Highway 5g and optical fiber sensing composite network signal attenuation compensation method

By collecting and diagnosing the state parameters of communication and sensing signals in real time through 5G fronthaul networks and distributed optical fiber sensing systems along highways, and generating differentiated compensation strategies at multiple time scales, communication signals and sensing signals can be mutually referenced and coordinated for adjustment. This solves the problem of independent compensation in existing technologies and improves the adaptability and effectiveness of compensation.

CN123122652APending Publication Date: 2026-11-17HEBEI JITONG ROAD&BRIDGE CONSTRUCT CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611278978.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-21
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

In the existing 5G fronthaul network and distributed optical fiber sensing system co-transmission along highways, the compensation of communication signals and sensing signals are carried out independently, failing to effectively utilize the compensation information of the other party, and the compensation methods lack adaptability to environmental changes.

Method used

By acquiring the state parameters of communication signals and sensor signals, as well as environmental physical parameters in real time, online diagnosis is performed to generate differentiated compensation strategies at multiple time scales. These strategies serve as mutual references during the compensation process, enabling closed-loop adjustment. The channel state information of the communication signals provides a global reference baseline for the sensor signals, while the environmental parameters of the sensor signals provide feedforward control for the communication signals.

Benefits of technology

It achieves coordinated compensation of communication signals and sensor signals, improves adaptability to environmental changes, enhances the rationality and effectiveness of compensation, and adapts to continuous changes along the highway.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123122652A_ABST
    Figure CN123122652A_ABST
Patent Text Reader

Abstract

This invention relates to the field of optical communication technology, specifically disclosing a method for signal attenuation compensation in a 5G and fiber optic sensing composite network for highways. The method involves real-time acquisition of communication signal state parameters, sensor signal state parameters, and environmental physical parameters along the fiber optic cable. Through correlation analysis of these three data sources, it diagnoses the attenuation type of each fiber segment online, classifying attenuation into four types: intrinsic attenuation, temperature drift attenuation, vibration stress attenuation, and sudden attenuation. Based on the diagnosed attenuation type, a corresponding compensation strategy is matched from a preset strategy library. During compensation, the communication signal is predictively adjusted based on the environmental physical parameters extracted from the sensor signal, while the sensor signal is baseline-calibrated based on the channel state information extracted from the communication signal. The compensation effect is evaluated after execution. This invention enables the communication signal and sensor signal to reference each other and adjust collaboratively during the compensation process, overcoming the technical deficiencies of existing technologies.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of optical communication technology, and in particular to a method for signal attenuation compensation in a high-speed 5G and fiber optic sensing composite network. Background Technology

[0002] With the large-scale deployment of 5G communication networks and the widespread application of distributed fiber optic sensing technology in infrastructure monitoring, 5G fronthaul networks are beginning to appear along highways. The practical need for shared-fiber transmission in distributed fiber optic sensing systems. Utilizing a single optical fiber to simultaneously carry communication and sensing signals can significantly reduce the deployment cost of sensing systems along highways.

[0003] In existing co-fiber transmission schemes, compensation for communication signals and compensation for sensing signals are basically performed independently.

[0004] For communication links, the common practice is to insert optical amplifiers into the link to compensate for fiber loss, or to use an equalizer at the receiving end to compensate for and restore the signal.

[0005] For sensing systems, the main focus is on To compensate for the interference fading problem.

[0006] As for the integration of communication and sensing, most current solutions focus on how to prevent communication and sensing signals from interfering with each other during transmission, such as through wavelength division multiplexing and planning reverse transmission for isolation.

[0007] Analysis of the above-mentioned existing technologies reveals that the communication link and the sensing link are treated as two unrelated systems. The channel state information carried by the communication signal during transmission can actually reflect the loss status of the optical fiber link, but this information is not transmitted to the sensing system as a reference.

[0008] Moreover, environmental changes such as temperature and vibration along the optical fiber, which are sensed by the sensor signals, can actually predict the impending attenuation of the communication signal, but these are not used for communication compensation.

[0009] Both types of signals possess the compensation information needed by the other, but existing solutions fail to utilize them. Furthermore, existing compensation methods are mostly open-loop, ending once compensation is completed, lacking verification of compensation effectiveness and subsequent parameter optimization, making it difficult to adapt to the continuously changing environmental conditions along highways. Summary of the Invention

[0010] To address the aforementioned technical problems, this invention provides a method for signal attenuation compensation in a highway 5G and fiber optic sensing composite network, enabling communication signals and sensing signals to reference each other and adjust collaboratively during the compensation process.

[0011] The signal attenuation compensation method for highway 5G and fiber optic sensing composite networks of the present invention includes the following steps:

[0012] Step 1: Real-time acquisition of the first state parameters of communication signals and the second state parameters of sensing signals in the composite network, and simultaneous acquisition of environmental physical parameters along the optical fiber. Based on the correlation analysis of the first state parameters, the second state parameters and the environmental physical parameters, online diagnosis of signal attenuation sources along the optical fiber is performed, and attenuation type labels for each optical fiber segment are output.

[0013] Step 2: Based on the attenuation type label, match the corresponding compensation strategy from the preset differentiated compensation strategy library to generate a multi-timescale compensation scheme that includes compensation methods and compensation parameters;

[0014] Step 3: Execute the compensation scheme. During the execution process, the environmental physical parameters extracted from the communication signal are predictively adjusted based on the compensation reference sensor signal, and the channel state information extracted from the communication signal is baseline calibrated based on the compensation reference sensor signal.

[0015] Step four: Evaluate the quality of the compensated communication signal and the quality of the sensor signal, calculate the compensation gain of the communication signal and the compensation gain of the sensor signal respectively, and adjust the diagnostic threshold in step one and the compensation parameters in step two in a closed loop based on the comparison results of the compensation gain and the preset threshold.

[0016] Furthermore, in step one, the attenuation type label includes intrinsic attenuation, temperature drift attenuation, vibration stress attenuation, and sudden attenuation;

[0017] The correlation analysis is implemented in the following manner:

[0018] 1.1 Compare the communication signal amplitude attenuation curve in the first state parameter with the backscattered light intensity distribution curve in the second state parameter segment by segment, and determine the intrinsic attenuation based on the comparison results;

[0019] 1.2 Extract the temperature change rate and backscattered light intensity change rate in each fiber segment, and perform time-domain cross-correlation analysis on the two in each segment. Determine the temperature drift attenuation segment by segment based on the cross-correlation analysis results.

[0020] 1.3 Perform coherent analysis on the vibration spectrum intensity in the environmental physical parameters and the time-domain fluctuation amplitude of the backscattered light intensity in the second state parameter, and determine the vibration stress attenuation based on the coherent analysis results;

[0021] 1.4 When the signal attenuation changes abruptly beyond a preset threshold within a preset time window, it is determined to be a sudden attenuation.

[0022] Furthermore, in step two,

[0023] The multi-timescale compensation scheme comprises four compensation levels:

[0024] For intrinsic attenuation of the underlying baseline, the adjustment period for the compensation parameters is on the order of minutes.

[0025] For mid-layer tracking compensation to address temperature drift decay, the compensation parameter adjustment cycle is on the order of seconds.

[0026] For real-time compensation of vibration stress attenuation in the upper layer, the adjustment cycle of the compensation parameters is in the millisecond range.

[0027] The top-level emergency compensation for sudden attenuation has a response time in the microsecond range;

[0028] The above four levels of compensation operate in parallel, with the top-level emergency compensation having the highest execution priority;

[0029] The residual error signal of the upper layer compensation serves as the error correction reference for the middle layer compensation, and the DC drift output of the middle layer compensation serves as the gain fine-tuning reference for the lower layer compensation. When the top-level emergency compensation is triggered, an interrupt signal is generated, which freezes the parameter updates of the other three layers at the current moment. After the freeze is lifted, the compensation parameters of the lower, middle, and upper layers smoothly transition to the target value in a step-by-step approximation manner.

[0030] Furthermore, in step three,

[0031] The baseline calibration is achieved in the following manner:

[0032] 3.1 Extract channel state information from the receiving end of the communication signal, wherein the channel state information includes the amplitude response, phase response and nonlinear distortion of the optical fiber link;

[0033] 3.2 Calculate the total end-to-end attenuation of the entire link based on the reference power at the transmitting end and the measured power at the receiving end. The calculation formula is as follows:

[0034] in, The reference power for the transmitter. The measured power at the receiving end is used as the global reference baseline for sensor signal compensation.

[0035] 3.3 When the overall amplitude of the sensing signal deviates from the global baseline by more than the preset deviation range, the deviation amount is used as a compensation control signal.

[0036] Furthermore, in step three, the degree of nonlinear distortion extracted from the channel state information is used to limit the maximum output power of the optical amplifier during the compensation process;

[0037] When the degree of nonlinear distortion exceeds the preset nonlinear threshold, the upper limit of the gain of the optical amplifier is temporarily reduced.

[0038] Furthermore, in step three,

[0039] Predictive adjustments are achieved in the following ways:

[0040] Environmental physical parameters are extracted from the receiving end of the sensor signal. When the temperature or vibration parameter change of the optical fiber segment exceeds the preset range, the additional attenuation estimate of the segment is calculated according to the pre-calibrated environmental attenuation coefficient. This estimate is used as the feedforward control quantity to adjust the transmission power or equalization parameters of the communication signal in advance.

[0041] Furthermore, in step four, the closed-loop adjustment is implemented as follows:

[0042] The difference between the actual compensation gain and the expected compensation gain of each fiber segment is taken as the deviation, and the deviation is attributed in the following order:

[0043] First, if the characteristics of the compensated residual signal do not match the typical characteristics of the attenuation type diagnosed in step one, it is determined that the attenuation type diagnosis is incorrect, and the corresponding diagnosis threshold is adjusted; if the characteristics of the residual signal match the typical characteristics of the diagnosed attenuation type, it is further determined whether the correlation coefficient between the compensated residual attenuation waveform and the typical compensation curve corresponding to the current strategy is lower than the preset matching threshold. If so, it is determined that the compensation strategy selection is inappropriate, and the compensation strategy is re-matched from the strategy library.

[0044] If the correlation coefficient is not lower than the preset matching threshold, the deviation is attributed to the deviation of the compensation parameter execution, and the compensation parameter is adjusted in reverse within the allowable adjustable range.

[0045] The data from each compensation operation is stored in a knowledge base for continuous optimization of the initial parameters of the compensation strategy.

[0046] Furthermore, it also includes a dynamic power allocation step, which runs in parallel with steps one through four, as detailed below:

[0047] Under the premise of not exceeding the maximum allowable output power of the optical transmitter and the total transmit power limit defined by the fiber nonlinear effect threshold in the composite network, a joint inductive-synthetic efficiency function is established. ,in For communication quality indicators, As a sensor accuracy indicator, and These are the communication weighting coefficient and the sensing weighting coefficient, respectively. Dynamically adjust according to the real-time operating conditions of the highway. and The values ​​of are determined according to the linear mapping rules. , The power distribution ratio of communication signals and sensor signals is adjusted synchronously, among which... This represents the upper limit of the total transmission power.

[0048] A signal attenuation compensation system for a 5G and fiber optic sensing composite network on a highway, the system being used to implement the above method, specifically including:

[0049] The acquisition unit is used to acquire in real time the first state parameters of communication signals and the second state parameters of sensing signals in the composite network, as well as the environmental physical parameters along the optical fiber.

[0050] The diagnostic unit is used to receive the data output by the acquisition unit and perform correlation analysis to perform online diagnosis of signal attenuation sources along the optical fiber and output attenuation type labels for each optical fiber segment.

[0051] The strategy generation unit is used to receive the attenuation type label output by the diagnostic unit, and accordingly match the corresponding compensation strategy from the preset differential compensation strategy library to generate a multi-time scale compensation scheme containing compensation methods and compensation parameters.

[0052] The compensation execution unit is used to receive the compensation scheme output by the strategy generation unit and execute it. During the execution process, the environmental physical parameters extracted from the communication signal are predictively adjusted based on the compensation reference sensor signal, and the channel state information extracted from the communication signal is baseline calibrated based on the compensation reference sensor signal.

[0053] The feedback optimization unit is used to receive the compensated signal output by the compensation execution unit and evaluate the compensation effect, calculate the compensation gain, adjust the diagnostic threshold in the diagnostic unit and the compensation parameters in the strategy generation unit in a closed loop according to the comparison result of the compensation gain and the preset threshold, and send the adjusted parameters back to the diagnostic unit and the strategy generation unit respectively.

[0054] The power allocation unit is used to establish a joint communication and sensing performance function, dynamically adjust the values ​​of communication weight coefficient and sensing weight coefficient according to the real-time operating conditions of the highway, and synchronously adjust the transmission power allocation ratio of communication signal and sensing signal according to the linear mapping rule.

[0055] A computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the method as claimed in any one of claims 1 to 8.

[0056] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0057] This invention establishes an online attenuation source diagnosis mechanism, which can classify signal attenuation in fiber optic links into four types: intrinsic attenuation, temperature drift attenuation, vibration stress attenuation, and sudden attenuation, and match differentiated compensation strategies based on the diagnosis results. For situations where multiple attenuation types exist simultaneously in the same segment, a priority rule is established to resolve decision-making conflicts. This mechanism overcomes the shortcomings of existing technologies that use a uniform compensation logic regardless of the attenuation source, making the correspondence between compensation measures and attenuation causes more reasonable and avoiding the blindness of compensation methods.

[0058] Furthermore, a mutually reinforcing relationship was established between communication signals and sensing signals. Specifically, the channel state information of the communication signal was processed to extract the total end-to-end attenuation of the entire link, providing a global reference baseline for the compensation of the sensing signal; the environmental physical parameters sensed by the sensing signal were calibrated to provide feedforward control quantities for the communication signal, thereby achieving predictive compensation. The compensation of the two types of signals is no longer performed independently, but rather incorporates the information of the other into its own compensation decision. This solves the problem mentioned in the background technology that communication compensation and sensing compensation are independent of each other and fail to utilize complementary information.

[0059] Finally, this invention employs a multi-timescale hierarchical design in its compensation architecture. The four compensation layers correspond to four attenuation types and operate in parallel with different cycles at the minute, second, millisecond, and microsecond levels. Information transmission and conflict avoidance are achieved through cascading collaboration between the layers. Simultaneously, through quantitative evaluation of compensation gain and hierarchical attribution of deviations, closed-loop adjustment of diagnostic thresholds, strategy matching, and compensation parameters is realized, enabling the entire compensation method to continuously adapt to environmental changes. Furthermore, the operating condition-driven dynamic power allocation mechanism allows the compensation strategy to automatically adapt to actual road conditions such as peak traffic periods, nighttime off-peak periods, and extreme weather, further enhancing the engineering applicability of the solution in highway scenarios. Attached Figure Description

[0060] Figure 1 This is a flowchart of the overall method of the present invention;

[0061] Figure 2 This is a schematic diagram of the overall system architecture of the present invention;

[0062] Figure 3 This is a flowchart of the online diagnosis of attenuation sources in step one of the present invention;

[0063] Figure 4 This is a structural diagram of the multi-timescale compensation scheme in step two of the present invention;

[0064] Figure 5 This is a flowchart of the reciprocal enhancement compensation execution process for step three of the present invention;

[0065] Figure 6 This is the overall closed-loop control flowchart of the present invention; Detailed Implementation

[0066] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and are not intended to limit the scope of the invention.

[0067] like Figures 1 to 6 As shown:

[0068] I. Example 1:

[0069] This embodiment provides a method for signal attenuation compensation in a 5G and fiber optic sensing composite network for highways. This method is applied to 5G fronthaul networks deployed along highways and... Scenarios of shared fiber transmission in distributed optical fiber sensing systems.

[0070] Before implementation, the system must be configured.

[0071] More specifically, fiber optic segments to be monitored are selected along the highway; these segments simultaneously carry 5G fronthaul communication signals and... Distributed sensing signals. The operating wavelength of the communication signal is selected in the 1550nm band. The detection optical pulses of the sensing signal are selected from different wavelengths within the 1550nm band, and the two are combined into the same optical fiber through a wavelength division multiplexer.

[0072] Furthermore, for receiving communication signals, an optical power detection module and a baseband processing unit are installed at the RRU end of the 5G base station. The optical power detection module preferably uses an InGaAs photodetector with a response wavelength range covering 1520nm to 1580nm. It converts the received optical signal into a current signal, which is then read by the edge computing node after analog-to-digital conversion. The baseband processing unit extracts the bit error rate, signal-to-noise ratio, and dispersion cost from the digital domain of the received signal. The edge computing node reads these parameters through a data interface and records the corresponding timestamps and fiber location information.

[0073] Regarding the reception of sensor signals, in The receiving end of the system is equipped with a coherent receiving module, preferably consisting of a local oscillating laser, a 90-degree optical mixer, and a balanced photodetector.

[0074] The probe light pulse is frequency-shifted by an acousto-optic modulator and injected into the sensing fiber. The backscattered Rayleigh light returns and is coherently mixed with the local oscillator light. The mixed optical signal is converted into an electrical signal by a balanced photodetector. Edge computing nodes sample the output of the balanced detector using a high-speed analog-to-digital converter. The sampled data is then digitally orthogonally demodulated, allowing for the simultaneous extraction of intensity, phase, polarization state, and interference fading depth information of the backscattered light.

[0075] The digital orthogonal demodulation is used for coherent detection. Conventional signal processing techniques in the system are used to extract the in-phase and quadrature components of the backscattered light from the beat frequency signal.

[0076] As a preferred approach, for the acquisition of environmental physical parameters, sensing nodes are set up at regular intervals along the optical fiber, equipped with temperature sensors, vibration sensors, and strain sensors. The output signals of each sensor are aggregated to the edge computing node via a fieldbus, and the edge computing node is deployed in the 5G base station equipment room along the highway. The acquisition of data from each channel is synchronized in time with the system clock of the edge computing node.

[0077] Furthermore, regarding the division of fiber optic segments, the deployment location of 5G base stations along the highway is used as a reference point, and the fiber optic segment between two adjacent base stations is considered as an independent segment.

[0078] More specifically, if the distance between adjacent base stations is too long, a dividing point can be added in the middle to ensure that the length of each segment does not exceed a preset length threshold.

[0079] As a preferred approach, this threshold can be set to 1 kilometer, with the specific value adjustable based on actual fiber quality and system requirements. After segmentation, a unique segment number is assigned to each segment, and a mapping relationship between the segment number and the mileage marker is established. This mapping relationship is stored in the configuration file of the edge computing node for use during attenuation location diagnosis.

[0080] The following are the specific implementation steps of the signal attenuation compensation method for the highway 5G and fiber optic sensing composite network of the present invention:

[0081] Step 1: Data acquisition and online diagnosis of attenuation sources.

[0082] The edge computing node reads three data streams in real time at a preset sampling period, as follows:

[0083] The first path consists of the first state parameters of the communication signal, including received optical power, bit error rate, signal-to-noise ratio, and dispersion cost;

[0084] The second path contains the second state parameters of the sensing signal, including backscattered light intensity, phase noise, polarization state, and interference fading depth.

[0085] The third path includes environmental physical parameters, such as temperature distribution, vibration spectrum, and strain distribution along the optical fiber.

[0086] More specifically, after the edge computing node acquires the above three data streams, it performs correlation analysis to diagnose the attenuation type of each fiber segment.

[0087] In this embodiment, the attenuation types include four types: intrinsic attenuation, temperature drift attenuation, vibration stress attenuation, and sudden attenuation. The specific details of the four types are as follows:

[0088] Intrinsic attenuation refers to the decrease in signal amplitude caused by the inherent losses of the optical fiber itself, including absorption loss and Rayleigh scattering loss of the fiber material. This type of attenuation is essentially determined after the optical fiber is manufactured and changes extremely slowly over time, usually without significant changes due to short-term fluctuations in environmental factors.

[0089] The intrinsic attenuation is determined by comparing the communication signal amplitude attenuation curve with the backscattered light intensity distribution curve for each fiber segment at the edge computing node, and calculating the Pearson correlation coefficient between the two.

[0090] More specifically, suppose there are n sampling points in a certain optical fiber segment, and the communication signal amplitude attenuation value sequence is as follows:

[0091] , ,..., ,

[0092] The sequence of backscattered light intensity values ​​is as follows:

[0093] , ,…, ,

[0094] The means of the two are as follows:

[0095]

[0096] ,

[0097] Based on the above calculations, the Pearson correlation coefficient r is calculated using the following formula:

[0098]

[0099] When the calculated correlation coefficient r is greater than a first preset threshold, and the attenuation fluctuation of the segment is less than a preset gradual change threshold within multiple consecutive monitoring periods, the edge computing node determines that the segment has intrinsic attenuation. A preferred example value for the gradual change threshold is an attenuation change rate of less than 0.01 dB / hour.

[0100] More specifically, the specific value of the first preset threshold can be set according to the fiber type and system configuration. In a preferred example, the first preset threshold is 0.85. This value is derived from the typical correlation between the theoretical loss spectrum and the actual loss spectrum of a standard single-mode fiber in the 1550nm window. This threshold is subject to closed-loop dynamic adjustment in subsequent steps.

[0101] Temperature drift attenuation refers to the additional loss in optical fibers caused by changes in ambient temperature. Its physical mechanisms include: thermal expansion or contraction of the fiber material due to temperature changes leading to microbending loss; changes in the refractive index distribution of the fiber caused by temperature changes; and the difference in the coefficients of thermal expansion between the fiber and the cable sheath materials, resulting in additional stress on the fiber. The rate of this type of attenuation is positively correlated with the rate of temperature change; both increases and decreases in temperature can lead to increased loss, and the faster the temperature change, the more significant the attenuation. Once the temperature stabilizes, the attenuation tends to stabilize as well.

[0102] The method for determining temperature drift attenuation is as follows: Edge computing nodes extract the rate of temperature change and the rate of change of backscattered light intensity within each fiber segment and compare them segment by segment. For each segment, a time-domain cross-correlation analysis is performed between the average rate of temperature change of the temperature sensor within that segment and the average rate of change of backscattered light intensity within that segment. When the maximum value of the cross-correlation function is greater than a second preset threshold, and the corresponding lag time is within a preset range, it is determined that temperature drift attenuation exists in that segment.

[0103] More specifically, the edge computing node takes a time window preceding the current moment and collects discrete sampling sequences within the window.

[0104] As a preferred approach, the time window can be set to 60 seconds. Let there be N sampling points within the window, and the temperature change rate sequence be... The sequence of received optical power change rate is k=1,2,...,N.

[0105] Edge computing nodes calculate the cross-correlation function between the two:

[0106]

[0107] Where m is the number of lag sampling points.

[0108] When the maximum value of the cross-correlation function Greater than the second preset threshold, and the corresponding lag time When within the preset range, where The sampling interval is used to determine the presence of temperature drift attenuation in edge computing nodes.

[0109] A preferred example value for the second preset threshold is 0.7, which is derived from the standard correlation coefficient between the decay change and the temperature change in a temperature cycling experiment under laboratory conditions.

[0110] Vibration stress attenuation refers to the changes in fiber microbending loss or polarization-dependent loss caused by external vibrations or stresses such as traffic loads and geological movements. Its physical mechanisms include vibration causing microbending of the fiber, leading to radiation loss, and vibration-induced stress changes altering the fiber's birefringence, thus causing polarization-dependent loss. This type of attenuation changes rapidly and is directly related to the vibration intensity; as the vibration intensity increases, the attenuation fluctuation amplitude increases, and the attenuation usually recovers after the vibration disappears.

[0111] Furthermore, the method for determining vibration stress attenuation is as follows: the edge computing node compares the vibration spectrum intensity in the environmental physical parameters with the temporal fluctuation amplitude of the backscattered light intensity in the second state parameter.

[0112] Specifically, the edge computing nodes perform fast Fourier transform on the vibration signal and the light intensity signal respectively, and calculate the coherence function of the two in the main vibration frequency band;

[0113] It should be noted that before calculating the coherence function, the edge computing nodes first evaluate the data quality of the intensity data of each fiber segment based on the interference fading depth extracted in step one.

[0114] If the interference fading depth of a certain fiber segment exceeds the preset fading threshold, the intensity data of that segment will be marked as unreliable data and will not be included in subsequent coherence analysis.

[0115] If the proportion of unreliable data segments across the entire line exceeds a preset threshold, the determination of vibration stress attenuation will be paused, and a system prompt message will be output.

[0116] A preferred example of the fading threshold is a signal amplitude decrease of more than 50%, and a preferred example of the proportion threshold is 30%.

[0117] Let the Fourier transform of the vibration signal be... The Fourier transform of the light intensity signal is The cross-power spectral density of the two is Their respective automatic rate spectral densities are and The coherence function is then calculated using the following formula:

[0118]

[0119] More specifically, when the average value of the coherence function within a preset frequency band When the value exceeds the third preset threshold, vibration stress attenuation is determined to exist. A preferred example value for the third preset threshold is 0.6, which is set based on the statistical results of field-measured vibration-attenuation coherence under typical traffic loads on highways.

[0120] Sudden attenuation refers to a sudden change in signal amplitude caused by sudden events such as fiber bending, breakage, or connector loosening. This type of attenuation is characterized by its transient nature, with a large jump occurring within a very short time, which is fundamentally different from the three types of continuously changing attenuation mentioned above.

[0121] Furthermore, the method for determining sudden attenuation is as follows: if none of the above three attenuation types are triggered, but the received optical power of the signal undergoes a sudden change exceeding a fourth preset threshold within a preset time window, then the edge computing node determines that sudden attenuation exists. A preferred example value for the fourth preset threshold is: the average signal power decreases by more than 3dB within a 100-millisecond window. This value is set based on typical engineering criteria for sudden loss at fiber optic splices or bends.

[0122] As a preferred approach, when the same fiber segment simultaneously meets the criteria for multiple attenuation types, the edge computing node determines the primary attenuation type according to the priority order of sudden attenuation, vibration stress attenuation, temperature drift attenuation, and intrinsic attenuation.

[0123] More specifically, the priority order is set based on the time-varying characteristics of various types of decay and their impact on system security.

[0124] Sudden decay is given the highest priority due to its transient nature and the greatest threat to system security; once triggered, the corresponding emergency compensation is executed first.

[0125] Vibration stress decay is given the second highest priority because its time-varying rate is faster than that of temperature drift decay, requiring a rapid response to avoid continuous deterioration of signal quality.

[0126] Temperature drift decays at a slower rate than vibration stress decays, and can be tracked and compensated on a slightly larger time scale, so it is given less priority.

[0127] The intrinsic decay changes the slowest and can be compensated for as the underlying baseline, so it has the lowest priority.

[0128] Diagnostic results for non-priority attenuation types are not discarded but cached and used as additional corrections to compensation parameters. That is, when executing the compensation strategy corresponding to the primary type, additional corrections are applied to the compensation amount based on the severity of the secondary attenuation type.

[0129] Through the above diagnostic process, the edge computing node outputs the attenuation type label and the corresponding attenuation severity index for each fiber segment.

[0130] The severity index of attenuation is calculated as follows:

[0131] For intrinsic attenuation, the degree of deviation between the measured loss value and the theoretical loss value is used as an index. Let the measured loss value be... The theoretical loss value is The intrinsic attenuation severity index for:

[0132]

[0133] In this formula, when the measured loss value equals the theoretical loss value, that is... = hour, =0; when the measured loss value reaches twice the theoretical loss value, that is =2 hour, =1; take 0 when the measured loss value is less than the theoretical loss value, and take 1 when it is more than twice the theoretical loss value. The value between 0 and 1 is determined by linear interpolation.

[0134] For temperature drift attenuation, the ratio of the temperature change to the allowable temperature range of the optical fiber is used as the exponent. Let the current temperature change be... The allowable temperature variation range of optical fiber is The severity index of temperature drift decay for:

[0135]

[0136] More specifically, the index is 0 when the temperature does not change, 1 when the temperature changes to the upper limit of the allowable range, and the intermediate value is determined by linear interpolation, taking 1 when it exceeds the allowable range.

[0137] For vibration stress attenuation, the ratio of the attenuation fluctuation amplitude caused by vibration to the signal dynamic range is used as the exponent. Let the attenuation fluctuation amplitude caused by the current vibration be... The dynamic range of the signal is The severity index of vibration stress attenuation for:

[0138]

[0139] The index is 0 when there is no vibration, 1 when the fluctuation amplitude reaches the upper limit of the signal's dynamic range, the intermediate value is determined by linear interpolation, and 1 is taken when it exceeds the dynamic range.

[0140] For sudden decay, the ratio of the mutation magnitude to the mutation trigger threshold is used as the index.

[0141] More specifically, let the current mutation magnitude be... The fourth preset threshold is The severity index of sudden decay for:

[0142]

[0143] The exponent is 1 when the mutation magnitude equals the trigger threshold, and the exponent is greater than 1 when the mutation magnitude exceeds the trigger threshold.

[0144] 2. Step Two:

[0145] Differentiated compensation strategy matching and multi-timescale compensation scheme generation.

[0146] After receiving the attenuation type label and attenuation severity index from step one, the edge computing node performs policy matching and scheme generation.

[0147] More specifically, edge computing nodes are pre-configured with a differentiated compensation strategy library. This strategy library is a compensation strategy mapping table that stores the correspondence between attenuation type labels and compensation strategy entries.

[0148] Furthermore, each compensation strategy entry contains the following fields:

[0149] The compensation method identifier, the initial value and adjustable range of the compensation parameters, the time scale for compensation execution, and the applicable fiber optic segment length range.

[0150] This strategy library is pre-built based on fiber type and system design specifications before system deployment and stored in the memory of edge computing nodes.

[0151] More specifically, the strategy library is constructed based on the physical mechanisms of various attenuations and corresponding engineering compensation experience.

[0152] Intrinsic attenuation is a broadband flat loss, which is suitable for broadband gain compensation using optical amplifiers.

[0153] Temperature drift decay is a slow-varying type of loss, which is suitable for compensation using bias correction and slow-varying equalization.

[0154] Vibration stress attenuation is a type of fast-deformation loss, which is suitable for rapid tracking and compensation using adaptive filtering.

[0155] Sudden attenuation is a transient loss and is suitable for handling using transient response compensation and redundancy switching.

[0156] The specific execution method of strategy matching is as follows: the edge computing node uses the attenuation type label output in step one as an index to perform a table lookup in the compensation strategy library and extract the corresponding compensation strategy entry.

[0157] Furthermore, taking intrinsic attenuation as an example, the compensation strategy entries obtained by looking up the compensation strategy mapping table include:

[0158] The compensation method involves adjusting the gain coefficient of the tunable amplifier, with the compensation parameter being the gain value G, and the initial value being the theoretical loss value of the fiber segment. Plus the preset redundancy Adjustable range is The specific range depends on the model of the optical amplifier.

[0159] The edge computing nodes then use the attenuation severity index output from step one. The initial execution value of the compensation parameter is determined within the adjustable range.

[0160] The determination method uses linear interpolation:

[0161]

[0162] When the severity index The lower limit of the adjustable range is taken at that time. ,when The upper limit of the adjustable range is taken at this time. The intermediate value is determined by a linear ratio.

[0163] As a preferred option, the multi-timescale compensation scheme generated in this step includes four compensation levels. The settings of each level correspond one-to-one with the four attenuation types in step one, and the time scale of each level is determined according to the time-varying characteristics of the corresponding attenuation type.

[0164] The bottom-level baseline compensation targets intrinsic attenuation and uses the gain coefficient adjustment of the tunable amplifier as a compensation method. Since intrinsic attenuation changes extremely slowly, the adjustment period for the compensation parameters at this level is set to the minute level, specifically based on the minute-by-minute update value of the moving average of the intrinsic attenuation severity diagnosed in step one.

[0165] For mid-level tracking compensation targeting temperature drift attenuation, dynamic bias correction of the photodetector and updating of the tap coefficients of the slow-changing equalizer are used as compensation methods. Since the time scale of temperature changes is typically on the order of seconds to minutes, the adjustment cycle of the compensation parameters at this level is set to the order of seconds, and the specific adjustment is based on the update value of the temperature change rate per second.

[0166] The upper-level real-time compensation targets vibration stress attenuation and employs adaptive filter weight updates as a compensation method. Since the time scale of the vibration signal is on the millisecond level, the adjustment period of the compensation parameters at this level is set to the millisecond level, specifically based on the millisecond update value of the vibration spectral intensity.

[0167] The top-level emergency compensation system targets sudden signal attenuation, employing transient response compensation and rapid switching of redundant links as compensation methods. Since sudden attenuation is a transient event, the response time at this level is set to the microsecond level, and the triggering condition is that the sudden change in signal attenuation exceeds the fourth preset threshold.

[0168] The above four levels of compensation operate in parallel.

[0169] The coordination rules between different levels are as follows:

[0170] The residual error signal of the upper-layer adaptive filter is fed to the error evaluation end of the middle-layer equalizer to correct the update direction of the equalizer tap coefficients.

[0171] More specifically, when the residual error of the adaptive filter contains a specific frequency component, this information prompts the mid-level equalizer to adjust the tap coefficients of the corresponding frequency band.

[0172] The DC drift output of the middle equalizer is fed to the gain fine-tuning terminal of the bottom optical amplifier to compensate for the slow-varying correction of the gain coefficient.

[0173] More specifically, when there is a DC offset in the equalizer output, the offset indicates that the gain of the optical amplifier should be adjusted accordingly.

[0174] The top-level emergency compensation has the highest execution priority. When it is triggered, an interrupt signal is generated. This interrupt signal freezes the updated values ​​of the compensation parameters of the bottom, middle and top layers at the current moment. The freeze is lifted after the emergency compensation is completed and the confirmation signal is restored. The parameters of each layer continue to be updated based on the values ​​before the freeze.

[0175] It should be noted that after the freeze is lifted, the compensation parameters of the bottom, middle and top layers do not jump directly to the values ​​saved before the freeze, but instead use a gradual approximation method for a smooth transition.

[0176] Specifically, the difference between the parameter value saved before freezing and the target value calculated based on the actual signal quality at the current time is used as the deviation. This deviation is gradually eliminated by first-order inertial filtering according to a preset time constant. A preferred example value of the time constant is 200 milliseconds. Alternatively, a limiting and gradual change method is used, limiting the maximum parameter change per millisecond to no more than 0.5% of the adjustable range of the parameter, until the parameter smoothly transitions to the target value.

[0177] Through the aforementioned cascading and collaborative relationships, compensation at each level is coordinated in an orderly manner, avoiding conflicts or duplication among compensation methods.

[0178] 3. Step Three:

[0179] Reciprocity enhances the implementation of compensation.

[0180] The edge computing node parses the compensation scheme generated in step two into control commands that can be recognized by each execution device and sends them to the corresponding execution device.

[0181] Furthermore, for sensor signal compensation, the edge computing nodes employ baseline calibration. The core idea of ​​baseline calibration is to utilize the link state information naturally carried by the communication signal during transmission in the optical fiber as a reference benchmark for sensor signal compensation, rather than adding additional detection equipment.

[0182] The specific implementation process of baseline calibration is as follows:

[0183] Edge computing nodes extract channel state information from the receiving end of communication signals. This information includes the amplitude response, phase response, and degree of nonlinear distortion of the fiber optic link.

[0184] Furthermore, the edge computing node obtains the reference power of the communication signal transmitter. and the measured power at the receiving end The total end-to-end attenuation of the entire link is calculated using the following formula. :

[0185] This total attenuation reflects the overall loss experienced by the communication signal during transmission throughout the entire fiber optic link. Since the communication signal is transmitted point-to-point, this end-to-end total attenuation reflects the cumulative loss of the entire link, rather than the point-to-point loss along the route.

[0186] More specifically, this total attenuation It serves as a global reference baseline for sensor signal compensation.

[0187] As a preferred approach, when the edge computing node detects that the total attenuation of the communication signal is equal to... When the deviation between the average loss calculated by the system and the total loss of the entire link exceeds a preset deviation range, it is determined that there is a local abnormal attenuation event in the current optical fiber link. At this time, the use of the total attenuation of the communication signal as the global reference baseline for the sensing signal is suspended, and the system switches to using... The system's own end-to-end average loss is used as a replacement baseline. Once the deviation between the two falls back to within the preset deviation range, the total attenuation of the communication signal is used as the global reference baseline again.

[0188] A preferred example value for this preset deviation range is 3dB.

[0189] Once the deviation between the two falls back to the preset deviation range, the total attenuation of the communication signal is restored as the global reference baseline. During restoration, the value of the communication baseline is not directly switched to the value of the alternative baseline. Instead, the difference between the alternative baseline and the communication baseline is gradually eliminated by first-order inertial filtering according to the preset time constant. A preferred example value of the time constant is 500 milliseconds to avoid transient disturbances introduced by the baseline switching.

[0190] At the sensor signal receiving end, the edge computing node will The average level of the overall loss distribution across the entire link, measured by the system, is compared with this global reference baseline. When the overall amplitude level of the sensing signal deviates from this global baseline by more than a preset deviation range, the edge computing node identifies this deviation as a systematic attenuation offset that needs to be compensated, and drives... The digital signal processing module at the receiving end corrects the global gain of the sensor signal.

[0191] It should be noted that the point-to-point distributed loss localization capability of the sensor signal is still provided by The system's own reflection and scattering mechanism provides point-by-point loss detection for each fiber segment by measuring the distribution of backscattered Rayleigh light along the fiber length. The global baseline provided by the communication signal is used to calibrate the overall amplitude level of the sensing signal, rather than replacing the point-by-point positioning function of the sensing signal. The two types of signals each perform their respective functions and cooperate with each other: the communication signal provides a globally accurate loss reference, while the sensing signal provides distributed positioning capability.

[0192] Meanwhile, the degree of nonlinear distortion extracted from the channel state information by the edge computing node is used to limit the maximum output power of the optical amplifier during the compensation process.

[0193] Specifically, when the detected nonlinear distortion exceeds a preset nonlinear threshold, the upper limit of the optical amplifier's gain is reduced. The gain is temporarily reduced to avoid triggering nonlinear effects such as stimulated Brillouin scattering due to excessive compensation gain, ensuring that the compensation operation does not introduce new signal degradation. Once the nonlinear distortion level falls below the threshold, the original upper limit of the optical amplifier's gain is restored.

[0194] Furthermore, for communication signal compensation, edge computing nodes employ a predictive adjustment method. The core idea of ​​predictive adjustment is to utilize the sensing capabilities of sensor signals to perceive the physical parameters of the environment along the optical fiber, predict the attenuation that the communication signal will experience, and compensate in advance, rather than reacting after attenuation occurs.

[0195] As a preferred approach, the specific implementation process of predictive adjustment is as follows:

[0196] Edge computing nodes extract environmental physical parameters from the receiver of sensor signals, including temperature distribution along the optical fiber. Vibration spectrum and strain distribution .

[0197] Furthermore, taking the additional attenuation caused by temperature as an example, the edge computing node monitors the temperature change rate of each fiber segment. When the rate of temperature rise in a fiber segment exceeds a preset threshold, it means that the loss in that segment will increase accordingly. Edge computing nodes then adjust the temperature-attenuation coefficient based on a pre-defined temperature-attenuation factor. Calculate the estimated additional attenuation expected to occur in this section:

[0198]

[0199] in, This represents the temperature change in that section. This is the length of the segment.

[0200] Using this estimated value as the feedforward control variable, the edge computing node sends a control command to the optical amplifier on the communication signal link in advance, increasing its gain value based on the current value. This is to offset the additional attenuation that is about to occur.

[0201] If multiple segments show estimated attenuation values ​​simultaneously, the estimated attenuation values ​​of each segment are superimposed, and the transmission power of the communication signal is adjusted globally.

[0202] Furthermore, taking the additional attenuation caused by vibration as an example, the edge computing node monitors the vibration intensity of each fiber segment. When the vibration intensity increase of a certain fiber segment exceeds a preset increase threshold, the vibration-attenuation coefficient is pre-calibrated. Calculate the estimated additional attenuation for this section:

[0203]

[0204] Similarly, the gain of the optical amplifier or the output power of the transmitter for the communication signal are adjusted in advance using a feedforward method.

[0205] Temperature-decrease coefficient and vibration-damping coefficient The pre-calibration was completed through laboratory testing before system deployment.

[0206] More specifically, the calibration method is as follows: Take optical fiber samples of the same model and batch as those used on-site, measure the transmission loss of the optical fiber at different temperature gradients in a temperature-controlled chamber, record the corresponding data of temperature and loss, and obtain the temperature-attenuation coefficient through linear regression fitting. ;

[0207] The transmission loss of the optical fiber was measured on a vibration table at different vibration frequencies and amplitudes. The corresponding data of vibration parameters and loss were recorded, and the vibration-attenuation coefficient was obtained by linear regression fitting. .

[0208] After calibration, the coefficients are stored in the configuration file of the edge computing node for runtime use.

[0209] In addition, edge computing nodes dynamically adjust the prediction step size of the communication signal equalizer according to the rate of change of environmental physical parameters. The prediction step size of the equalizer determines how quickly the equalizer coefficients track channel changes. A step size that is too large will introduce noise, while a step size that is too small will result in slow tracking speed.

[0210] Specifically, when the rate of temperature change or rate of change of vibration intensity When the value is large, the prediction step size The corresponding increase allows the equalizer to converge to a near-optimal coefficient state before actual attenuation occurs; when environmental changes are gradual, the prediction step size is increased. The corresponding reduction is made to decrease the steady-state error.

[0211] Through the aforementioned mutually reinforcing mechanism, the compensation of the two types of signals forms a two-way positive feedback relationship: the communication signal provides a global reference baseline for the link state of the sensing signal, and the sensing signal provides prior information about environmental changes for the communication signal. The compensation effects of the two are mutually reinforcing rather than independent of each other.

[0212] After the compensation is completed, the edge computing node records the execution results as data for this compensation operation, including the actual compensation parameters, execution time, and preliminary indicators of signal quality after compensation.

[0213] 4. Step Four:

[0214] Compensation effect evaluation and closed-loop adjustment.

[0215] After the compensation operation is completed, the edge computing node re-executes data acquisition, reads the received optical power, bit error rate, and signal-to-noise ratio of the communication signal, and the backscattered light intensity and signal-to-noise ratio of the sensor signal, and forms the compensated dataset.

[0216] Edge computing nodes calculate the compensation gain for communication signals and the compensation gain for sensing signals, respectively.

[0217] More specifically, the compensation gain is defined as the ratio of the signal-to-noise ratio after compensation to the signal-to-noise ratio before compensation, expressed in decibels.

[0218] Furthermore, let the signal-to-noise ratio before compensation be... The signal-to-noise ratio after compensation is Then compensate for the gain. for:

[0219]

[0220] A positive compensation gain indicates improved signal quality, while a negative compensation gain indicates that the compensation operation actually degrades the signal quality.

[0221] If the communication signal compensation gain of a certain optical fiber segment Below the preset gain threshold Or sensor signal compensation gain Below the preset gain threshold Then the edge computing node will initiate compensation deviation analysis.

[0222] The gain threshold can be set according to the system design specifications.

[0223] As a preferred method, the gain threshold of the communication signal Set to 3dB, this value corresponds to the engineering reference criterion for halving optical power and is a commonly used reference value in fiber optic communication system design; the gain threshold of the sensing signal. The value is set to 2dB, which corresponds to the minimum acceptable improvement in the signal-to-noise ratio of the sensing system. A value lower than this means that the actual benefit of the compensation operation is not significant.

[0224] It should be noted that the above values ​​are preferred examples. In actual engineering deployments, adjustments can be made according to the specific design specifications of the system. For example, if the system adopts a higher-order modulation format, which has higher requirements for signal-to-noise ratio, the gain threshold can be increased accordingly; if the system adopts stronger forward error correction coding, the gain threshold can be appropriately reduced.

[0225] More specifically, the specific process of compensation deviation analysis is as follows:

[0226] Edge computing nodes will compensate for the actual gain of each fiber segment. Compared with expected compensation gain The difference between them is taken as the deviation:

[0227]

[0228] Expected compensation gain The compensation scheme generated in step two and the pre-stored compensation effect model are used for estimation. This model is continuously updated during system operation.

[0229] As a preferred approach, edge computing nodes perform attribution judgments in the following order:

[0230] First-level attribution: Attenuation type diagnosis error. The edge computing node compares the attenuation type label diagnosed in step one with the characteristics of the compensated residual signal. The residual signal refers to the difference between the compensated signal and the ideal signal.

[0231] Furthermore, if the characteristics of the residual signal do not match the typical characteristics of the diagnosed attenuation type, for example, in a segment determined to be intrinsic attenuation, the compensated residual signal exhibits a fluctuation pattern different from the spectral characteristics of that attenuation type, then the edge computing node will attribute the deviation to an attenuation type diagnosis error.

[0232] At this point, adjust the corresponding diagnostic threshold: if a certain type of attenuation actually exists but is not triggered for judgment, then lower the corresponding threshold to increase sensitivity;

[0233] If a certain type of decay does not actually exist but is triggered incorrectly, the threshold is increased to improve specificity.

[0234] If the characteristics of the residual signal match the typical characteristics of the diagnosed attenuation type, the diagnosis is correct, and the second-level attribution process continues.

[0235] More specifically, the threshold is adjusted as follows:

[0236] Let the current threshold be Adjust step size Can be an initial value If it is 5%, then the adjusted threshold is... for:

[0237]

[0238] The upward adjustment is indicated by a plus sign, which increases specificity; the downward adjustment is indicated by a minus sign, which increases sensitivity.

[0239] As a preferred approach, the adjustment range in a single instance shall not exceed 20% of the initial value, that is:

[0240]

[0241] This is to avoid system oscillations caused by excessively large adjustments in a single instance.

[0242] Second-level attribution: Inappropriate selection of compensation strategy. If the attenuation type diagnosis is correct, the edge computing node will perform correlation analysis between the compensated residual attenuation waveform and the typical compensation curves of each compensation method in the strategy library.

[0243] Furthermore, a typical compensation curve refers to the attenuation characteristic curve that each compensation method can offset under ideal conditions.

[0244] For example, the typical compensation curve of an optical amplifier is a flat gain type, where the gain value does not change with frequency within the compensation band; the typical compensation curve of an adaptive filter is a frequency-selective gain type, providing different gain values ​​at different frequencies.

[0245] More specifically, if the correlation coefficient between the residual attenuation waveform and the typical compensation curve corresponding to the current strategy is lower than the preset matching threshold, the edge computing node will attribute the deviation to an improper selection of the compensation strategy. In this case, a suboptimal matching strategy will be selected from the strategy library to re-execute the compensation.

[0246] Third-level attribution: Compensation parameter execution deviation.

[0247] If the strategy selection is correct, the edge computing node attributes the deviation to the execution deviation of the compensation parameters. At this point, the edge computing node adjusts the compensation parameters in the opposite direction of the deviation, within the adjustable range allowed by the strategy entry.

[0248] If the actual compensation is insufficient, i.e. If so, adjust the parameters in the direction of increasing the compensation amount;

[0249] If the actual compensation amount is too large, that is If so, adjust the parameters in the direction of reducing the compensation amount.

[0250] More specifically, the adjustment amount is calculated using a proportional-integral controller. Let the current deviation be... The adjustment amount output by the controller for:

[0251]

[0252] in, This is the proportionality coefficient. This is the integral coefficient. The proportional part determines the adjustment range based on the current deviation, while the integral part eliminates static error based on the cumulative value of the deviation.

[0253] The proportional coefficient of the proportional-integral controller and integral coefficient The system response characteristics are preset and adaptively corrected during operation based on the adjustment effect. If the deviation decreases after adjustment, the coefficient remains unchanged; if the deviation increases, the proportional coefficient is appropriately reduced to avoid oscillation.

[0254] The adjusted diagnostic threshold will be used in the determination process of step one in the next diagnostic cycle, and the adjusted compensation parameters will be used in the next compensation execution of the same type of attenuation, thus forming a complete closed-loop feedback chain.

[0255] Edge computing nodes store the data of each compensation operation in a knowledge base.

[0256] More specifically, the knowledge base can be implemented using a lightweight relational database and deployed in the local storage of edge computing nodes.

[0257] The stored data includes: operation time, operating conditions at the time, such as traffic flow, temperature, vibration intensity, attenuation type diagnosis results, compensation strategy and specific parameters adopted, actual compensation gain, deviation analysis results and attribution conclusions.

[0258] As runtime accumulates, the data gathered in the knowledge base can be used to continuously optimize the initial parameters of the compensation strategy.

[0259] For example, statistical analysis can be used to find the optimal combination of compensation parameters under different working conditions and update them to new initial values.

[0260] 5. Step Five:

[0261] Dynamic power allocation.

[0262] Based on steps one to four above, this embodiment also includes a step of dynamically allocating the power of communication signals and sensing signals.

[0263] This step is independent of the main compensation process, but runs in parallel with it.

[0264] More specifically, the edge computing nodes are pre-defined with a synesthetic joint performance function E:

[0265]

[0266] in, For communication quality indicators, As a sensor accuracy indicator, and These are the communication weighting coefficient and the sensing weighting coefficient, respectively, satisfying... The constraints.

[0267] As a preferred method, and The values ​​range from 0.1 to 0.9.

[0268] The specific calculation method is as follows:

[0269] The normalized value is the reciprocal of the bit error rate (BER) of the communication signal. The value of .

[0270] The normalized benchmark is the bit error rate from to Logarithmic mapping between:

[0271]

[0272] when hour, This indicates the best communication quality;

[0273] when hour, This indicates that the communication quality is close to being unusable;

[0274] The intermediate value is determined by logarithmic interpolation.

[0275] The use of logarithmic normalization is based on the following considerations: the bit error rate in optical fiber communication systems is usually characterized by orders of magnitude, and the change in its logarithmic value can more accurately reflect the engineering perception of the communication link quality. The change in system margin corresponding to each order of magnitude increase in the bit error rate is roughly the same.

[0276] Furthermore,

[0277] The specific calculation method is as follows: based on the signal-to-noise ratio of the sensing signal. The normalized value is used as The value of is determined by the normalization benchmark, which is a linear mapping of the signal-to-noise ratio from 0 dB to 40 dB.

[0278]

[0279] when hour, This indicates the highest sensing accuracy;

[0280] when hour, This indicates that the sensing signal has been overwhelmed by noise.

[0281] The intermediate value is determined by linear interpolation.

[0282] The use of linear normalization is based on the following considerations: the signal-to-noise ratio (SNR) of sensor signals is usually measured in engineering as a linear amplitude, and the improvement in signal discernibility corresponding to the same dB increase in SNR is roughly the same. Although both use different normalization benchmarks, they are both mapped to a unified [0,1] interval, and after weighted combination, the overall synesthetic performance can be comprehensively evaluated on the same scale.

[0283] Let the upper limit of the total transmit power of the optical transmitter in the composite network be . This upper limit is constrained by the maximum allowable output power of the optical transmitter and the nonlinear effect threshold of the optical fiber. Edge computing nodes are dynamically adjusted based on the real-time operating conditions of the highway. and The value of is determined, and the power distribution ratio of the communication signal and the sensor signal is adjusted synchronously according to the following linear mapping rule:

[0284] Communication signal transmission power Sensor signal transmission power ,and .

[0285] More specifically, real-time operating condition identification can be achieved in the following ways:

[0286] Edge computing nodes connect to the highway traffic management system to obtain traffic flow data for the current road segment;

[0287] Obtain the current time using the clock module;

[0288] Obtain the current weather conditions through a meteorological data interface.

[0289] Specifically, during peak traffic periods, such as holidays or weekday morning and evening rush hours, edge computing nodes can, as a preferred method, [provide services / services]. Adjust it to between 0.7 and 0.9 to prioritize communication quality; accordingly... .

[0290] During off-peak hours at night, such as from midnight to 5 a.m. each day, edge computing nodes can, as a preferred approach, [perform certain actions]. Adjusting it to between 0.6 and 0.8 improves sensing accuracy. .

[0291] Furthermore, during extreme weather or emergencies, such as heavy rain, dense fog, or traffic accidents, edge computing nodes can, as a preferred approach, [provide additional information / services]. Adjust the value to between 0.8 and 0.9 to prioritize the sensor warning function. .

[0292] Power allocation adjustment is achieved through drive current control of the optical transmitter in the composite network, and the communication signal transmission power is related to... The values ​​are positively correlated. As the value increases, the transmission power of the communication signal increases accordingly;

[0293] Sensor signal transmission power and The values ​​are positively correlated. As the value increases, the transmission power of the sensor signal increases accordingly. The specific method for controlling the transmission power is as follows: based on... and The change in the amount of the drive current is adjusted proportionally, and the upper limit of the adjustment is limited by the maximum allowable output power of the optical transmitter and the nonlinear effect threshold of the optical fiber.

[0294] After adjusting the transmit power, the power distribution unit sends the adjusted transmit power value to the compensation execution unit. When adjusting the optical amplifier gain, the compensation execution unit dynamically calculates the remaining available gain margin based on the current transmit power level, ensuring that the superposition effect of transmit power and link gain does not exceed the nonlinear effect threshold of the optical fiber. Specifically, when the transmit power is adjusted to the upper limit, the gain adjustment of the optical amplifier decreases accordingly; when the transmit power decreases, the gain adjustment of the optical amplifier can increase accordingly.

[0295] The following are instructions regarding threshold calibration during the initial system deployment:

[0296] The initial values ​​of the aforementioned preset thresholds are reference values ​​determined under standard laboratory conditions using optical fibers of the same type and specifications as those used in the field. In actual engineering deployments, due to differences in optical fiber manufacturing batches, laying conditions, and individual differences in optoelectronic devices, the initial values ​​of each threshold may not be optimal.

[0297] Therefore, after system deployment and before formal operation, it is recommended to collect at least 24 hours of continuous operating data to calibrate the initial thresholds using the statistical results. The calibration method is as follows: calculate the actual statistical distribution of various attenuation characteristic parameters and adjust each threshold to a reasonable quantile in the statistical distribution. After calibration, the system enters normal operation mode, and each threshold is continuously optimized under the closed-loop adjustment mechanism in step four.

[0298] II. Example 2:

[0299] This embodiment provides a signal attenuation compensation system for a 5G and fiber optic sensing composite network on a highway. This system is used to implement the method described in Embodiment 1.

[0300] The system comprises a data acquisition unit, a diagnostic unit, a strategy generation unit, a compensation execution unit, and a feedback optimization unit. Each unit can be deployed at edge computing nodes along the highway, or, as needed, on the centralized server of the road section management center; the specific deployment method will be determined based on the actual engineering conditions.

[0301] The acquisition unit is used to acquire in real time the first state parameters of communication signals and the second state parameters of sensor signals in the composite network, as well as the environmental physical parameters along the optical fiber. The acquisition unit connects to the optical power detection module and the baseband processing unit via a data interface. The coherent receiver module, temperature sensor, vibration sensor, and strain sensor are connected. The acquisition unit reads the output data from each sensor according to a preset sampling period, performs timestamp alignment and format standardization processing, and then outputs the data to the diagnostic unit.

[0302] The diagnostic unit receives data output from the acquisition unit and performs correlation analysis to diagnose signal attenuation sources along the optical fiber online, outputting attenuation type labels for each optical fiber segment. The diagnostic unit internally stores initial values ​​for a first preset threshold, a second preset threshold, a third preset threshold, and a fourth preset threshold, and performs attenuation type determination and priority sorting according to the method described in step one of Embodiment 1.

[0303] The strategy generation unit receives the attenuation type label and attenuation severity index output by the diagnostic unit, and accordingly matches the corresponding compensation strategy from a preset differentiated compensation strategy library to generate a multi-timescale compensation scheme containing compensation methods and compensation parameters. The strategy generation unit internally stores a compensation strategy mapping table and performs strategy lookup and parameter interpolation according to the method described in step two of embodiment one.

[0304] The compensation execution unit receives and executes the compensation scheme output by the strategy generation unit. During execution, the environmental physical parameters extracted from the communication signal are predictively adjusted based on the compensation reference sensor signal, and the channel state information extracted from the communication signal is used for baseline calibration based on the compensation reference sensor signal. The compensation execution unit is connected to the tunable optical amplifier, photodetector bias control terminal, adaptive filter, transient response compensation module, and redundant link switching switch via a control interface, and controls the operation of each execution device according to the instructions in the compensation scheme.

[0305] The feedback optimization unit receives the compensated signal output by the compensation execution unit, evaluates the compensation effect, calculates the compensation gain, and adjusts the diagnostic threshold in the diagnostic unit and the compensation parameters in the strategy generation unit in a closed loop based on the comparison result of the compensation gain and the preset threshold. The adjusted parameters are then sent back to the diagnostic unit and the strategy generation unit, respectively. The feedback optimization unit internally stores the proportional coefficient and integral coefficient of the proportional-integral controller and performs deviation attribution and parameter adjustment according to the method described in step four of embodiment one.

[0306] The power distribution unit is used to establish the inductive joint performance function. The communication weighting coefficient is dynamically adjusted based on the real-time operating conditions of the highway. and sensing weight coefficient The value of the signal is determined, and the transmission power allocation ratio of the communication signal and the sensor signal is adjusted synchronously according to the linear mapping rule. The real-time operating conditions include peak traffic periods, nighttime off-peak periods, and extreme weather or emergency conditions. The power allocation unit identifies the real-time operating conditions by obtaining traffic flow data through the highway traffic management system, obtaining the current time through the clock module, and obtaining weather conditions through the meteorological data interface. The adjustment of the power allocation is achieved through the drive current control of the optical transmitter in the composite network. The transmission power of the communication signal is related to the signal transmission power. The value is positively correlated with the sensor signal transmission power. The values ​​are positively correlated.

[0307] The data flow relationships between the units are as follows: the output data from the acquisition unit flows to the diagnostic unit, the output data from the diagnostic unit flows to the strategy generation unit, the output data from the strategy generation unit flows to the compensation execution unit, the output signal from the compensation execution unit flows to the feedback optimization unit, and the adjustment parameters from the feedback optimization unit are fed back to the diagnostic unit and the strategy generation unit, respectively, forming a complete closed-loop data flow. The power allocation unit is connected to the compensation execution unit, obtains the quality indicators of the communication signal and the sensor signal from the compensation execution unit, and sends the adjusted power allocation parameters back to the compensation execution unit; at the same time, the power allocation unit is connected to the acquisition unit, obtains state parameters such as bit error rate and signal-to-noise ratio from the acquisition unit for calculating the joint communication and sensing performance function.

[0308] Units communicate with each other through standardized data interfaces. Data can be in common serialization formats such as JSON or Protocol Buffers to ensure interoperability and consistency of data parsing between units.

[0309] The initialization of each unit is completed automatically during system startup. The initialization process includes: each unit reading initial parameters from the configuration file, the diagnostic unit loading initial values ​​of preset thresholds, the strategy generation unit loading the compensation strategy mapping table, and the feedback optimization unit loading the initial coefficients of the proportional-integral controller. After initialization is complete, the system enters normal operation.

[0310] III. Regarding the storage media, the details are as follows:

[0311] Those skilled in the art will understand that the methods provided in the embodiments of the present invention can be implemented by computer program instructions, which can be stored in a computer-readable storage medium. When the computer program instructions stored in the computer-readable storage medium are read and executed by a processor, all or part of the steps of the method described in Embodiment 1 can be implemented.

[0312] The computer-readable storage media include, but are not limited to, various physical media capable of storing program code, such as ROM, RAM, magnetic disks, optical disks, USB flash drives, and portable hard drives, as well as non-physical storage media such as cloud storage.

[0313] Taking this embodiment as an example, the computer programs corresponding to steps one to five executed by the edge computing node can be pre-programmed into the read-only memory of the edge computing node, or loaded into memory from an external storage medium for execution when the system starts.

[0314] Those skilled in the art should also understand that the functions of the acquisition unit, diagnosis unit, strategy generation unit, compensation execution unit, and feedback optimization unit described in Embodiment 2 can be implemented through software modules, hardware circuits, or a combination of software and hardware. When implemented in software, the functional modules corresponding to the above units can be integrated into the same processor or distributed across multiple processors for collaborative operation.

[0315] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A highway 5G and optical fiber sensing composite network signal attenuation compensation method, characterized in that, Includes the following steps: Step 1: Real-time acquisition of the first state parameters of communication signals and the second state parameters of sensing signals in the composite network, and simultaneous acquisition of environmental physical parameters along the optical fiber. Based on the correlation analysis of the first state parameters, the second state parameters and the environmental physical parameters, online diagnosis of signal attenuation sources along the optical fiber is performed, and attenuation type labels for each optical fiber segment are output. Step 2: Based on the attenuation type label, match the corresponding compensation strategy from the preset differentiated compensation strategy library to generate a multi-timescale compensation scheme that includes compensation methods and compensation parameters; Step 3: Execute the compensation scheme. During the execution process, the environmental physical parameters extracted from the communication signal are predictively adjusted based on the compensation reference sensor signal, and the channel state information extracted from the communication signal is baseline calibrated based on the compensation reference sensor signal. Step four: Evaluate the quality of the compensated communication signal and the quality of the sensor signal, calculate the compensation gain of the communication signal and the compensation gain of the sensor signal respectively, and adjust the diagnostic threshold in step one and the compensation parameters in step two in a closed loop based on the comparison results of the compensation gain and the preset threshold.

2. The highway 5G and fiber sensing composite network signal attenuation compensation method of claim 1, wherein, In step one, the attenuation type label includes intrinsic attenuation, temperature drift attenuation, vibration stress attenuation, and sudden attenuation; The correlation analysis is implemented in the following manner: 1.1 Compare the communication signal amplitude attenuation curve in the first state parameter with the backscattered light intensity distribution curve in the second state parameter segment by segment, and determine the intrinsic attenuation based on the comparison results; 1.2 Extract the temperature change rate and backscattered light intensity change rate in each fiber segment, and perform time-domain cross-correlation analysis on the two in each segment. Determine the temperature drift attenuation segment by segment based on the cross-correlation analysis results. 1.3 Perform coherent analysis on the vibration spectrum intensity in the environmental physical parameters and the time-domain fluctuation amplitude of the backscattered light intensity in the second state parameter, and determine the vibration stress attenuation based on the coherent analysis results; 1.4 When the signal attenuation changes abruptly beyond a preset threshold within a preset time window, it is determined to be a sudden attenuation.

3. The highway 5G and fiber sensing composite network signal attenuation compensation method of claim 1, wherein, In step two, The multi-timescale compensation scheme comprises four compensation levels: For intrinsic attenuation of the underlying baseline, the adjustment period for the compensation parameters is on the order of minutes. For mid-layer tracking compensation to address temperature drift decay, the compensation parameter adjustment cycle is on the order of seconds. For real-time compensation of vibration stress attenuation in the upper layer, the adjustment cycle of the compensation parameters is in the millisecond range. The top-level emergency compensation for sudden attenuation has a response time in the microsecond range; The above four levels of compensation operate in parallel, with the top-level emergency compensation having the highest execution priority; The residual error signal of the upper layer compensation serves as the error correction reference for the middle layer compensation, and the DC drift output of the middle layer compensation serves as the gain fine-tuning reference for the lower layer compensation. When the top layer emergency compensation is triggered, an interrupt signal is generated, which freezes the parameter updates of the other three layers at the current moment. After the freeze is lifted, the compensation parameters of the bottom, middle and top layers are smoothly transitioned to the target value by gradually approximating them.

4. The highway 5G and fiber sensing composite network signal attenuation compensation method of claim 1, wherein, In step three, The baseline calibration is achieved in the following manner: 3.1 Extract channel state information from the receiving end of the communication signal, wherein the channel state information includes the amplitude response, phase response and nonlinear distortion of the optical fiber link; 3.2, according to the reference power of the communication signal transmitting end and the measured power of the receiving end, the end-to-end total attenuation of the whole link is calculated, and the calculation formula is wherein, is the reference power at the transmitting end, is the measured power at the receiving end, which total attenuation is used as a global reference baseline for the compensation of the sensing signal; 3.3 When the overall amplitude of the sensing signal deviates from the global baseline by more than the preset deviation range, the deviation amount is used as a compensation control signal.

5. The highway 5G and fiber sensing composite network signal attenuation compensation method of claim 4, wherein, In step three, the degree of nonlinear distortion extracted from the channel state information is used to limit the maximum output power of the optical amplifier during the compensation process. When the degree of nonlinear distortion exceeds the preset nonlinear threshold, the upper limit of the gain of the optical amplifier is temporarily reduced.

6. The highway 5G and fiber sensing composite network signal attenuation compensation method of claim 1, wherein, In step three, Predictive adjustments are achieved in the following ways: Environmental physical parameters are extracted from the receiving end of the sensor signal. When the temperature or vibration parameter change of the optical fiber segment exceeds the preset range, the additional attenuation estimate of the segment is calculated according to the pre-calibrated environmental attenuation coefficient. The estimated value is used as the feedforward control quantity to adjust the optical amplifier gain at the corresponding position of the optical fiber segment in advance; or the transmission power of the communication signal is globally supplemented and adjusted according to the estimated total attenuation of multiple segments in the whole link.

7. The highway 5G and fiber sensing composite network signal attenuation compensation method of claim 1, wherein, In step four, the closed-loop adjustment is implemented as follows: The difference between the actual compensation gain and the expected compensation gain of each fiber segment is taken as the deviation, and the deviation is attributed in the following order: First, if the characteristics of the compensated residual signal do not match the typical characteristics of the attenuation type diagnosed in step one, it is determined that the attenuation type diagnosis is incorrect, and the corresponding diagnosis threshold is adjusted; if the characteristics of the residual signal match the typical characteristics of the diagnosed attenuation type, it is further determined whether the correlation coefficient between the compensated residual attenuation waveform and the typical compensation curve corresponding to the current strategy is lower than the preset matching threshold. If so, it is determined that the compensation strategy selection is inappropriate, and the compensation strategy is re-matched from the strategy library. If the correlation coefficient is not lower than the preset matching threshold, the deviation is attributed to the deviation of the compensation parameter execution, and the compensation parameter is adjusted in reverse within the allowable adjustable range. The data from each compensation operation is stored in a knowledge base for continuous optimization of the initial parameters of the compensation strategy.

8. The highway 5G and fiber sensor composite network signal attenuation compensation method of claim 1, wherein, It also includes a dynamic power allocation step, which runs in parallel with steps one through four, as detailed below: Under the premise that the total transmission power does not exceed the upper limit defined by the maximum allowable output power of the optical transmitter in the composite network and the threshold of the optical fiber nonlinear effect wherein is a communication quality index, is a sensing accuracy index, and are a communication weight coefficient and a sensing weight coefficient, respectively, ; According to the real-time working condition of the expressway, the dynamic adjustment is performed and The value of is adjusted according to the linear mapping rule 、 The transmission power distribution ratio of the communication signal and the sensing signal is synchronously adjusted, wherein is the upper limit of the total transmission power. 9.A system for compensating signal attenuation of a highway 5G and optical fiber sensing composite network, characterized in that, The system is used to implement the method as described in any one of claims 1-8, specifically including: The acquisition unit is used to acquire in real time the first state parameters of communication signals and the second state parameters of sensing signals in the composite network, as well as the environmental physical parameters along the optical fiber. The diagnostic unit is used to receive the data output by the acquisition unit and perform correlation analysis to perform online diagnosis of signal attenuation sources along the optical fiber and output attenuation type labels for each optical fiber segment. The strategy generation unit is used to receive the attenuation type label output by the diagnostic unit, and accordingly match the corresponding compensation strategy from the preset differential compensation strategy library to generate a multi-time scale compensation scheme containing compensation methods and compensation parameters. The compensation execution unit is used to receive the compensation scheme output by the strategy generation unit and execute it. During the execution process, the environmental physical parameters extracted from the communication signal are predictively adjusted based on the compensation reference sensor signal, and the channel state information extracted from the communication signal is baseline calibrated based on the compensation reference sensor signal. The feedback optimization unit is used to receive the compensated signal output by the compensation execution unit and evaluate the compensation effect, calculate the compensation gain, adjust the diagnostic threshold in the diagnostic unit and the compensation parameters in the strategy generation unit in a closed loop according to the comparison result of the compensation gain and the preset threshold, and send the adjusted parameters back to the diagnostic unit and the strategy generation unit respectively. The power allocation unit is used to establish a joint communication and sensing performance function, dynamically adjust the values ​​of communication weight coefficient and sensing weight coefficient according to the real-time operating conditions of the highway, and synchronously adjust the transmission power allocation ratio of communication signal and sensing signal according to the linear mapping rule.

10. A computer readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 8.