An ODN attenuation optimization method and system based on multi-parameter matching

Abstract

A method and system for optimizing ODN attenuation based on multi-parameter matching, relating to the field of electronic digital data processing, is disclosed. The method includes: reading multiple candidate topology schemes of the ODN network to be planned, generating a multi-dimensional parameter vector set, and performing similarity matching calculations with the parameter vectors of completed networks in a historical engineering database to determine a reference ODN network identifier; extracting real received optical power telemetry data based on the reference ODN network identifier to generate a real optical power dataset; reading original design data, calculating the theoretical estimated optical power of each optical network unit according to a preset standard attenuation linear formula to generate a theoretical optical power dataset; determining multi-parameter joint compensation coefficients, performing basic attenuation calculations on multiple candidate topology schemes based on the standard attenuation linear formula, and calculating corrected attenuation values; and determining the candidate topology scheme as the attenuation optimization scheme for the ODN network to be planned. Implementing this application can improve the accuracy of scheme optimization for actual optical distribution networks.

Description

Technical Field

[0001] This application relates to the field of electronic digital data processing, and in particular to an ODN attenuation optimization method and system based on multi-parameter matching. Background Technology

[0002] Optical Distribution Network (ODN), as the core infrastructure of fiber optic access networks, is primarily responsible for distributing optical signals from the Optical Line Terminal (OLT) to various Optical Network Units (ONUs) via passive devices such as optical splitters. During the planning, construction, and operation of the ODN, the overall attenuation value of the optical link is a key indicator determining network communication quality and coverage. To ensure stable optical signal transmission and meet the sensitivity requirements of the receiver, engineers must accurately calculate and optimize the attenuation caused by multiple parameters such as cable length, fusion splices, movable connectors, and optical splitters. This aims to maximize network coverage and optimize construction costs within a limited optical power budget.

[0003] In related technologies, a linear accumulation and comparison screening method based on standard empirical coefficients is adopted. That is, the system first generates multiple candidate cabling topology schemes based on the building floor plan; then, it extracts the optical cable design length and the number of passive nodes from each candidate topology scheme, and performs mathematical multiplication and addition operations strictly according to industry standard fixed attenuation coefficients (such as fixed attenuation rate per kilometer of fiber and fixed connector insertion loss value) to obtain the theoretical estimated attenuation value of each candidate topology scheme; the system compares these theoretical estimated attenuation values ​​with the preset optical power budget, and selects the candidate topology scheme with the smallest theoretical estimated attenuation value as the final engineering implementation drawing.

[0004] However, since the relevant technologies rely solely on fixed standard empirical coefficients for purely theoretical deduction and optimization, they often encounter severe unforeseen losses after actual construction and implementation. This results in the actual optical power of the final deployed network deviating significantly from the safety budget, causing large-scale signal blind spots. In other words, the planning of optical distribution networks in complex buildings using the relevant technologies is not accurate enough. Summary of the Invention

[0005] This application provides an ODN attenuation optimization method and system based on multi-parameter matching, which can improve the accuracy of scheme optimization in practical optical distribution networks.

[0006] Firstly, this application provides an ODN attenuation optimization method based on multi-parameter matching, applied to a data processing system. The method includes: reading multiple candidate topologies of the ODN network to be planned; extracting environmental features, pipeline features, and equipment features from each candidate topology to generate a multi-dimensional parameter vector set; performing similarity matching operations between the multi-dimensional parameter vector set and the as-built network parameter vectors in a historical engineering database to determine a reference ODN network identifier that meets preset similarity conditions; extracting the actual received optical power telemetry data reported by the corresponding online optical network unit from the network management system based on the reference ODN network identifier to generate an actual optical power dataset; reading the original design data corresponding to the reference ODN network identifier and calculating the theoretical estimated optical power of each optical network unit according to a preset standard attenuation linear formula to generate a theoretical optical power dataset; determining multi-parameter joint compensation coefficients based on the actual optical power dataset and the theoretical optical power dataset; performing basic attenuation calculations on multiple candidate topologies based on the standard attenuation linear formula and generating corrected attenuation values ​​for the candidate topologies based on the multi-parameter joint compensation coefficients; and determining the candidate topology that meets the optical power budget threshold and has the smallest corrected attenuation value as the attenuation optimization scheme for the ODN network to be planned.

[0007] In the above embodiments, the data processing system introduces telemetry data from historical real projects through multi-dimensional feature matching, and constructs multi-parameter joint compensation coefficients based on the difference between real and theoretical optical power to correct the basic attenuation. This changes the limitation of traditional theoretical deduction that relies solely on fixed standard empirical coefficients. It uses the real operating status of the isomorphic historical network to feed back the attenuation calculation of the network to be planned, thereby improving the accuracy of ODN network scheme optimization and the reliability of actual deployment.

[0008] In conjunction with some embodiments of the first aspect, in some embodiments, the step of reading multiple candidate topology schemes of the ODN network to be planned, extracting environmental features, pipeline features, and equipment features from each candidate topology scheme, and generating a multidimensional parameter vector set specifically includes: reading multiple candidate topology schemes of the ODN network to be planned, determining vertical cabling sections by combining the three-dimensional spatial orientation data of each candidate topology scheme; extracting the sag drop features and horizontal corner features of the vertical cabling sections to generate a spatial stress feature set; extracting the environmental features, pipeline features, and equipment features of each candidate topology scheme to generate a basic feature set; and concatenating the spatial stress feature set and the basic feature set to generate a multidimensional parameter vector set.

[0009] In the above embodiments, the data processing system extracts the sag drop features and horizontal corner features of the vertical cabling section to generate a spatial force feature set, and splices it with the basic feature set. This incorporates the physical force factors brought about by the three-dimensional spatial structure into the feature evaluation dimension, making up for the deficiency of traditional two-dimensional planar planning in not covering the additional attenuation of optical cable spatial force deformation, and improving the expression accuracy of multi-dimensional parameter vectors for complex cabling scenarios.

[0010] In conjunction with some embodiments of the first aspect, in some embodiments, after concatenating the spatial force feature set and the basic feature set to generate a multidimensional parameter vector set, the method further includes: calculating the Euclidean distance between the spatial force feature set in the multidimensional parameter vector set and the historical force feature set in the as-built network parameter vector to generate a spatial isomorphic distance value; removing as-built network parameter vectors in the historical engineering database whose spatial isomorphic distance value is greater than a preset distance threshold to generate a candidate as-built network set; and determining as-built network identifiers in the candidate as-built network set whose global feature matching degree is higher than a preset matching threshold as reference ODN network identifiers that meet the preset similarity conditions.

[0011] In the above embodiments, the data processing system first eliminates historical networks with excessive differences by using spatial isomorphic distance values, and then performs global feature matching. This constructs a progressive screening mechanism from local spatial physical morphology to global multidimensional features, avoiding the computational power consumption caused by global matching of massive historical data. While ensuring the physical homogeneity of the reference network, it improves the computational efficiency of similarity matching.

[0012] In conjunction with some embodiments of the first aspect, in some embodiments, before determining the multi-parameter joint compensation coefficients based on the real optical power dataset and the theoretical optical power dataset, the method further includes: dividing the optical network units in the real optical power dataset and the theoretical optical power dataset into multiple homogeneous optical network unit clusters according to the splitter topology hierarchy identified by the reference ODN network; calculating the deviation values ​​between the real received optical power and the theoretically estimated optical power of all optical network units within each homogeneous optical network unit cluster to generate a node deviation matrix; and extracting the mean deviation value of each deviation value within the same homogeneous optical network unit cluster in the node deviation matrix to generate a common-mode deviation component.

[0013] In the above embodiments, the data processing system divides the optical network unit clusters of the same source according to the topology hierarchy of the optical splitter, and extracts the mean deviation of the node deviation matrix within the same cluster to generate common mode deviation components. Based on the physical topology, it separates the systematic attenuation error caused by the shared backbone link or the upper-level equipment, eliminates the interference of single node anomalies on the overall compensation calculation, and enhances the structural targeting of the deviation analysis.

[0014] In conjunction with some embodiments of the first aspect, in some embodiments, after the step of extracting the mean deviation of each deviation value within the same optical network unit cluster in the node deviation matrix to generate a common-mode deviation component, the method further includes: subtracting the common-mode deviation component from each deviation value in the node deviation matrix to generate a differential-mode deviation component; mapping the common-mode deviation component to the backbone pipeline features identified by the reference ODN network to generate a backbone-level compensation coefficient; mapping the differential-mode deviation component to the branch pipeline features identified by the reference ODN network to generate a branch-level compensation coefficient; and concatenating the backbone-level compensation coefficient and the branch-level compensation coefficient to generate a multi-parameter joint compensation coefficient.

[0015] In the above embodiments, the data processing system subtracts the common-mode deviation component from the node deviation to generate the differential-mode deviation component, and maps both to the trunk and branch pipeline characteristics to generate corresponding compensation coefficients. This achieves decoupled analysis of the common attenuation of the trunk and the individual attenuation of the branches, so that the construction of the compensation coefficients can accurately correspond to different physical levels of the network topology, and improves the calculation accuracy of the final corrected attenuation value.

[0016] In conjunction with some embodiments of the first aspect, in some embodiments, the step of performing basic attenuation calculations on multiple candidate topology schemes based on the standard attenuation linear formula and generating corrected attenuation values ​​for the candidate topology schemes by combining multi-parameter joint compensation coefficients specifically includes: performing basic attenuation calculations on multiple candidate topology schemes based on the standard attenuation linear formula to generate basic attenuation values; extracting micro-environmental disturbance node data from the environmental features of the candidate topology schemes, adding environmental fluctuation margins to the multi-parameter joint compensation coefficients to generate compensation coefficients; generating corrected attenuation envelopes for the candidate topology schemes based on the basic attenuation values ​​and compensation coefficients, and using the corrected attenuation envelopes as corrected attenuation values; the corrected attenuation envelopes include a lower attenuation limit and an upper attenuation limit.

[0017] In the above embodiments, the data processing system extracts the environmental fluctuation margin from the micro-environmental disturbance node data and generates a corrected attenuation envelope containing upper and lower limits. This quantifies the dynamic environmental change factors and introduces them into the attenuation calculation process, breaking the traditional evaluation mode of a single fixed attenuation value. By presenting the attenuation change range in the form of an interval envelope, the robustness of the attenuation evaluation to complex environmental fluctuations is improved.

[0018] In conjunction with some embodiments of the first aspect, in some embodiments, after generating the modified attenuation envelope of the candidate topology scheme based on the basic attenuation value and the compensation coefficient, and using the modified attenuation envelope as the modified attenuation value, the method further includes: comparing the upper limit of the attenuation of the modified attenuation envelope of each candidate topology scheme with the optical power budget threshold to generate a subset of candidate schemes that do not exceed the optical power budget threshold; calculating the difference between the upper limit of attenuation and the lower limit of attenuation of each candidate topology scheme in the subset of candidate schemes to generate the envelope fluctuation amplitude; and selecting the candidate topology scheme with the smallest envelope fluctuation amplitude from the subset of candidate schemes as the stable optimization scheme of the ODN network to be planned.

[0019] In the above embodiments, the data processing system selects the scheme with the smallest envelope fluctuation amplitude from the subset of candidate schemes that meet the optical power budget as the stable optimization scheme. Under the premise of ensuring the basic communication threshold, the focus of the scheme evaluation is shifted to the stability of the network against environmental interference, which effectively reduces the risk of communication interruption caused by severe attenuation fluctuations in the network during long-term operation and improves the operational stability of the ODN network.

[0020] In a second aspect, embodiments of this application provide a data processing system comprising: one or more processors and a memory; the memory is coupled to the one or more processors and is used to store computer program code, the computer program code including computer instructions, wherein the one or more processors invoke the computer instructions to cause the data processing system to perform the method described in the first aspect and any possible implementation thereof.

[0021] Thirdly, embodiments of this application provide a computer-readable storage medium including instructions that, when executed on a data processing system, cause the data processing system to perform the method described in the first aspect and any possible implementation thereof.

[0022] Fourthly, embodiments of this application provide a computer program product containing instructions that, when the computer program product is run on a data processing system, cause the data processing system to perform the method described in the first aspect and any possible implementation thereof.

[0023] Understandably, the data processing system provided in the second aspect, the computer storage medium provided in the third aspect, and the computer program product provided in the fourth aspect are all used to execute the methods provided in the embodiments of this application. Therefore, the beneficial effects they can achieve can be referred to the beneficial effects in the corresponding methods, and will not be repeated here.

[0024] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages:

[0025] 1. By matching the multi-dimensional parameter vectors of candidate topologies with historical completed networks to determine the reference ODN network, and determining the multi-parameter joint compensation coefficient based on the difference between the actual optical power and the theoretically estimated optical power of the reference network, and then introducing the compensation coefficient into the basic attenuation calculation to generate the corrected attenuation value and optimize the scheme, it is possible to perform multi-dimensional joint compensation for pure theoretical attenuation based on historical real operating data. This effectively solves the problem in existing technologies where relying solely on fixed empirical coefficients leads to deviations of optical power from the budget after actual construction, thereby improving the accuracy of optical distribution network scheme optimization.

[0026] 2. By combining the three-dimensional spatial orientation data of candidate topology schemes to extract the sag drop features and horizontal corner features of vertical cabling sections, and splicing them with environmental, pipeline and equipment features to generate a multi-dimensional parameter vector set, spatial stress factors can be included in the feature extraction range. This effectively solves the problem in the existing technology that the influence of complex spatial structures on the stress and additional attenuation of optical cables is not considered, thereby enriching the feature vector dimensions and improving the accuracy of historical network similarity matching.

[0027] 3. By using micro-environmental disturbance node data extracted from environmental features as compensation coefficients with added environmental fluctuation margins, and combining the basic attenuation value to generate a corrected attenuation envelope containing upper and lower attenuation limits as the final corrected attenuation value, the dynamic fluctuation factors of the environment can be introduced into the attenuation calculation process. This effectively solves the problem that a single fixed attenuation value in the existing technology cannot cope with the complex environmental changes in actual engineering, thereby improving the robustness of network attenuation assessment to environmental fluctuations. Attached Figure Description

[0028] Figure 1 This is a flowchart illustrating an ODN attenuation optimization method based on multi-parameter matching in an embodiment of this application.

[0029] Figure 2 This is another flowchart illustrating the ODN attenuation optimization method based on multi-parameter matching in the embodiments of this application;

[0030] Figure 3 This is a schematic diagram of the physical device structure of a data processing system in an embodiment of this application. Detailed Implementation

[0031] The terminology used in the following embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. As used in the specification of this application, the singular expressions “a,” “an,” “the,” “the,” and “this” are intended to include the plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this application refers to any or all possible combinations including one or more of the listed items.

[0032] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature, and in the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more.

[0033] In the construction and operation of Optical Distribution Networks (ODNs), the planning of the network topology directly determines the quality of signal transmission. Traditional planning methods often rely on static empirical coefficients, failing to fully consider the complex factors in the actual construction environment, such as pipeline stress and micro-environmental disturbances. This application proposes an ODN attenuation optimization method based on multi-parameter matching. By introducing real optical power telemetry data from historical projects and combining multi-dimensional spatial and environmental characteristics, a multi-parameter joint compensation mechanism is constructed, thereby achieving accurate prediction of attenuation values ​​and optimal scheme selection during the planning stage.

[0034] The following describes the process of the method provided in this implementation. Please refer to [link / reference]. Figure 1 This is a flowchart illustrating an ODN attenuation optimization method based on multi-parameter matching in an embodiment of this application.

[0035] S101. Read multiple candidate topology schemes of the ODN network to be planned, extract the environmental features, pipeline features and equipment features of each candidate topology scheme, and generate a multi-dimensional parameter vector set.

[0036] Among them, the candidate topology scheme represents the preset optical cable cabling path and node distribution scheme in a specific building or area; environmental characteristics refer to external environmental parameters such as temperature, humidity, and vibration in the cabling area; pipeline characteristics are used to represent the optical cable's material, wire diameter, laying method, and other attributes; equipment characteristics represent the model and specifications of passive devices such as splitters and junction boxes; and the multidimensional parameter vector set refers to the vector data set formed by numerically encoding and combining the above-mentioned characteristics.

[0037] Specifically, in the early stages of ODN network planning, the data processing system receives multiple candidate topology schemes output by the design tools. The system parses the data files of these schemes, extracting environmental characteristics, pipeline characteristics, and equipment characteristics. By standardizing and vectorizing these characteristic data, the system generates a set of multi-dimensional parameter vectors that comprehensively describe the physical characteristics of each candidate scheme.

[0038] It should be noted that the data processing system employs pre-defined quantization mapping rules during the generation of the basic feature set. For example, temperature parameters in environmental features are discretized and encoded according to preset temperature ranges (e.g., -10℃ to 0℃ is encoded as 1, 0℃ to 20℃ as 2, and above 20℃ as 3); for pipeline features, different materials (e.g., G.652D, G.657A) and wire diameters are mapped to corresponding standard attenuation reference value vectors; for equipment features, the splitting ratio of the beam splitter (e.g., 1:8, 1:16) and junction box type are converted into corresponding nominal insertion loss values. Through the above quantization mapping, unstructured physical attributes are transformed into a numerical basic feature set with unified dimensions, thereby ensuring the mathematical operability of subsequent similarity calculations.

[0039] In some embodiments, feature extraction and vector generation can be achieved in multiple ways: Optionally, the data processing system parses the CAD drawing files of candidate topology schemes to extract spatial coordinates and equipment attribute information; queries a preset feature mapping table based on the extracted information to obtain the corresponding numerical features; and concatenates the numerical features according to a preset dimensional order to generate a multi-dimensional parameter vector. Optionally, the data processing system receives BIM model data through an interface, uses geometric analysis algorithms to extract pipeline routing and environmental parameters; normalizes the extracted parameters; and combines the normalized parameters into a multi-dimensional parameter vector. It is understood that other methods can also be used to achieve feature extraction and vector generation, which are not limited here.

[0040] S102. Perform similarity matching operation between the multidimensional parameter vector set and the as-built network parameter vector in the historical project database to determine the reference ODN network identifier that meets the preset similarity conditions.

[0041] Among them, the historical project database represents the storage system that stores ODN network data that has been completed and put into operation; the completed network parameter vector refers to the feature vector representation of the historical network at the time of completion; the similarity matching operation is used to represent the process of calculating the data similarity between the multidimensional parameter vector and the completed network parameter vector; and the reference ODN network identifier represents the unique identifier of the completed network with the highest matching degree and meeting the conditions.

[0042] Specifically, after generating a multidimensional parameter vector set, the data processing system accesses the historical project database to retrieve the as-built network parameter vectors stored therein. The system then uses a pre-defined similarity algorithm (such as cosine similarity or Euclidean distance) to calculate the similarity value between each vector in the multidimensional parameter vector set and the as-built network parameter vectors. Based on the calculation results, the system filters out as-built networks with similarity values ​​greater than a preset threshold and extracts their corresponding network identifiers as reference ODN network identifiers for subsequent retrieval of their actual operational data.

[0043] In some embodiments, similarity matching can be implemented in several ways: Optionally, the data processing system calculates the cosine similarity between the multidimensional parameter vector and each completed network parameter vector; sorts the similarity results in descending order; and selects the network identifier with the highest ranking and a value greater than a threshold as the reference ODN network identifier. Optionally, the data processing system uses the K-nearest neighbor algorithm to find the K nearest completed network parameter vectors to the multidimensional parameter vector in the historical project database; calculates the weighted average similarity of these K vectors; and selects the network identifier with the highest weighted average similarity as the reference ODN network identifier. It is understood that other methods can also be used to implement similarity matching, and this is not limited here.

[0044] In some embodiments, the sheer volume of data in the historical engineering database can lead to excessively long matching calculation times. To address this, the data processing system can employ the Locality Sensitive Hash (LSH) algorithm to reduce the dimensionality of historical data and create an index, thereby accelerating the similarity matching process.

[0045] S103. Based on the reference ODN network identifier, extract the actual received optical power telemetry data reported by the corresponding online optical network unit from the network management system, and generate an actual optical power dataset.

[0046] Among them, the network management system refers to the software platform responsible for monitoring and managing the operating status of the ODN network; the online optical network unit refers to the ONU device that is currently active and communicating; the actual received optical power telemetry data is used to represent the received optical signal strength value actually measured and reported by the ONU device; the actual optical power dataset refers to the data set formed by summarizing all relevant ONU telemetry data.

[0047] Specifically, the data processing system uses a defined reference ODN network identifier to initiate a data query request to the network management system. The network management system locates the corresponding historical network based on this identifier and extracts the actual received optical power telemetry data reported by all online optical network units within that network during a specific time period. The data processing system receives this data, performs cleaning and noise reduction processing, removes outliers, and ultimately generates a true optical power dataset reflecting the actual operating status of the historical network.

[0048] In some embodiments, data extraction and generation can be achieved in multiple ways: Optionally, the data processing system sends a query command containing the reference ODN network identifier to the network management system via an API interface; receives JSON-formatted telemetry data returned by the network management system; parses the JSON data and extracts the optical power field, storing it in a database to generate a real optical power dataset. Optionally, the data processing system directly accesses the underlying database of the network management system, executes SQL queries to obtain telemetry records with corresponding identifiers; performs time-series smoothing on the obtained records; and exports the smoothed data as a CSV file to generate a real optical power dataset. It is understood that other methods can also be used to achieve data extraction and generation, which are not limited here.

[0049] S104. Read the original design data corresponding to the reference ODN network identifier, calculate the theoretical estimated optical power of each optical network unit according to the preset standard attenuation linear formula, and generate a theoretical optical power dataset.

[0050] Among them, the original design data refers to the design drawings and parameter information of the ODN network during the planning stage; the standard attenuation linear formula refers to the industry-standard calculation formula based on a fixed coefficient for attenuation accumulation; the theoretical estimated optical power is used to represent the theoretical value of the optical signal strength at the ONU receiver calculated based on the design parameters; the theoretical optical power dataset refers to the data set formed by summarizing all theoretical estimated values.

[0051] Specifically, the data processing system retrieves the corresponding original design data from the design archive based on the reference ODN network identifier, including fiber optic cable length, number of connectors, and splitter splitting ratio. Applying a pre-defined standard attenuation linear formula and combining it with the design parameters of each link segment, the system calculates the theoretical total attenuation of the optical signal transmitted from the OLT to each optical network unit, thereby deriving the theoretical estimated optical power of each optical network unit. The data processing system then summarizes these theoretical calculation results to generate a theoretical optical power dataset.

[0052] In some embodiments, theoretical calculations and generation can be achieved in multiple ways: Optionally, the data processing system parses the original design data, extracts the length and node type of each link segment; multiplies the length by the fiber attenuation coefficient according to the standard attenuation linear formula, and adds the node insertion loss; subtracts the total attenuation from the OLT transmit power to obtain the theoretically estimated optical power and generate a dataset. Optionally, the data processing system constructs a topology tree model of a reference ODN network; traverses from the root node (OLT) to the leaf nodes (ONU), accumulating the standard attenuation value layer by layer; calculates the theoretically estimated optical power of each leaf node and summarizes it to generate a theoretical optical power dataset. It is understood that other methods can also be used to achieve theoretical calculations and generation, which are not limited here.

[0053] S105. Determine the multi-parameter joint compensation coefficients based on the real optical power dataset and the theoretical optical power dataset.

[0054] Among them, the multi-parameter joint compensation coefficient represents a set of numerical multipliers or additive terms used to correct the theoretical attenuation calculation deviation.

[0055] Specifically, the data processing system aligns the corresponding node data in the real optical power dataset and the theoretical optical power dataset. The system calculates the difference between the real optical power and the theoretically estimated optical power for each node, and analyzes the distribution of this difference across different link levels and device types. Based on the analysis results, the system uses multiple regression or machine learning algorithms to extract multi-parameter joint compensation coefficients that reflect the actual attenuation characteristics, for use in subsequent correction calculations.

[0056] It should be noted that the data processing system can use a multiple linear regression model based on the least squares method to determine the joint compensation coefficients for multiple parameters. Specifically, the difference between the actual optical power and the theoretically estimated optical power is set as the target variable Y, and the various features in the multidimensional parameter vector set (such as overhang difference, horizontal rotation angle, temperature encoding, etc.) are used as the set of independent variables. The data processing system constructs a regression equation:

[0057] ,in It is the nth independent variable The regression coefficients represent the degree of influence of the nth physical feature on the light attenuation deviation (i.e., the compensation weight); the equation is fitted using a historical real optical power dataset to obtain the regression coefficient vector:

[0058] The regression coefficient vector β serves as the multi-parameter joint compensation coefficient, and the values ​​of each dimension directly reflect the compensation weight of the corresponding physical feature on the actual optical attenuation. It is the random error term or residual in a statistical model, representing the random errors in the model that cannot be fixed. The portion explained by these known variables is due to the fact that in actual ODN (Optical Distribution Network) projects, no matter how many features we extract (temperature, stress, material, etc.), there will always be some unquantifiable, accidental factors that cause changes in optical power. For example, extremely small differences in loss caused by slight hand tremors when splicing optical fibers, extremely small dust particles at the joint, or measurement errors of the optical power meter itself.

[0059] In some embodiments, the compensation coefficients can be determined in several ways: Optionally, the data processing system calculates the node deviation matrix between the real and theoretical data; performs principal component analysis (PCA) on the deviation matrix to extract the main influencing factors; and calculates the multi-parameter joint compensation coefficients based on the main influencing factors. Optionally, the data processing system inputs the real and theoretical data into a pre-trained neural network model; the model outputs the weight adjustment values ​​of each parameter; and generates the multi-parameter joint compensation coefficients based on the weight adjustment values. It is understood that other methods can also be used to determine the compensation coefficients, and this is not limited here.

[0060] In some embodiments, the deviation between real and theoretical data may be extremely uneven. To address this, the data processing system can use clustering algorithms to group nodes and calculate local multi-parameter joint compensation coefficients for different groups. In some embodiments, after grouping nodes using clustering algorithms and calculating local multi-parameter joint compensation coefficients, the rigid division of cluster boundaries may lead to adjacent optical network units in the physical topology being assigned to different clusters. This can result in abrupt changes in compensation values ​​between adjacent nodes when applying local compensation coefficients, violating the continuity of physical link attenuation. To address this, the data processing system introduces a fuzzy membership smoothing mechanism. For optical network units located in the cluster boundary region, the data processing system calculates their fuzzy membership weights to multiple adjacent clusters. Subsequently, the data processing system extracts the local multi-parameter joint compensation coefficients of these adjacent clusters and performs a weighted fusion calculation based on the fuzzy membership weights to generate a smooth transition compensation coefficient suitable for the boundary node. This eliminates spatial abrupt changes in compensation values ​​and ensures the continuity and rationality of attenuation compensation in the physical topology.

[0061] S106. Based on the standard attenuation linear formula, perform basic attenuation calculations on multiple candidate topology schemes respectively, and generate corrected attenuation values ​​for candidate topology schemes by combining multi-parameter joint compensation coefficients.

[0062] Among them, the basic attenuation calculation refers to the preliminary attenuation calculation process using only the standard formula; the corrected attenuation value refers to the attenuation estimate that is closer to the actual situation after adjustment by the compensation coefficient.

[0063] Specifically, for multiple candidate topology schemes of the ODN network to be planned, the data processing system first calculates the basic attenuation value of each scheme using the standard linear attenuation formula. Then, the data processing system introduces the multi-parameter joint compensation coefficient determined in step S105, and applies this coefficient to the calculation result of the basic attenuation value to correct theoretical deviations. Through this combined calculation, the data processing system generates corrected attenuation values ​​for each candidate topology scheme, making them more consistent with actual engineering expectations.

[0064] In some embodiments, the generation of the corrected attenuation value can be achieved in several ways: Optionally, the data processing system calculates the basic attenuation value of the candidate topology scheme; performs matrix multiplication on the basic attenuation value and the multi-parameter joint compensation coefficient; and outputs the calculation result as the corrected attenuation value. Optionally, the data processing system decomposes the basic attenuation value into multiple attenuation components; adds the multi-parameter joint compensation coefficient of the corresponding dimension to each attenuation component; and summarizes the adjusted attenuation components to generate the corrected attenuation value. It is understood that other methods can also be used to generate the corrected attenuation value, which are not limited here.

[0065] S107. Determine the candidate topology scheme that meets the optical power budget threshold and minimizes the modified attenuation value, and use it as the attenuation optimization scheme for the ODN network to be planned.

[0066] Among them, the optical power budget threshold represents the maximum link attenuation limit allowed for network devices to communicate normally; the attenuation optimization scheme refers to the optimal network cabling and node configuration scheme selected after evaluation.

[0067] Specifically, the data processing system obtains a preset optical power budget threshold and compares the corrected attenuation value of each candidate topology scheme with this threshold one by one. The data processing system eliminates schemes whose corrected attenuation values ​​exceed the threshold, forming a set of feasible schemes. In the set of feasible schemes, the data processing system compares the magnitude of the corrected attenuation values ​​of each scheme, selects the scheme with the smallest value as the attenuation optimization scheme for the ODN network to be planned, and outputs the detailed design parameters of this scheme for engineering implementation.

[0068] In some embodiments, scheme determination can be achieved in multiple ways: Optionally, the data processing system iterates through the corrected attenuation values ​​of all candidate topology schemes; determines whether the corrected attenuation value is less than the optical power budget threshold; and selects the scheme with the smallest corrected attenuation value among the schemes that meet the condition as the attenuation optimization scheme. Optionally, the data processing system defines the difference between the corrected attenuation value and the optical power budget threshold as the margin; sorts all candidate topology schemes in descending order according to the margin; and selects the scheme ranked first as the attenuation optimization scheme. It is understood that other methods can also be used to determine the scheme, and this is not limited here.

[0069] In some embodiments, there may be multiple candidate topology schemes with identical corrected attenuation values ​​that all meet the threshold. In such cases, the data processing system can further compare the pipeline laying costs of these schemes and select the scheme with the lowest cost as the final attenuation optimization scheme.

[0070] To further improve the precision of feature extraction and address the attenuation evaluation problem under complex spatial structures, this application provides further embodiments based on the above embodiments. The method provided in this embodiment is described in more detail below. Please refer to... Figure 2 This is another flowchart illustrating the ODN attenuation optimization method based on multi-parameter matching in this application embodiment.

[0071] S201. Read multiple candidate topology schemes of the ODN network to be planned, and determine the vertical cabling section by combining the three-dimensional spatial orientation data of each candidate topology scheme.

[0072] Among them, the candidate topology scheme represents the preset optical cable cabling path and node distribution scheme in a specific building or area; the three-dimensional spatial direction data represents the spatial path information of the optical cable in the X, Y, and Z coordinate axes; the vertical cabling section refers to the cable segment in the optical cable path where the Z-axis coordinate changes significantly.

[0073] Specifically, the data processing system reads candidate topology schemes and analyzes their three-dimensional spatial orientation data. It extracts the coordinates of key nodes along the fiber optic cable route and identifies link segments where the elevation difference exceeds a preset threshold by calculating the elevation difference between adjacent nodes. The system marks these link segments as vertical cabling sections for subsequent targeted extraction of stress characteristics.

[0074] In some embodiments, the determination of vertical cabling sections can be achieved in multiple ways: Optionally, the data processing system parses the building BIM model file and extracts the Z-axis coordinate set of the pipeline path; the data processing system performs a difference operation on adjacent coordinates to obtain the elevation change; the data processing system compares the elevation change with a preset judgment threshold and outputs a list of vertical cabling sections that meet the conditions. Optionally, the data processing system reads CAD 3D drawing data and extracts the elevation attributes of pipeline nodes; the data processing system uses a spatial geometric algorithm to fit the pipeline routing curve and calculates the derivative of the curve in the Z-axis direction; the data processing system extracts curve segments with derivative values ​​greater than a preset slope as vertical cabling sections. It is understood that other methods can also be used to determine vertical cabling sections, which are not limited here.

[0075] In some embodiments, there may be situations where inclined wiring on slopes or stairwells causes a large change in the Z-axis, but the force is not purely vertical. To address this, the data processing system introduces an XY plane projection length comparison mechanism. Only when the ratio of the Z-axis change to the horizontal projection length is greater than a preset slope threshold is it determined to be a vertical wiring segment, thus eliminating interference from inclined wiring.

[0076] S202. Extract the sag drop features and horizontal corner features of the vertical wiring section to generate a spatial force feature set.

[0077] Among them, the sag drop feature represents the absolute value of the elevation difference between the two ends of the vertical cabling section; the horizontal turning angle feature refers to the bending angle of the optical cable on the horizontal plane projection; the spatial stress feature set is used to represent the set of parameters reflecting the gravity and bending stress of the optical cable when it is laid in space.

[0078] Specifically, for the marked vertical cabling section, the data processing system extracts the Z-axis coordinates of its start and end nodes, calculates the absolute value of the coordinate difference between the two as the suspension drop feature, and uses this value to quantify the vertical gravity load on the optical cable itself. Simultaneously, the data processing system calculates the projection trajectory of this section on the XY plane, extracts the angle at the trajectory's turning point as the horizontal rotation angle feature, and uses this feature to quantify the additional stress generated by the cable's bending. The data processing system numerically integrates the suspension drop feature and the horizontal rotation angle feature to generate a spatial force feature set.

[0079] In some embodiments, feature extraction and generation can be achieved in multiple ways: Optionally, the data processing system extracts the Z-coordinates of the start and end points of the vertical wiring section and calculates the difference to obtain the overhang drop feature; the data processing system extracts the coordinates of the horizontal nodes and calculates the vector angle to obtain the horizontal turning angle feature; the data processing system concatenates the two types of features according to a preset data structure to generate a spatial force feature set. Optionally, the data processing system uses a spatial geometry algorithm to fit the pipeline curve and extracts the maximum vertical tangent length as the overhang drop feature; the data processing system extracts the horizontal radius of curvature and converts it into a horizontal turning angle feature; the data processing system normalizes the extracted features and then combines them into a spatial force feature set. It is understood that other methods can also be used to achieve feature extraction and generation, which are not limited here.

[0080] S203. Extract the environmental features, pipeline features, and equipment features of each candidate topology scheme to generate a basic feature set.

[0081] Among them, environmental characteristics refer to external environmental parameters such as temperature, humidity, and vibration within the cabling area; pipeline characteristics refer to attributes such as the material, diameter, and laying method of the optical cable; equipment characteristics are used to represent the model and specifications of passive devices such as splitters and junction boxes; and the basic characteristic set represents the parameter set that integrates the above three types of basic physical attributes.

[0082] Specifically, the data processing system parses the data files of candidate topology schemes, extracting environmental features of the cabling area, pipeline features of the optical cable link, and equipment features of passive nodes. The data processing system standardizes and vectorizes the extracted feature data to eliminate format differences between different data types, and finally combines them to generate a basic feature set.

[0083] In some embodiments, the generation of the basic feature set can be achieved in several ways: Optionally, the data processing system reads the design specification document and extracts environmental parameters using text parsing technology; the data processing system queries the equipment material library to match pipeline and equipment specification parameters; the data processing system converts the extracted parameters into numerical codes and concatenates them to generate the basic feature set. Optionally, the data processing system parses the engineering material list to extract pipeline and equipment attributes; the data processing system combines the GIS system to obtain the historical average environmental monitoring values ​​for the corresponding area; the data processing system performs normalization operations on all attribute data and then merges them into the basic feature set. It is understood that other methods can also be used to generate the basic feature set, and this is not limited here.

[0084] S204. Combine the spatial force feature set and the basic feature set to generate a multidimensional parameter vector set.

[0085] Among them, splicing refers to the operation of combining feature data of different dimensions into a single data structure according to specific rules; a multidimensional parameter vector set refers to a comprehensive vector data set containing multidimensional information such as spatial forces, environment, pipelines and equipment.

[0086] Specifically, the data processing system acquires the generated spatial force feature set and basic feature set, aligning their feature dimensions. Following a preset feature arrangement order, the system appends the spatial force feature set to the end of the basic feature set, performing a matrix or vector concatenation operation. Through this concatenation operation, the system generates a unified set of multidimensional parameter vectors, providing complete data input for subsequent similarity matching.

[0087] In some embodiments, feature set concatenation can be achieved in multiple ways: Optionally, the data processing system reads the spatial force feature set and the basic feature set according to a preset field order; the data processing system executes an array merging instruction to fuse the two into a one-dimensional array; the data processing system outputs this one-dimensional array as a multi-dimensional parameter vector set. Optionally, the data processing system performs principal component extraction on the spatial force feature set and the basic feature set respectively to reduce dimensionality; the data processing system performs weighted fusion and concatenation of the dimensionality-reduced core features; the data processing system encapsulates the fused data to generate a multi-dimensional parameter vector set. It is understood that other methods can also be used to achieve feature set concatenation, which are not limited here.

[0088] In some embodiments, there may be a significant difference in the dimensions between the spatial force feature set and the basic feature set, leading to an imbalance in the weights of some features after stitching in subsequent matching. To address this, the data processing system performs Z-Score standardization on various feature sets before performing the stitching operation to unify the distribution scale of the data across all dimensions.

[0089] In some embodiments, after generating a multidimensional parameter vector set, in order to improve matching efficiency, the data processing system can perform Euclidean distance calculation on the spatial stress feature set in the multidimensional parameter vector set and the historical stress feature set in the as-built network parameter vector to generate a spatial isomorphic distance value; remove as-built network parameter vectors in the historical project database whose spatial isomorphic distance value is greater than a preset distance threshold to generate a candidate as-built network set; and determine the as-built network identifiers in the candidate as-built network set whose global feature matching degree is higher than a preset matching threshold as reference ODN network identifiers that meet the preset similarity conditions.

[0090] Among them, the historical stress feature set represents the characteristic data of the historical completed network in terms of spatial stress; the spatial isomorphism distance value refers to the numerical value that measures the degree of difference between the two networks in the spatial structural stress state; the preset distance threshold represents the maximum allowable spatial structural difference limit; the candidate completed network set refers to the set of historical network data retained after preliminary screening; and the global feature matching degree is used to represent the similarity index calculated by comprehensively considering all feature dimensions.

[0091] Specifically, the data processing system extracts the spatial stress feature set from the multidimensional parameter vector set and calculates the Euclidean distance between it and the historical stress feature set of each as-built network parameter vector in the historical engineering database to obtain a spatial isomorphic distance value. The data processing system compares this distance value with a preset distance threshold, discards historical data with excessively large distances, and generates a candidate as-built network set. Subsequently, within the candidate as-built network set, the data processing system calculates the global feature matching degree using the complete feature vectors, and selects network identifiers with matching degrees higher than the preset matching threshold as reference ODN network identifiers.

[0092] In some embodiments, candidate set generation and identifier determination can be achieved in multiple ways: Optionally, the data processing system executes the Euclidean distance calculation formula to obtain the spatial isomorphic distance value; filters out records greater than a threshold using conditional statements to generate a candidate set; calculates the cosine similarity in the candidate set as the global feature matching degree and selects the identifiers that meet the conditions. Optionally, the data processing system inputs the force characteristics into the distance calculation module and outputs the spatial isomorphic distance value; uses the database's filtering function to remove data that does not meet the conditions to generate a candidate set; and uses a weighted feature matching algorithm to calculate the global feature matching degree and determine the identifier. It is understood that other methods can also be used to achieve candidate set generation and identifier determination, which are not limited here.

[0093] S205. Perform similarity matching operation between the multidimensional parameter vector set and the as-built network parameter vector in the historical project database to determine the reference ODN network identifier that meets the preset similarity conditions.

[0094] Refer to step S102, which will not be repeated here.

[0095] S206. Based on the reference ODN network identifier, extract the actual received optical power telemetry data reported by the corresponding online optical network unit from the network management system, and generate an actual optical power dataset.

[0096] Refer to step S103, which will not be repeated here.

[0097] S207. Read the original design data corresponding to the reference ODN network identifier, calculate the theoretical estimated optical power of each optical network unit according to the preset standard attenuation linear formula, and generate a theoretical optical power dataset.

[0098] Refer to step S104, which will not be repeated here.

[0099] In some embodiments, to more accurately separate systematic biases, before determining the multi-parameter joint compensation coefficients, the data processing system can divide the optical network units in the real optical power dataset and the theoretical optical power dataset into multiple homogeneous optical network unit clusters based on the splitter topology hierarchy identified by the reference ODN network; calculate the bias values ​​between the real received optical power and the theoretically estimated optical power of all optical network units within each homogeneous optical network unit cluster to generate a node bias matrix; and extract the mean bias value of each bias value within the same homogeneous optical network unit cluster in the node bias matrix to generate a common-mode bias component.

[0100] Among them, the optical splitter topology hierarchy represents the hierarchical structure of optical splitter cascading in the ODN network; the common-source optical network unit cluster refers to the set of ONU devices connected under the same final-stage optical splitter; the node deviation matrix is ​​used to represent the data structure that records the difference between the actual and theoretical optical power of each node; the common-mode deviation component represents the average attenuation deviation within the same cluster due to common factors.

[0101] Specifically, the data processing system analyzes the topology of the reference ODN network, identifies the splitter topology hierarchy, and groups optical network units connected to the same splitter port into a cluster of homogeneous optical network units. The system then calculates the difference between the actual received optical power and the theoretically estimated optical power for each optical network unit within each cluster, constructing a node bias matrix. Subsequently, the system calculates the arithmetic mean of the differences in the node bias matrix for each cluster, extracting the common-mode bias component representing the common attenuation characteristics of that cluster.

[0102] In some embodiments, deviation calculation and component extraction can be implemented in multiple ways: Optionally, the data processing system traverses the topology tree to divide the optical network into clusters of homogeneous optical network units; performs subtraction operations to generate a node deviation matrix; and calculates the mean of the matrix along the cluster dimension to generate common-mode deviation components. Optionally, the data processing system uses graph theory algorithms to identify connected subgraphs as clusters of homogeneous optical network units; inputs real and theoretical data into the calculation module to output a node deviation matrix; and applies statistical functions to extract the mean to generate common-mode deviation components. It is understood that other methods can also be used to implement deviation calculation and component extraction, which are not limited here.

[0103] In some embodiments, in order to achieve independent compensation for trunk and branch links, after extracting the common-mode deviation component, the data processing system can subtract the common-mode deviation component from each deviation value in the node deviation matrix to generate the differential-mode deviation component; map the common-mode deviation component to the trunk pipeline characteristics identified by the reference ODN network to generate trunk-level compensation coefficients; map the differential-mode deviation component to the branch pipeline characteristics identified by the reference ODN network to generate branch-level compensation coefficients; and concatenate the trunk-level compensation coefficients and the branch-level compensation coefficients to generate multi-parameter joint compensation coefficients.

[0104] Among them, the differential mode deviation component represents the individualized deviation value remaining after deducting the common deviation from the node deviation; the backbone pipeline characteristic refers to the link attribute connecting the OLT and the backbone splitter; the backbone level compensation coefficient is used to correct the multiplier in the backbone link attenuation calculation; the branch pipeline characteristic represents the link attribute connecting the splitter and the ONU; the branch level compensation coefficient is used to correct the multiplier in the branch link attenuation calculation.

[0105] Specifically, the data processing system subtracts the common-mode deviation component of its cluster from each deviation value in the node deviation matrix to obtain the differential-mode deviation component, which reflects the individual differences of branch links. The system then performs correlation analysis between the common-mode deviation component and the characteristics of the main pipeline to derive the main-level compensation coefficient; similarly, it performs correlation analysis between the differential-mode deviation component and the characteristics of the branch pipeline to derive the branch-level compensation coefficient. Finally, the system concatenates the main-level and branch-level compensation coefficients to construct a complete multi-parameter joint compensation coefficient.

[0106] It should be noted that the distinction between common-mode and differential-mode has a clear physical topological significance: since all ONUs within the same optical network unit cluster share the same backbone link connecting to the OLT and the upstream backbone splitter, the average value of the node deviation within the cluster (i.e., the common-mode deviation component) can accurately characterize the systematic attenuation error caused by common physical factors such as aging of the shared backbone link material or poor construction of the backbone connectors. In contrast, the differential-mode deviation component, after deducting the common-mode deviation, completely eliminates the influence of the backbone link and purely reflects the individual attenuation error caused by the branch drop cables (such as different degrees of bending and different end-pigeon losses) unique to each ONU. Through this error decoupling based on physical topology, the mapping of the compensation coefficients more closely conforms to the actual physical transmission patterns of the ODN network.

[0107] In some embodiments, coefficient generation and splicing can be achieved in multiple ways: Optionally, the data processing system performs matrix subtraction to generate differential mode deviation components; uses a linear regression model to fit the common mode deviation components to the trunk pipeline features to generate trunk-level compensation coefficients; uses the same model to fit the differential mode deviation components to the branch pipeline features to generate branch-level compensation coefficients and splices them. Optionally, the data processing system generates differential mode deviation components by subtracting each item; establishes a mapping table to convert the common mode deviation components into trunk-level compensation coefficients; establishes a mapping table to convert the differential mode deviation components into branch-level compensation coefficients and merges them. It is understood that other methods can also be used to achieve coefficient generation and splicing, which are not limited here. In some embodiments, there may be cases where branch pipeline feature data is missing, resulting in the inability to map differential mode deviation components. In this case, the data processing system can use default branch-level empirical compensation coefficients to replace the missing parts to ensure the smooth progress of the splicing process.

[0108] S208. Determine the multi-parameter joint compensation coefficients based on the real optical power dataset and the theoretical optical power dataset.

[0109] Refer to step S105, which will not be repeated here.

[0110] S209. Based on the standard attenuation linear formula, perform basic attenuation calculations on multiple candidate topology schemes to generate basic attenuation values.

[0111] Refer to step S106, which will not be repeated here.

[0112] S210. Extract the micro-environmental disturbance node data from the environmental features of the candidate topology scheme, add environmental fluctuation margin to the multi-parameter joint compensation coefficient, and generate the compensation coefficient.

[0113] Among them, the micro-environmental disturbance node data represents the location information of unstable factors such as drastic temperature changes or mechanical vibrations on the wiring path; the environmental fluctuation margin refers to the additional attenuation compensation value added to cope with micro-environmental disturbances; and the compensation coefficient refers to the comprehensive multiplier used for the final correction calculation.

[0114] Specifically, the data processing system analyzes the environmental characteristics of candidate topology schemes and locates nodes with micro-environmental disturbances. Based on the type and intensity of the disturbances, the data processing system calculates the corresponding environmental fluctuation margin and superimposes it onto the multi-parameter joint compensation coefficient to generate the final compensation coefficient, thereby enhancing the robustness of the attenuation assessment.

[0115] During the planning phase, the data processing system does not rely on real-time sensor monitoring. Instead, it obtains the aforementioned micro-environmental disturbance node data by accessing prior spatial data from the city's GIS system or building BIM model. Specifically, the data processing system performs spatial collision detection between the 3D cabling paths of candidate topologies and the building structure model. It identifies cabling nodes that are less than a preset safety distance from known heat sources (such as boiler rooms or HVAC duct intersections) or known strong vibration sources (such as elevator shafts or large computer room units), marking them as micro-environmental disturbance nodes. This prediction mechanism based on prior spatial data enables the system to quantify the potential attenuation impact of environmental fluctuations on the long-term operation of optical cables before construction.

[0116] S211. Generate the modified attenuation envelope of the candidate topology scheme based on the basic attenuation value and the compensation coefficient, and use the modified attenuation envelope as the modified attenuation value.

[0117] Here, the modified attenuation envelope represents an interval data structure that includes the possible range of attenuation changes; the lower limit of attenuation refers to the minimum possible attenuation value within the interval; and the upper limit of attenuation refers to the maximum possible attenuation value within the interval.

[0118] Specifically, the data processing system uses the base attenuation value and compensation coefficient to perform interval calculations, determining the lower limit of attenuation under favorable conditions and the upper limit of attenuation under unfavorable conditions, thereby generating a corrected attenuation envelope. The data processing system uses this entire envelope as the corrected attenuation value for subsequent scheme comparisons.

[0119] In some embodiments, in order to select the most stable scheme while meeting the budget, after generating the modified attenuation envelope, the data processing system can compare the upper limit of the attenuation of the modified attenuation envelope of each candidate topology scheme with the optical power budget threshold to generate a subset of candidate schemes that do not exceed the optical power budget threshold; calculate the difference between the upper limit and the lower limit of attenuation of each candidate topology scheme in the subset of candidate schemes to generate the envelope fluctuation amplitude; and select the candidate topology scheme with the smallest envelope fluctuation amplitude from the subset of candidate schemes as the stable optimization scheme of the ODN network to be planned.

[0120] Among them, the candidate scheme subset represents the set of schemes that meet the optical power budget requirements; the envelope fluctuation amplitude refers to the span between the upper and lower limits of attenuation; the stable optimization scheme is used to represent the preferred scheme with the smallest attenuation change under various environmental fluctuations.

[0121] Specifically, the data processing system extracts the upper limit of attenuation for each candidate topology scheme to correct the attenuation envelope, determines whether it is less than or equal to the optical power budget threshold, and assigns schemes that meet the condition to a subset of candidate schemes. For each scheme in the subset, the data processing system calculates the difference between the upper and lower limit of attenuation to obtain the envelope fluctuation amplitude. The data processing system compares the envelope fluctuation amplitudes of each scheme and selects the scheme with the smallest value as the stable optimization scheme for the ODN network to be planned, ensuring the reliability of network operation.

[0122] It should be noted that in practical engineering applications, to avoid selecting a suboptimal solution with extremely small fluctuations but an overall attenuation value approaching the critical budget, the data processing system, when executing the above selection logic, will first set a safety margin threshold in the subset of candidate solutions (for example, requiring the upper limit of attenuation to be at least 2dB lower than the optical power budget threshold). Only when this safety margin threshold is met will the system further compare the envelope fluctuation amplitude, thereby selecting a stable and optimized solution that has sufficient optical power redundancy and can maintain minimal attenuation changes under complex environmental fluctuations.

[0123] In some embodiments, the selection of a stable optimization scheme can be achieved in several ways: Optionally, the data processing system performs a comparison operation to generate a subset of candidate schemes; performs a subtraction operation to generate the envelope fluctuation amplitude; sorts the envelope fluctuation amplitudes and selects the scheme corresponding to the minimum value. Optionally, the data processing system uses a filtering function to select schemes that meet the budget to form a subset; calls the calculation module to output the envelope fluctuation amplitude; and applies an optimization algorithm to locate the scheme with the minimum envelope fluctuation amplitude in the subset. It is understood that other methods can also be used to select a stable optimization scheme, which are not limited here.

[0124] In some embodiments, multiple schemes within a subset of candidate schemes may exhibit the same envelope fluctuation amplitude. In response, the data processing system can further compare the lower limits of these schemes' attenuation values ​​and select the scheme with the smaller lower limit as the stable optimization scheme.

[0125] S212. Determine the candidate topology scheme that meets the optical power budget threshold and minimizes the modified attenuation value, and use it as the attenuation optimization scheme for the ODN network to be planned.

[0126] Refer to step S107, which will not be repeated here.

[0127] In this embodiment, by extracting multi-dimensional features and combining them with real historical data to generate multi-parameter joint compensation coefficients to correct theoretical attenuation, it can effectively cope with unpredictable losses in complex engineering environments. It effectively solves the problem in the prior art that relying solely on fixed standard empirical coefficients leads to actual optical power deviating from the budget, thereby improving the accuracy of scheme optimization for actual optical distribution networks.

[0128] The data processing system in the embodiments of this invention is described below from a hardware processing perspective. Please refer to [link / reference needed]. Figure 3 This is a schematic diagram of the physical device structure of a data processing system in an embodiment of this application.

[0129] It should be noted that, Figure 3 The structure of the data processing system shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of the present invention.

[0130] like Figure 3 As shown, the data processing system includes a CPU 301, which can perform various appropriate actions and processes according to a program stored in ROM 302 or a program loaded from storage section 308 into RAM 303, such as executing the methods described in the above embodiments. RAM 303 also stores various programs and data required for system operation. The CPU 301, ROM 302, and RAM 303 are interconnected via bus 304. I / O interface 305 is also connected to bus 304.

[0131] The following components are connected to I / O interface 305: input section 306 including audio input devices, push-button switches, etc.; output section 307 including a liquid crystal display (LCD) and audio output devices, indicator lights, etc.; storage section 308 including a hard disk, etc.; and communication section 309 including a network interface card such as a LAN (Local Area Network) card, modem, etc. Communication section 309 performs communication processing via a network such as the Internet. Drive 310 is also connected to I / O interface 305 as needed. Removable media 311, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., are installed on drive 310 as needed so that computer programs read from them can be installed into storage section 308 as needed.

[0132] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing computer programs for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 309, and / or installed from removable medium 311. When the computer program is executed by CPU 301, it performs the various functions defined in the present invention.

[0133] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. Each block in a flowchart or block diagram may represent a module, program segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those shown in the drawings.

[0134] Specifically, the data processing system of this embodiment includes a processor and a memory. The memory stores a computer program. When the computer program is executed by the processor, it implements the ODN attenuation optimization method based on multi-parameter matching provided in the above embodiment.

[0135] In another aspect, the present invention also provides a computer-readable storage medium, which may be included in the data processing system described in the above embodiments; or it may exist independently and not assembled into the data processing system. The storage medium carries one or more computer programs that, when executed by a processor of the data processing system, cause the data processing system to implement the ODN attenuation optimization method based on multi-parameter matching provided in the above embodiments.

[0136] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit it. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

[0137] As used in the above embodiments, depending on the context, the term "when..." can be interpreted as meaning "if...", "after...", "in response to determining...", or "in response to detecting...". Similarly, depending on the context, the phrase "when determining..." or "if (the stated condition or event) is interpreted as meaning "if determining...", "in response to determining...", "when (the stated condition or event) is detected", or "in response to detecting (the stated condition or event)".

Claims

1. A method for optimizing ODN attenuation based on multi-parameter matching, characterized in that, Applied to a data processing system, the method includes: Read multiple candidate topology schemes of the ODN network to be planned, extract environmental features, pipeline features and equipment features from each candidate topology scheme, and generate a multi-dimensional parameter vector set; The multidimensional parameter vector set is matched with the as-built network parameter vector in the historical engineering database to determine the reference ODN network identifier that meets the preset similarity conditions. Based on the reference ODN network identifier, extract the actual received optical power telemetry data reported by the corresponding online optical network unit from the network management system to generate an actual optical power dataset. Read the original design data corresponding to the reference ODN network identifier, calculate the theoretical estimated optical power of each optical network unit according to the preset standard attenuation linear formula, and generate a theoretical optical power dataset. Based on the real optical power dataset and the theoretical optical power dataset, determine the multi-parameter joint compensation coefficients; Based on the standard attenuation linear formula, the basic attenuation calculation is performed on the multiple candidate topology schemes respectively, and the corrected attenuation value of the candidate topology scheme is generated by combining the multi-parameter joint compensation coefficient. The candidate topology that satisfies the optical power budget threshold and minimizes the modified attenuation value is determined as the attenuation optimization scheme for the ODN network to be planned.

2. The method according to claim 1, characterized in that, The step of reading multiple candidate topology schemes of the ODN network to be planned, extracting environmental features, pipeline features, and equipment features from each candidate topology scheme, and generating a multi-dimensional parameter vector set specifically includes: Read multiple candidate topology schemes of the ODN network to be planned, and combine the three-dimensional spatial orientation data of each candidate topology scheme to determine the vertical cabling segment; Extract the sag drop features and horizontal corner features of the vertical wiring section to generate a spatial force feature set; Extract the environmental features, pipeline features, and equipment features of each of the candidate topology schemes to generate a basic feature set; The spatial force feature set and the basic feature set are concatenated to generate the multidimensional parameter vector set.

3. The method according to claim 2, characterized in that, After the step of concatenating the spatial force feature set and the basic feature set to generate the multidimensional parameter vector set, the method further includes: The spatial force feature set in the multidimensional parameter vector set and the historical force feature set in the as-built network parameter vector are calculated using Euclidean distance to generate a spatial isomorphic distance value. Remove the as-built network parameter vectors from the historical project database whose spatial isomorphic distance values ​​are greater than a preset distance threshold, and generate a candidate as-built network set; In the candidate completed network set, the completed network identifiers with a global feature matching degree higher than a preset matching threshold are determined as reference ODN network identifiers that meet the preset similarity conditions.

4. The method according to claim 1, characterized in that, Before the step of determining the multi-parameter joint compensation coefficients based on the real optical power dataset and the theoretical optical power dataset, the method further includes: Based on the splitter topology hierarchy identified by the reference ODN network, the optical network units in the real optical power dataset and the theoretical optical power dataset are divided into multiple homogeneous optical network unit clusters. Calculate the deviation between the actual received optical power and the theoretically estimated optical power of all optical network units within each of the aforementioned co-origin optical network unit clusters, and generate a node deviation matrix; The mean deviation of each deviation value within the same optical network unit cluster from the node deviation matrix is ​​extracted to generate the common-mode deviation component.

5. The method according to claim 4, characterized in that, After the step of extracting the mean deviation of each deviation value within the same optical network cell cluster in the node deviation matrix and generating the common-mode deviation component, the method further includes: Subtract the common mode deviation component from each deviation value in the node deviation matrix to generate the differential mode deviation component; The common-mode deviation component is mapped to the backbone pipeline characteristics of the reference ODN network identifier to generate backbone-level compensation coefficients. The differential mode deviation component is mapped to the branch pipeline characteristics identified by the reference ODN network to generate branch-level compensation coefficients; The main-level compensation coefficient and the branch-level compensation coefficient are concatenated to generate the multi-parameter joint compensation coefficient.

6. The method according to claim 1, characterized in that, The step of performing basic attenuation calculations on the multiple candidate topology schemes based on the standard attenuation linear formula, and generating corrected attenuation values ​​for the candidate topology schemes by combining the multi-parameter joint compensation coefficients, specifically includes: Based on the standard attenuation linear formula, basic attenuation calculations are performed on the multiple candidate topology schemes to generate basic attenuation values; Extract micro-environmental disturbance node data from the environmental features of the candidate topology scheme, add environmental fluctuation margin to the multi-parameter joint compensation coefficient, and generate compensation coefficient; The modified attenuation envelope of the candidate topology scheme is generated based on the basic attenuation value and the compensation coefficient, and the modified attenuation envelope is used as the modified attenuation value; the modified attenuation envelope includes a lower attenuation limit and an upper attenuation limit.

7. The method according to claim 6, characterized in that, After the step of generating a modified attenuation envelope of the candidate topology scheme based on the base attenuation value and the compensation coefficient, and using the modified attenuation envelope as the modified attenuation value, the method further includes: The upper limit of the attenuation of the modified attenuation envelope of each candidate topology scheme is compared with the optical power budget threshold to generate a subset of candidate schemes that do not exceed the optical power budget threshold. Calculate the difference between the upper and lower limits of attenuation for each candidate topology scheme in the subset of candidate schemes to generate the envelope fluctuation amplitude; The candidate topology scheme with the smallest envelope fluctuation amplitude is selected from the subset of candidate schemes and used as the stable optimization scheme for the ODN network to be planned.

8. A data processing system, characterized in that, The data processing system includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code including computer instructions, and the one or more processors call the computer instructions to cause the data processing system to perform the method as described in any one of claims 1-7.

9. A computer-readable storage medium comprising instructions, characterized in that, When the instructions are executed on the data processing system, the data processing system performs the method as described in any one of claims 1-7.

10. A computer program product, characterized in that, When the computer program product is run on a data processing system, the data processing system performs the method as described in any one of claims 1-7.

Images