrams can intuitively display the high and low levels, opening height and width of the signal, and can simultaneously display multiple signal characteristics, while observing the impact of attenuation on multiple parameters.

[0100] It's important to note that a time-domain waveform is a graph that shows how signal amplitude changes over time. An eye diagram is generated by superimposing multiple cycles of time-domain waveforms. The eye diagram overlays repeated signal waveforms, with time as the horizontal axis and amplitude as the vertical axis. The characteristics of the time-domain waveform (such as rise time, fall time, amplitude, and noise) directly influence the shape and characteristics of the eye diagram. For example, the faster the rise and fall rates of the time-domain waveform, the larger the eye opening.

[0101] Specifically, the eye opening height H and opening width E are obtained; then, combined with the attenuation possibility of the time domain data, the variability of the time domain data is obtained, and the calculation formula is as follows:

[0102]

[0103] Where K represents the variability of the time domain data, w represents the attenuation possibility of the time domain data, H represents the opening height of the eye diagram, and E represents the opening width of the eye diagram.

[0104] It should be noted that the greater the possibility of attenuation of time domain data, the greater the abnormal change of the signal in the time domain; and the decrease in opening height and opening width indicates that the amplitude of the signal decreases, which is caused by attenuation, so the variability of time domain data is greater.

[0105] It should be further explained that when the signal is not attenuated, the opening of the eye diagram should be symmetrical and regular in shape; if the signal is attenuated, the opening of the eye diagram will be asymmetrical and tilted to one side; and the opening of the eye diagram may become blurred or even completely closed. The attenuation degree of the time domain data is obtained based on the above characteristics.

[0106] Specifically, such as Figure 4 As shown, first obtain the upper and lower vertices of the eye diagram, as well as the two middle intersection points; use the line connecting the upper vertex and the lower fixed point as the symmetry axis, fold the eye diagram in half, and obtain the spatial distance d between the two middle intersection points; then use the template matching algorithm to match the eye diagrams on both sides of the symmetry axis to obtain the matching degree g of the eye diagrams on both sides of the symmetry axis; the attenuation degree calculation formula of the time domain data is as follows:

[0107]

[0108] Wherein, γ represents the attenuation degree of time domain data, K represents the variability of time domain data, d represents the spatial distance between the two intermediate intersection points, g represents the matching degree of the eye diagrams on both sides of the symmetry axis, and sigmoid() represents the sigmoid function. This embodiment is used for normalization processing.

[0109] At this point, the attenuation degree of the time domain data is obtained.

[0110] Step S004: Filter out abnormal data according to the attenuation degree of the time domain data, and detect the fault of the DAC high-speed cable.

[0111] It should be noted that after obtaining the attenuation degree of the time domain data, it is then determined based on the attenuation degree of the signal whether there is a signal attenuation fault when testing the DAC high-speed cable.

[0112] Specifically, a fault threshold is preset. In this embodiment, the fault threshold is described as 0.6. If the attenuation degree of the time domain data is greater than the fault threshold, it means that the test data is attenuated to a large extent. The main cause of signal attenuation is electromagnetic interference. Therefore, it is necessary to eliminate signal interference in a timely manner to avoid signal loss during transmission. If the test data is abnormal, the DAC high-speed cable is faulty.

[0113] Through the above steps, a fault detection method for a DAC high-speed cable is completed.

[0114] Another embodiment of the present invention provides a DAC high-speed cable fault detection system, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, a DAC high-speed cable fault detection method in steps S001 to S004 is implemented.

[0115] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A DAC high-speed cable fault detection method, characterized in that: The method comprises the following steps: Acquire test data of a DAC high-speed cable, wherein the test data includes time domain data and eye diagram data; Decompose the time domain data to obtain multiple component signals. The degree of change of the component signals is obtained based on the fluctuation of each component signal. The sliding window is adjusted based on the degree of change of the component signals. The directional anisotropy of adjacent sliding windows is obtained based on the change trend of the data in the sliding window. The change complexity of each component signal is obtained based on the numerical value of the data in the sliding window. The attenuation possibility of the time domain data is obtained based on the change complexity and frequency of each component signal. According to the attenuation possibility of the time domain data and the opening height and opening width of the eye diagram data, the variability of the time domain data is obtained; according to the variability of the time domain data and the distribution characteristics of the eye diagram data, the attenuation degree of the time domain data is obtained; Filter out abnormal data based on the attenuation level of time domain data and detect DAC high-speed cable faults; The specific steps of obtaining the directional anisotropy of adjacent sliding windows according to the changing trend of the data in the sliding window are as follows: Based on the length of the sliding window, several sliding windows are obtained for each component signal. The step size and length of the sliding window are equal, so there is no duplicate data between different sliding windows. Least squares fitting is performed on the data in any sliding window to obtain a fitting curve. The principal component direction of the fitting curve is obtained by the PCA algorithm and used as the data distribution direction of the sliding window. The calculation formula for obtaining the directional anisotropy of adjacent sliding windows is as follows: ； Where, Indicates the The first component signal The sliding window and The directional anisotropy of the sliding window, Indicates the The first component signal The data distribution direction of the sliding window, Indicates the The first component signal The data distribution direction of the sliding window; The specific steps of obtaining the attenuation degree of the time domain data according to the variability of the time domain data and the distribution characteristics of the eye diagram data are as follows: Obtain the upper and lower vertices of the eye diagram, as well as the two middle intersection points. Use the line connecting the upper and lower vertices as the symmetry axis, fold the eye diagram in half, and obtain the spatial distance between the two middle intersection points. Use the template matching algorithm to match the eye diagrams on both sides of the symmetry axis to obtain the matching degree of the eye diagrams on both sides of the symmetry axis. The calculation method for obtaining the attenuation degree of time domain data is as follows: ； Where, Indicates the attenuation degree of time domain data, represents the variability of time domain data, Represents the spatial distance between the two intermediate intersection points, Indicates the matching degree of the eye diagrams on both sides of the symmetry axis, Represents the sigmoid function.

2. The method for detecting a fault of a DAC high-speed cable according to claim 1, wherein: Decomposing the time domain data to obtain multiple component signals and obtaining the degree of change of the component signals according to the fluctuation of each component signal includes the following specific steps: Perform EMD decomposition on the time domain data to obtain multiple component signals; the degree of change of the component signals is obtained, and the calculation formula is as follows: ； Where, Indicates the degree of change of the component signal, represents the number of component signals, Indicates the The variance of all amplitudes in the component signal, Represents the sigmoid function.

3. The fault detection method for a DAC high-speed cable according to claim 1, wherein: The specific steps of adjusting the sliding window according to the degree of change of the component signal are as follows: The initial length of the sliding window is preset and adjusted according to the degree of change of the component signal. The calculation formula is as follows: ； Where, represents the length of the sliding window, Indicates the degree of change of the component signal, Indicates the preset length, Represents the ceiling function.

4. The method for detecting a fault of a DAC high-speed cable according to claim 1, wherein: The specific steps of obtaining the change complexity of each component signal are as follows: ； Where, Indicates the The complexity of the change of the component signal, Indicates the The maximum value of the data mean in all sliding windows of the component signal, Indicates the The minimum value of the data mean in all sliding windows of the component signal, represents the number of sliding windows in the component signal, Indicates the The first component signal The sliding window and The directional anisotropy of the sliding window, Represents an exponential function with a natural constant as its base.

5. The fault detection method for a DAC high-speed cable according to claim 1, wherein: The specific steps of obtaining the attenuation possibility of the time domain data according to the change complexity and frequency of each component signal include the following: ； Where, represents the attenuation possibility of time domain data, Indicates the The frequency of the component signal, Indicates the The complexity of the change of the component signal, represents the number of component signals, Represents the sigmoid function.

6. The method for detecting a fault of a DAC high-speed cable according to claim 1, wherein: The method of obtaining the variability of the time domain data based on the attenuation possibility of the time domain data and the opening height and opening width of the eye diagram data includes the following specific steps: ； Where, represents the variability of time domain data, Represents the attenuation possibility of time domain data, Indicates the opening height of the eye diagram, Indicates the width of the eye opening.

7. The method for detecting a fault of a DAC high-speed cable according to claim 1, wherein: The specific steps of screening out abnormal data according to the attenuation degree of time domain data are as follows: A fault threshold is preset. If the attenuation of the time domain data is greater than the fault threshold, the test data is abnormal and the DAC high-speed cable is faulty.

8. A DAC high-speed cable fault detection system, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the computer program is executed by a processor, the steps of the fault detection method for a DAC high-speed cable according to any one of claims 1 to 7 are implemented.

Images