CN115278407B

CN115278407B - Method and device for determining topological structure of passive optical network

Info

Publication number: CN115278407B
Application number: CN202110485308.XA
Authority: CN
Inventors: 宁列; 郭健; 刘伟
Original assignee: China Mobile Communications Group Co Ltd; China Mobile Group Shanxi Co Ltd
Current assignee: China Mobile Communications Group Co Ltd; China Mobile Group Shanxi Co Ltd
Priority date: 2021-04-30
Filing date: 2021-04-30
Publication date: 2024-07-05
Anticipated expiration: 2041-04-30
Also published as: CN115278407A

Abstract

The embodiment of the application provides a topology structure determining method and device of a passive optical network, wherein the method comprises the following steps: performing feature extraction on original time sequence data of a Passive Optical Network (PON) port to obtain time sequence feature data corresponding to the PON port; the time sequence feature data comprises global feature data, local feature data and statistical feature data; determining a co-ordination matrix based on the global feature data, the local feature data and the statistical feature data; hierarchical clustering is carried out on the co-ordination matrix, and a classification result is obtained; and determining the topological structure corresponding to the PON port based on the classification result. By the method and the device for determining the topological structure of the passive optical network, the optical power data of the receiving end and the transmitting end are effectively considered, the original time sequence data of the PON ports of the passive optical network are subjected to feature extraction, a co-ordination matrix is constructed, the determination of the PON topological structure of the passive optical network is realized, the robustness is higher, and the compatibility is stronger.

Description

Method and device for determining topological structure of passive optical network

Technical Field

The present application relates to the technical field of passive optical networks, and in particular, to a method and an apparatus for determining a topology structure of a passive optical network.

Background

The passive optical network (passive optical network, PON) adopts a tree topology structure to connect an optical line terminal (optical LINE TERMINAL, OLT) and an optical network unit (Optical Network Unit, ONU) of a user, and the connection between the OLT and the ONU forms a tree topology structure through an optical splitter. The optical splitter can enable the OLT to mount more ONUs in a cascading mode, so that the capacity of the network is improved, and the construction cost of the PON network is reduced. The cascade of optical splitters is typically limited to a two-stage cascade. At present, the connection relationship between the optical splitter and the ONU is not clear for various reasons. First, early designs and construction lack the emphasis on splitters, resulting in loss of connection structure information of ONUs and OLTs in a part of the network. Second, since the optical splitter is a passive device, the connection condition cannot be obtained through the downlink instruction. Thirdly, the connection relation between the optical splitter and the ONU in the current system mainly depends on manual input and is influenced by the accuracy of the manual input. Fourth, PON networks often switch connection relationships between ONUs and splitters during operation and maintenance, so as to cope with new demands of users on the networks.

However, the existing method generally considers unilateral optical power data or early warning, fault and other information at different moments, adopts a single clustering algorithm to determine topology information of an optical distribution network (Optical Distribution Network, ODN) containing passive optical devices, and the obtained result has poor robustness and weak compatibility.

Therefore, the problem in the prior art is solved by effectively considering the fluctuation condition of the optical power data of the receiving end and the transmitting end, which is an important topic to be solved in the industry.

Disclosure of Invention

Aiming at the problems existing in the prior art, the embodiment of the application provides a topology structure determining method and device of a passive optical network.

In a first aspect, an embodiment of the present application provides a method for determining a topology structure of a passive optical network, including: performing feature extraction on original time sequence data of a Passive Optical Network (PON) port to obtain time sequence feature data corresponding to the PON port; the time sequence feature data comprises global feature data, local feature data and statistical feature data;

determining a co-ordination matrix CM based on the global feature data, the local feature data and the statistical feature data;

hierarchical clustering is carried out on the co-ordination matrix CM to obtain a classification result; wherein the classification result satisfies the following condition: the classification number and the estimated splitting ratio of the PON port meet the preset production requirement, and the classification number is at least two types and at most not more than half of the number of the Optical Network Units (ONU) under the PON port;

And determining a topological structure corresponding to the PON port based on the classification result.

Optionally, the feature extracting the original time sequence data of the PON port of the passive optical network to obtain time sequence feature data corresponding to the PON port includes:

performing data extraction, wherein the data comprises: ONU receives the optical power and transmits the optical power, PON port receives the optical power and transmits the optical power;

Carrying out data reconstruction based on the data extraction result, wherein the method specifically comprises the following steps: and calculating link loss data according to the ONU receiving optical power and the PON port transmitting optical power, and/or calculating link loss data according to the ONU transmitting optical power and the PON port receiving optical power.

Optionally, the global feature data includes a piecewise linear PLR feature and a PLRW feature, the PLR feature is a piecewise linear representation of the PON port original time-series data, and the PLRW feature is a feature constructed by adding an arithmetic mean of the original data over the PLR feature;

the local characteristic data comprises one or a combination of abnormal values, turning points and mutation points;

the statistical characteristic data includes one or a combination of maximum, minimum, polar error, average, divergence, skewness, quartile range, variance, and standard deviation.

Optionally, the determining the co-relation matrix CM based on the global feature data, the local feature data and the statistical feature data includes:

performing hierarchical-based cluster analysis on the global feature data, performing partition-based cluster analysis on the statistical feature data, and initializing a co-ordination matrix CM based on a cluster result of the cluster analysis;

constructing a distance matrix D based on the local feature data, and initializing the distance matrix D;

And calculating the Hadamard product of the co-ordinated matrix CM and the distance matrix D after zero removal to be used as a new co-ordinated matrix CM.

Optionally, the estimated splitting ratio of the PON port is determined according to the following steps:

extracting original time sequence data at a specific time point, and calculating the average loss of the beam splitter;

acquiring experience data, comparing loss data corresponding to the experience data with loss data of actual data under different light splitting ratios, and estimating the light splitting ratio of the light splitter; wherein the average loss of the optical splitter is the average value of the loss of the optical splitter of all ONU of the PON port;

The splitter loss is calculated as link loss data according to ONU received optical power and PON port transmitted optical power, and/or,

And calculating link loss data according to the ONU transmitting optical power and the PON port receiving optical power.

Optionally, the method further comprises: the co-ordination matrix CM is modified based on the history topology.

Optionally, the correcting the co-ordination matrix CM based on the history topology includes:

Step 1, initializing an all-zero matrix CM _i as a co-ordination matrix during the ith data acquisition; wherein i is any time of data acquisition, and i is more than or equal to 2;

step 2, loading a historical topological structure generated by the i-1 th calculation during the i-th data acquisition; the CM _i is modified according to the following rules: if the classification of the a-th ONU and the b-th ONU are consistent, making CM (a, b) =X, wherein X is a first parameter, and the value range of X is [1,2,4];

And 3, loading a history co-relation matrix from the ith time to the jth time during the ith time data acquisition to the ith time to the 2 th time calculation, and marking the history co-relation matrix as CM _i-j,CM_i-j+1,……,CM_i-2. And their data acquisition times T _i-j,T_i-j+1,……,T_i-2, j are thresholds;

Step 4, summing the history co-ordination matrix calculated from the i-j time to the i-2 time in the step 3 and the corrected co-ordination matrix according to the first parameter in the step 2, multiplying the sum by a second parameter P and a third parameter Q, and finally adding an initialized all-zero matrix CM _i to obtain the corrected co-ordination matrix; wherein Q is a second parameter used for describing forgetting of data, and 0<Q is less than or equal to 1; p is a third parameter used for measuring the influence degree of the correction term, and 0<P is less than or equal to 1.

In a second aspect, an embodiment of the present application provides a topology determining device of a passive optical network, including a memory, a transceiver, and a processor;

A memory for storing a computer program; a transceiver for transceiving data under control of the processor; a processor for executing the computer program in the memory and implementing the steps of:

Performing feature extraction on original time sequence data of a Passive Optical Network (PON) port to obtain time sequence feature data corresponding to the PON port; the time sequence feature data comprises global feature data, local feature data and statistical feature data;

In a third aspect, an embodiment of the present application provides a topology determining apparatus of a passive optical network, where the apparatus includes:

The feature extraction module is used for carrying out feature extraction on original time sequence data of a Passive Optical Network (PON) port to obtain time sequence feature data corresponding to the PON port; the time sequence feature data comprises global feature data, local feature data and statistical feature data;

A matrix determining module, configured to determine a co-ordination matrix CM based on the global feature data, the local feature data, and the statistical feature data;

The classification module is used for carrying out hierarchical clustering on the co-ordination matrix CM to obtain a classification result; wherein the classification result satisfies the following condition: the classification number and the estimated splitting ratio of the PON port meet the preset production requirement, and the classification number is at least two types and at most not more than half of the number of the Optical Network Units (ONU) under the PON port;

And the structure output module is used for determining the topological structure corresponding to the PON port based on the classification result.

In a fourth aspect, an embodiment of the present application provides a processor-readable storage medium storing a computer program for causing the processor to perform the steps of the topology determining method of the passive optical network according to the first aspect as described above.

According to the method and the device for determining the topological structure of the passive optical network, the original time sequence data of the PON ports of the passive optical network are subjected to feature extraction by effectively considering the optical power data of the receiving end and the transmitting end, and the co-ordination matrix is constructed, so that the determination of the PON topological structure of the passive optical network is realized, the robustness is higher, and the compatibility is stronger.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are some embodiments of the present application, and other drawings can be obtained according to these drawings without inventive effort for a person skilled in the art.

Fig. 1 is a diagram showing a specific location of link loss generation in a PON according to an embodiment of the present application;

fig. 2 is a schematic flow chart of a topology determining method of a passive optical network according to an embodiment of the present application;

fig. 3 is a second schematic flow chart of a topology determining method of a passive optical network according to an embodiment of the present application;

Fig. 4 is a third schematic flow chart of a topology determining method of a passive optical network according to an embodiment of the present application;

fig. 5 is a schematic flow chart of a topology determining method of a passive optical network according to an embodiment of the present application;

Fig. 6 is a schematic diagram of a complete flow of a topology determining method of a passive optical network according to an embodiment of the present application;

Fig. 7 is a flow chart of a topology determining method of a passive optical network according to another embodiment of the present application;

fig. 8 is a schematic structural diagram of a topology determining device of a passive optical network according to an embodiment of the present application;

fig. 9 is a schematic structural diagram of a topology determining apparatus of a passive optical network according to an embodiment of the present application.

Detailed Description

In the embodiment of the application, the term "and/or" describes the association relation of the association objects, which means that three relations can exist, for example, a and/or B can be expressed as follows: a exists alone, A and B exist together, and B exists alone. The character "/" generally indicates that the context-dependent object is an "or" relationship.

The term "plurality" in embodiments of the present application means two or more, and other adjectives are similar.

The following description of the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments, but not all embodiments of the present application. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.

At present, the topology structure of the ONU and the optical splitter is determined by the following method: 1) The PPPoE method can generally determine the connection relationship between the OLT and the ONU, and this connection relationship is accurate, but the connection structure between the ONU and the optical splitter is not clear. 2) The topology of the ONUs and splitters is typically maintained manually by service personnel. 3) And analyzing the topological structures of the ONU and the optical splitter by adopting a big data means. And determining the topological structures of the optical splitter and the ONU by analyzing the curve of the received optical power and taking the similar fluctuation curve as a standard.

In the prior art, a manual structure maintenance method is generally adopted, the accuracy is low, and the change of the topological structure cannot be effectively conducted. Schemes employing large data analysis methods typically require the storage of large amounts of data, which creates a data storage burden for network operators. With the increase of PON network construction time, some devices are aged to different degrees, and empirical data cannot guide analysis of network structure well. The prior art depends on the fact that the received optical power data can not well reflect the fluctuation condition of the optical power data of the transmitting end; the problem of data missing of ONU under different online conditions cannot be solved; the problem of topology change during ONU change cannot be adaptively solved.

The present application aims to solve the above technical problems. The optical splitter device in the PON network can be managed more conveniently and efficiently by a network operator.

The core of the application is developed around the construction, updating and correction processes of the co-ordination matrix, and the following targets are realized by combining the method for generating the distance matrix by using the co-ordination matrix: on-line learning of topology structure, different on-line ONU, and change of ONU,

Fig. 1 is a diagram illustrating a specific location of link loss generation in a PON according to an embodiment of the present application. As shown in fig. 1, link loss is largely divided into connection point loss, splitter loss, and line loss. The line loss is related to the line length and the type of optical fiber used, the connection point loss is related to the connection mode of the connection point, meanwhile, the loss of the connection point is also influenced by the ageing degree of the equipment, and the loss of the optical splitter is related to the optical splitting ratio of the optical splitter. The link loss data contains connection structure information of the link. The topology structure can be effectively obtained through analysis of the link loss data. The application aims at analyzing the link loss data and reasoning the topological structures of the ONU and the optical splitter.

Fig. 2 is a schematic flow chart of a topology determining method of a passive optical network according to an embodiment of the present application. As shown in fig. 2, the topology structure determining method of the passive optical network includes:

Step 201, extracting features of original time sequence data of a passive optical network PON port, obtaining time sequence feature data corresponding to the PON port; the time sequence feature data comprises global feature data, local feature data and statistical feature data;

Specifically, the passive optical network PON is mainly composed of an optical line terminal OLT, an optical distribution network ODN containing a passive optical device, and an optical network unit ONU at a user end, where all ONUs are hooked on PON ports of the OLT, that is, there is a certain topological relationship between the ONUs and the PON ports; in the specific implementation, the ONU can be hung on a PON port through a primary beam splitter and/or a secondary beam splitter, and a certain topological relation exists between the ONU and the secondary beam splitter. From the above, it is clear that one PON port and the corresponding ONU have a one-to-one or one-to-many relationship, and that one optical line terminal OLT has a plurality of PON ports.

Firstly, acquiring original optical power time sequence data of a first acquisition period of all ONUs under a first PON port, and then carrying out feature extraction on the original optical power time sequence data to obtain time sequence feature data corresponding to the PON port, wherein the time sequence feature data comprises global feature data, local feature data and statistical feature data;

Step 202, determining a co-ordination matrix CM based on the global feature data, the local feature data and the statistical feature data;

Specifically, the co-ordination matrix CM is defined as an n-dimensional square matrix, where n is determined by the number of ONUs attached under the PON port. And determining the co-ordination matrix CM according to global feature data, local feature data and statistical feature data obtained after feature extraction of the original optical power time sequence data.

Step 203, hierarchical clustering is carried out on the co-ordination matrix CM to obtain a classification result; wherein the classification result satisfies the following condition: the classification number and the estimated splitting ratio of the PON port meet the preset production requirement, and the classification number is at least two types and at most not more than half of the number of the Optical Network Units (ONU) under the PON port;

Specifically, hierarchical clustering is performed on the co-ordination matrix CM, and common algorithms include AGNES, BIRCH, CURE, ROCK, CHAMELEON, DIANA and the like.

The process of the AGNES algorithm is as follows:

① Each object is treated as an initial cluster.

② The nearest two clusters are found from the nearest data point in the two clusters.

③ The two clusters are merged to generate a new set of clusters.

④ ② And ③ are repeated until the defined number of clusters is reached. And outputting a cluster tree.

The cluster tree output here is broken down top-down into several categories of cluster trees. Decomposing the cluster tree needs to meet the following conditions, and a classification result can be generated: 1. the classification number and the estimated splitting ratio of the PON port meet the preset production conditions. 2. The generation classification number is at least 2 classes, and at most, the classification number is not more than half of the number of ONUs under the first PON port.

The preset production conditions are actual production conditions, namely, the obtained classification number meets the actual application conditions, and the common classification number is as follows: 1:2,1:4,1:8,1:16,1:32,1:64. for example, if the classification result of the actual output is 1: and 3, obviously failing to meet the actual production condition, or outputting a classification result of the first-stage optical splitter of 1:4, the output final classification result is 1:64, the classification result of the secondary beam splitter is necessarily 1:16 are all the practical production conditions.

The AGNES algorithm adopted by hierarchical clustering of the co-ordination matrix CM is simple to realize, and a cluster tree can be obtained more rapidly and efficiently.

Step 204, determining a topology structure corresponding to the PON port based on the classification result.

The cluster tree determined in step 203 is used for determining the optimal classification of the classification result by adopting a corresponding algorithm, and various algorithms are commonly used.

According to the method for determining the topological structure of the passive optical network, provided by the embodiment of the application, the original time sequence data of the PON port of the passive optical network is subjected to feature extraction by effectively considering all the optical power data of the PON port, so that the co-ordination matrix CM is constructed, the determination of the PON topological structure of the passive optical network is realized, the robustness is higher, and the compatibility is stronger.

On the basis of the foregoing embodiment, optionally, feature extraction is performed on original time sequence data of a PON port of a passive optical network to obtain time sequence feature data corresponding to the PON port, including: performing data extraction, wherein the data comprises: ONU receives the optical power and transmits the optical power, PON port receives the optical power and transmits the optical power;

Specifically, the data extraction process may adopt a method of recording the optical power data to be used and extracting the optical power data from the historical database in real time, which is not limited in the present application. The extraction of data includes the following: ONU received optical power and transmitted optical power, PON port received optical power and transmitted optical power.

Optionally, the original data may further include distance data between the PON port and the ONU.

After the data extraction, the corresponding link loss data, namely, the data reconstruction, needs to be obtained, the link loss data is calculated according to the ONU received optical power and the PON port transmitted optical power, and/or the link loss data is calculated according to the ONU transmitted optical power and the PON port received optical power.

When the data is rebuilt, the bidirectional link loss data can be considered, so that the loss or instability of unidirectional data is avoided, and the obtained link loss data is more effective.

According to the method for determining the topological structure of the passive optical network, the original time sequence data of the PON ports of the passive optical network are subjected to feature extraction by effectively considering the optical power data of the receiving end and the transmitting end, the co-ordination matrix CM is constructed, the determination of the PON topological structure of the passive optical network is realized, the robustness is higher, and the compatibility is stronger.

Fig. 3 is a second schematic flow chart of a topology determining method of a passive optical network according to an embodiment of the present application. As shown in fig. 3, in the method for determining the topology structure of the passive optical network, the global feature data includes a piecewise linear PLR feature and a PLRW feature, the PLR feature is a piecewise linear representation of the PON port original time-series data, and the PLRW feature is a feature constructed by adding an arithmetic mean of the original data above the PLR feature;

Specifically, global feature data extraction is performed on the original data of the first PON port, where the global features include PLR features and PLRW features. The PLR feature is a piecewise linear representation (PIECEWISE LINEAR presentation, PLR) of the original time series data of the first PON port, and the PLRW feature (PIECEWISE LINEAR presentation WITH WEIGHT, PLRW) is a feature constructed by adding an arithmetic mean of the original data over the PLR feature. The global feature data is data in a time-series format.

Extracting local feature data from the original data of the first PON port, wherein the feature extraction comprises the following aspects: outliers, turning points, mutation points. The local feature data may consist of one of these outliers, turning points and abrupt points or a combination of these three types of data. The outlier is defined as an absolute value of a difference between the loss data and the average data exceeding a threshold. The abrupt change point is defined as a point when the absolute value of the difference value between the loss data at the second moment and the loss data at the first moment exceeds a threshold value. The turning point is defined as a mutation point at the same time as the first time and the second time, and the mutation directions are different, and the second time is the turning point. The remaining time is calculated in the same manner as described above. The mutation directions herein mainly refer to different directions of data change, for example, the mutation directions of the corresponding mutation points at the first moment are increased, the mutation directions of the corresponding mutation points at the second moment are decreased, and then the mutation directions of the corresponding turning points are different.

Carrying out statistical feature data extraction on the original data of the first PON port, wherein the feature extraction comprises the following aspects: maximum, minimum, pole difference, average, divergence, bias, quartile range, variance, standard deviation. The statistical characteristic data may be composed of one of these maximum, minimum, polar error, average, divergence, skewness, quartile range, variance and standard deviation, or a combination of these nine types of data.

Resampling the global feature data and the local feature data. The resampling is defined as extracting data at every interval k time from each ONU time sequence characteristic data of the first PON port according to a preset resampling ratio k.

And extracting time sequence characteristic data from each ONU of the first PON port. For a first PON port comprising n ONUs, m resampling is performed, which comprises m groups of PLR features, m groups PLRW features, m groups of local feature data, 1 group of statistical feature data, i.e. for each ONU of the first PON port, 3m+1 groups of data are included.

The feature data of each first PON port further comprises: distance data between the OLT and the ONU and unique identification codes of the ONU. The distance data can be acquired through the OLT or manually input data, and the acquisition mode is not limited by the application. The unique identification code of the ONU may be a logical code or an actual device code, which is not limited in this application.

According to the topology structure determining method of the passive optical network, the original time series data of the PON ports of the passive optical network are subjected to feature extraction and resampling operation by effectively considering the optical power data of the receiving end and the transmitting end, more complex data acquisition granularity can be compatible, the data quantity can be adjusted, and the compatibility is stronger when the minimum data requirement is met. Meanwhile, preconditions are provided for subsequent analysis to obtain the PON topological structure of the passive optical network.

Fig. 4 is a third schematic flow chart of a topology determining method of a passive optical network according to an embodiment of the present application. As shown in fig. 4, the method for determining a passive optical network topology, which determines a co-ordination matrix CM based on the global feature data, the local feature data and the statistical feature data, includes:

step 401, performing hierarchical-based cluster analysis on the global feature data, performing partition-based cluster analysis on the statistical feature data, and initializing a co-ordination matrix CM based on a cluster result of the cluster analysis;

Specifically, the global feature data comprises PLR feature data and PLRW feature data, m resampling is carried out on the global feature data to obtain m groups of PLR features and m groups of PLRW features, hierarchical clustering analysis of time sequences is carried out on the resampled global features, and hierarchical clustering of the time sequences is carried out on the resampled global features to form 2m base clusters.

The clustering of global features is performed using a hierarchical based clustering algorithm, specifically, using a balanced iteration protocol and clustering (Balanced Iterative Reducing and Clustering Using Hierarchies, BIRCH) algorithm using a hierarchical approach.

The BIRCH algorithm requires the use of pattern distances to compute clusters of new samples with leaf nodes.

The mode distance is calculated as follows.

Where S1 and S2 are two time series, m _1i is the ith of the PLR signature of S1, and k is the length of S1 and S2.

The statistical features are obtained by resampling the statistical feature data once, the clustering analysis based on division is carried out, the clustering analysis based on division is adopted aiming at the statistical features, and a K-Means algorithm is specifically adopted to form 1 base clustering device.

Traversing the clustering results of the 2m+1 base clusters, aiming at counting the times of dividing the two ONU into the same class, filling the corresponding position of the co-ordination matrix CM, and realizing the initialization of the co-ordination matrix CM.

Step 402, constructing a distance matrix D based on the local feature data, and initializing the distance matrix D;

Specifically, the local feature data is subjected to comparative analysis, so that the initialization of the distance matrix D is realized. The method comprises the following specific steps:

① And constructing a distance matrix D. Assuming that the first PON port contains n ONUs, D is an n×n matrix, and all elements of the initialization matrix are 0.

② D (i, j) is the distance of the ith ONU from the jth ONU. The distance is defined as the number of times that the time series data of two ONUs simultaneously appear an outlier, a mutation point, and a turning point.

Step 403, calculating the Hadamard product of the co-ordinated matrix CM and the distance matrix D after zeroing, and using the Hadamard product as a new co-ordinated matrix CM.

Specifically, 1 is added to each element of the co-ordination matrix CM in step 401, and 1 is added to each element of the distance matrix D in step 402, so as to implement the operation of zeroing.

And calculating the Hadamard product of the co-ordinated matrix CM and the distance matrix D after zero removal, namely multiplying two matrices with the same dimension, and taking the product as a new co-ordinated matrix CM.

The topological structure determining method of the passive optical network provided by the embodiment of the application performs clustering integration based on hierarchical clustering analysis of global feature data and partition clustering analysis based on statistical feature data. The results of the multiple basic clusters can be comprehensively considered, and the random disturbance can be effectively filtered. The data characteristics are processed from multiple angles by designing the basic clusters of different parameters and algorithms, and the robustness of the corresponding result is high.

Fig. 5 is a flowchart of a topology determining method of a passive optical network according to an embodiment of the present application. As shown in fig. 5, the estimated splitting ratio of the PON port is determined according to the following steps:

Step 501, extracting original time sequence data at a specific time point;

specifically, data at a specific time point is extracted, and if the first PON port contains n ONUs, n numbers are collected in total.

Step 502, calculating average loss of the optical splitter;

Specifically, the splitter loss is calculated according to ONU received optical power and PON port transmitted optical power, and/or calculated according to ONU transmitted optical power and PON port received optical power.

The average loss of the optical splitter is defined as the average value of the loss of the optical splitter of all ONUs of the first PON port. And summing the link loss data corresponding to all the ONUs hung under all the first PON ports, dividing the sum by the number of all the ONUs, and obtaining the average loss of the optical splitter.

Step 503, acquiring experience data, comparing loss data corresponding to the experience data with loss data of actual data under different light splitting ratios, and estimating the light splitting ratio of the light splitter; wherein the average loss of the optical splitter is the average value of the loss of the optical splitter of all ONU of the PON port; the optical splitter loss is calculated according to the ONU received optical power and the PON port transmitted optical power, and/or calculated according to the ONU transmitted optical power and the PON port received optical power.

Specifically, the acquisition of the empirical data is performed according to the following steps:

① Loss data at a number of different spectral ratios are collected.

② The data were classified according to the spectral ratio, and the arithmetic mean and standard deviation of each group were calculated.

③ The normal distribution is fitted using the arithmetic mean and standard deviation calculated from the historical data for different spectral ratios. And R normal curve probability density equations are formed for R kinds of spectroscopies. Wherein the R kinds of spectroscope schemes include: 1:2,1:4,1:8,1:16,1:32,1:64.

④ Solving the intersection point of two normal curve probability density equations with similar spectral ratios between the two average values. The intersection point is the basis for judging the beam splitting ratio. Here, the close spectral ratios are compared by two with a close numerical ratio, for example, a 1:2 to 1:4 comparison, a 1:4 to 1:8 comparison, and a 1:8 to 1:16 comparison.

According to the method for determining the topological structure of the passive optical network, the average loss of the optical splitter at the specific moment is compared with the experience data, the estimated optical splitting ratio of the PON port is determined, and the method is combined with the co-ordination matrix CM, so that the determination of the topological structure of the passive optical network PON is achieved, the robustness is higher, and the compatibility is stronger.

The following describes in detail the entire flow of the topology determining method of the passive optical network according to an embodiment of the present application with reference to fig. 6.

Step 601, constructing data characteristics to obtain time sequence characteristic data;

Step 602, estimating a beam splitting ratio;

and calculating the average loss of the optical splitter at a specific time to obtain loss data of actual data, acquiring empirical data, comparing the loss data corresponding to the empirical data with the loss data of the actual data under a similar optical splitting ratio, and estimating the optical splitting ratio of the optical splitter.

Step 603, constructing a clustering device and a co-ordination matrix;

and constructing a clustering device of global feature data for m groups of PLR features and m groups PLRW features based on hierarchical clustering analysis, constructing a clustering device of statistical feature data for 1 group of statistical feature data based on partitioned clustering analysis, traversing the two clustering devices, and filling the times of counting the times that every two ONU are partitioned into the same class into the positions corresponding to the co-ordination matrix CM to finish the construction of the co-ordination matrix CM.

Step 604, hierarchical clustering is carried out on the co-ordination matrix;

Constructing a distance matrix according to the local characteristic data, and performing Hadamard product based on the distance matrix after zero removal and the consensus relation matrix to obtain a new consensus relation matrix, and performing hierarchical clustering on the consensus relation matrix by adopting an AGNES algorithm to obtain a corresponding cluster tree;

step 605, determining a topology;

And determining optimal classification of the cluster tree by adopting Calinski-Harabaz indexes, and determining a topological structure corresponding to the PON port based on an optimal classification result.

On the basis of the above embodiment, optionally, the method further includes: the co-ordination matrix CM is modified based on the history topology.

Specifically, in the data acquisition process, there may be a situation that part of ONUs are not online, a history topology structure is adopted to initialize the co-ordination matrix CM, and the co-ordination matrix CM at a specific moment is corrected based on history topology information at a moment before the specific moment and according to the history co-ordination matrix CM. For example, the local feature corrects the CM matrix, and the correction method is as follows: counting the times of occurrence of abnormal values, turning points and abrupt points of the first ONU and the second ONU at the same time, and multiplying the values by CM (1, 2). The same applies to other ONUs of the first PON port. And combining the history co-ordination matrix in a period of time before a specific moment, and correcting the co-ordination matrix by considering the corresponding influence degree.

The topology structure determining method of the passive optical network provided by the embodiment of the application effectively aims at the condition that ONU is not on line at the same time. When data acquisition is carried out, the ONU is difficult to ensure to be online simultaneously, so that partial ONU has the condition of data acquisition failure, the model and the history structure are subjected to persistence operation, and the co-ordination matrix CM is initialized through the history structure, so that analysis of the ONU which is online at different times is realized.

Fig. 7 is a flowchart of a topology determining method of a passive optical network according to another embodiment of the present application. As shown in fig. 7, the method for determining a passive optical network topology, where the correcting the co-ordination matrix CM based on the history topology includes:

Step 701, initializing a co-ordination matrix CM by adopting a history topological structure. The history topology is a set of class labels. The result of the initialization of the co-ordination matrix CM is determined by the history topology.

The method specifically comprises the following steps:

Step 7011, initializing an all-zero matrix CM _i as a co-ordination matrix CM during the ith data acquisition; wherein i is any time of data acquisition, and i is more than or equal to 2;

Step 7012, loading a history topological structure generated by the i-1 th calculation during the i-th data acquisition; the CM _i is modified according to the following rules: if the classification of the a-th ONU and the b-th ONU are consistent, making CM (a, b) =X, wherein X is a first parameter, and the value range of X is [1,2,4];

Specifically, during data acquisition, the history topological structure obtained by the local side at the moment before the acquisition moment is adopted to correct the co-coordination relation matrix CM at the acquisition moment, so that the influence caused by different on-line ONU can be corrected, and the value range of the first parameter X is one of 1,2 and 4. For example, loading the history topology generated by the i-1 th calculation; if the classifications of the a-th ONU and the b-th ONU agree, CM (a, b) =2 is made.

Step 7013, loading a history co-relation matrix from the ith time to the jth time when the ith time is acquired to the ith time and the ith time are calculated, wherein the history co-relation matrix is recorded as CM _i-j,CM_i-j+1,……,CM_i-2, and the data acquisition time T _i-j,T_i-j+1,……,T_i-2 and the j are thresholds;

Specifically, j is a threshold value, which is used to describe an event of data storage. j is equal to the number of days data storage times the number of data acquisitions per day. In this embodiment, j is 7, and data acquisition is performed once a day in this embodiment.

Step 7014, summing the history co-ordination matrices generated by the i-j-th to i-2-th calculations in step 7013, and the corrected co-ordination matrix according to the first parameter in step 7012, multiplying the sum by the second parameter P and the third parameter Q, and finally adding an initialized all-zero matrix CM _i to obtain the corrected co-ordination matrix; wherein Q is a second parameter used for describing forgetting of data, and 0<Q is less than or equal to 1; p is a third parameter used for measuring the influence degree of the correction term, and 0<P is less than or equal to 1.

Specifically, the method of correcting CMi by the history co-ordination matrix is as follows.

Where Q is a second parameter used to describe forgetting of the data. Wherein P is a third parameter for measuring the influence degree of the correction term. Wherein the method comprises the steps ofRepresenting rounding down in days.

In this embodiment, the value of P is 1, 0<P is equal to or less than 1, which represents that the influence degree of the correction term is the highest, the value of Q is 0.5, 0<Q is equal to or less than 1, which represents that the forgetting degree of the description data is moderate.

Other steps in this embodiment are the same as those in the previous embodiment, and will not be described here again.

And initializing a co-ordination matrix CM by using the historical topological structure, and combining the co-ordination matrix by taking the topological structure as a base clustering device result. And extracting time sequence characteristic data, constructing other base clusters, combining the co-ordination matrixes, and hierarchical clustering the co-ordination matrixes. And obtaining the updated topological structure according to the historical data.

The method for determining the topological structure of the passive optical network provided by the embodiment of the application can effectively cope with the condition that the ONU is not on line at the same time, perform the persistence operation on the model and the historical structure, initialize the co-ordination relation matrix through the historical structure and realize the analysis on the ONU which is not on line at the same time. But also can effectively process the change of OUN. The new addition, removal and change of the ONU occur at the time, and effective response is realized through a forgetting strategy of the co-ordination matrix, wherein the forgetting strategy refers to: the last estimated structure used for initializing the co-ordination matrix is forgotten for a longer structure, so that the history structure is considered during the change, and the history structure does not influence the new structure estimation process by excessive weight.

In summary, the present application differs from the prior art scheme:

1. The data feature-based clustering method and the partition-based clustering method are adopted for different purposes, and in order to provide robustness of the model, clustering integration is performed. The clustering integration method is a clustering mode with strong robustness, and can comprehensively consider the results of various basic clusters, thereby effectively realizing the filtration of random disturbance. The clustering integration algorithm is not limited to the method adopted by the base clustering device, and the processing of the data features from multiple angles can be realized by designing the base clustering device with different parameters and algorithms. The existing method adopts a single clustering algorithm, and the robustness of the result is poorer than that of the method.

2. The data acquisition granularity can be compatible with more complex data acquisition granularity through resampling operation, the data volume can be adjusted simultaneously, and the compatibility is stronger when the minimum data requirement is met again. And the link loss data calculation is adopted, so that the fluctuation condition of the optical power data of the receiving end and the transmitting end is effectively considered.

3. The method effectively aims at the situation that the ONU is not online at the same time. When data acquisition is carried out, the ONU is difficult to ensure to be online simultaneously, so that partial ONU has the condition of data acquisition failure, the model and the history structure are subjected to persistence operation, and the history structure is used for initializing a co-ordination matrix so as to analyze the ONU which is online at different times.

4. Changes to the OUN are handled efficiently. The new addition, removal and change of the ONU occur at the time, and effective response is realized through a forgetting strategy of the co-ordination matrix, wherein the forgetting strategy refers to: the last estimated structure used for initializing the co-ordination matrix is forgotten for a longer structure, so that the history structure is considered during the change, and the history structure does not influence the new structure estimation process by excessive weight.

The application is divided into a cold start process and an online update process in terms of implementation. The cold start process is as follows: acquiring ONU original optical power data, and calculating link loss according to the received optical power and the transmitted optical power; extracting various features from the link loss, constructing different base clusters for different features, and adopting a co-ordination matrix to perform clustering integration; and constructing a distance matrix by using the co-ordination matrix, and clustering the distance matrix based on layers to obtain a topological structure, wherein the specific content is as in the first embodiment. The online updating process comprises the following steps: in order to process the different online conditions of the ONU and the changing conditions of the ONU, data acquisition is carried out for a plurality of times, and the topological structure is corrected through historical data; correcting the co-ordination matrix by adopting a history topological structure; correcting the new co-ordination matrix by considering the history co-ordination matrix; and constructing a distance matrix by using the new consensus relation matrix, and clustering the distance matrix based on layers to obtain a topological structure, wherein the specific content is as in the second embodiment.

Fig. 8 is a schematic structural diagram of a topology determining device of a passive optical network according to an embodiment of the present application. As shown in fig. 8, the device includes a memory 820, a transceiver 810, and a processor 800; wherein processor 800 and memory 820 may also be physically separate.

A memory 820 for storing a computer program; a transceiver 810 for transceiving data under the control of the processor 800.

In particular, the transceiver 810 is used to receive and transmit data under the control of the processor 800.

Wherein in fig. 8, a bus architecture may comprise any number of interconnected buses and bridges, and in particular, one or more processors represented by processor 800 and various circuits of memory represented by memory 820, linked together. The bus architecture may also link together various other circuits such as peripheral devices, voltage regulators, power management circuits, etc., all as are well known in the art and, therefore, will not be described further herein. The bus interface provides an interface. The transceiver 810 may be a number of elements, i.e., including a transmitter and a receiver, providing a means for communicating with various other apparatus over a transmission medium, including wireless channels, wired channels, optical cables, etc. The processor 800 is responsible for managing the bus architecture and general processing, and the memory 820 may store data used by the processor 800 in performing operations.

Processor 800 may be CPU, ASIC, FPGA or a CPLD, and the processor may also employ a multi-core architecture.

Processor 800, by invoking a computer program stored in memory 820, is configured to perform any of the methods provided by embodiments of the present application in accordance with the obtained executable instructions, for example: performing feature extraction on original time sequence data of a Passive Optical Network (PON) port to obtain time sequence feature data corresponding to the PON port; the time sequence feature data comprises global feature data, local feature data and statistical feature data;

Determining a co-ordination matrix based on the global feature data, the local feature data and the statistical feature data;

Hierarchical clustering is carried out on the co-ordination matrix, and a classification result is obtained; wherein the classification result satisfies the following condition: the classification number and the estimated splitting ratio of the PON port meet the preset production requirement, and the classification number is at least two types and at most not more than half of the number of the Optical Network Units (ONU) under the PON port;

It should be noted that, the topology structure determining device for a passive optical network provided by the embodiment of the present application can implement all the method steps implemented by the method embodiment and achieve the same technical effects, and detailed descriptions of the same parts and beneficial effects as those of the method embodiment in the embodiment are omitted.

Fig. 9 is a schematic structural diagram of a topology determining apparatus of a passive optical network according to an embodiment of the present application. As shown in fig. 9, the apparatus includes:

The feature extraction module 901 is configured to perform feature extraction on original time sequence data of a PON port of a passive optical network, and obtain time sequence feature data corresponding to the PON port; the time sequence feature data comprises global feature data, local feature data and statistical feature data;

A matrix determining module 902, configured to determine a co-ordination matrix based on the global feature data, the local feature data, and the statistical feature data;

The classification module 903 is configured to perform hierarchical clustering on the co-ordination matrix to obtain a classification result; wherein the classification result satisfies the following condition: the classification number and the estimated splitting ratio of the PON port meet the preset production requirement, and the classification number is at least two types and at most not more than half of the number of the Optical Network Units (ONU) under the PON port;

and the structure output module 904 is configured to determine a topology structure corresponding to the PON port based on the classification result.

It should be noted that, in the embodiment of the present application, the division of the units is schematic, which is merely a logic function division, and other division manners may be implemented in actual practice. In addition, each functional unit in the embodiments of the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.

The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a processor-readable storage medium. Based on such understanding, the technical solution of the present application may be embodied in essence or a part contributing to the prior art or all or part of the technical solution in the form of a software product stored in a storage medium, including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) or a processor (processor) to execute all or part of the steps of the method according to the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), a magnetic disk, or an optical disk, or other various media capable of storing program codes.

It should be noted that, the above device provided in the embodiment of the present application can implement all the method steps implemented in the method embodiment and achieve the same technical effects, and detailed descriptions of the same parts and beneficial effects as those in the method embodiment in this embodiment are omitted.

In another aspect, an embodiment of the present application further provides a processor readable storage medium, where a computer program is stored, where the computer program is configured to cause the processor to execute the method for determining a topology of a passive optical network provided in the foregoing embodiments, where the method includes: performing feature extraction on original time sequence data of a Passive Optical Network (PON) port to obtain time sequence feature data corresponding to the PON port; the time sequence feature data comprises global feature data, local feature data and statistical feature data;

The processor-readable storage medium may be any available medium or data storage device that can be accessed by a processor, including, but not limited to, magnetic storage (e.g., floppy disks, hard disks, magnetic tape, magneto-optical disks (MOs), etc.), optical storage (e.g., CD, DVD, BD, HVD, etc.), and semiconductor storage (e.g., ROM, EPROM, EEPROM, non-volatile storage (NAND FLASH), solid State Disk (SSD)), etc.

It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, magnetic disk storage, optical storage, and the like) having computer-usable program code embodied therein.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer-executable instructions. These computer-executable instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These processor-executable instructions may also be stored in a processor-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the processor-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These processor-executable instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

It will be apparent to those skilled in the art that various modifications and variations can be made to the present application without departing from the spirit or scope of the application. Thus, it is intended that the present application also include such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.

Claims

1. A method for determining a topology of a passive optical network, comprising:

Hierarchical clustering is carried out on the co-ordination matrix CM to obtain a classification result; wherein the classification result satisfies the following condition: the classification number and the estimated splitting ratio of the PON port meet the preset production requirement, and the classification number is at least two types and at most not more than half of the number of the Optical Network Units (ONU) under the PON port; the light splitting ratio is determined based on loss data corresponding to experience data and loss data of actual data, the preset production requirement refers to an actual production condition, the actual production condition is a condition that the obtained classification number meets actual application, and the classification number comprises at least one of the following: 1:2,1:4,1:8,1:16,1:32,1:64;

2. The method for determining the topology of the PON according to claim 1, wherein performing feature extraction on the original time-series data of the PON port of the PON to obtain time-series feature data corresponding to the PON port comprises:

Carrying out data reconstruction based on the data extraction result, wherein the method specifically comprises the following steps: the link loss data is calculated from the ONU received optical power and the PON port transmitted optical power, and/or,

3. The method of claim 1, wherein the global feature data comprises piecewise linear PLR features, which are piecewise linear representations of the PON port raw time-series data, and PLRW features, and wherein the PLRW features are features constructed by adding an arithmetic mean of raw data over the PLR features;

4. The method for determining the topology of a passive optical network according to claim 1, wherein the determining the co-ordination matrix CM based on the global feature data, the local feature data and the statistical feature data comprises:

5. The method for determining the topology of a passive optical network according to claim 1, wherein the estimated split ratio of PON ports is determined according to the steps of:

6. The method for determining the topology of a passive optical network according to any one of claims 1 to 5, further comprising:

the co-ordination matrix CM is modified based on the history topology.

7. The method for determining the topology of a passive optical network according to claim 6, wherein said modifying said co-ordination matrix CM based on a historical topology comprises:

Step 1, initializing an all-zero matrix CM _i as a co-ordination matrix CM during the ith data acquisition; wherein i is any time of data acquisition, and i is more than or equal to 2;

Step 3, loading a history co-relation matrix from the ith time to the jth time when the ith time is acquired to the ith time and the ith time when the history co-relation matrix is calculated for the ith time to the ith time and the ith time is recorded as CM _i-j,CM_i-j+1,……,CM_i-2, and the data acquisition time T _i-j,T_i-j+1,……,T_i-2 and the j of the history co-relation matrix are used as thresholds;

8. The topology structure determining device for passive optical network includes memory, transceiver and processor;

9. A device for determining a topology of a passive optical network, the device comprising:

The classification module is used for carrying out hierarchical clustering on the co-ordination matrix CM to obtain a classification result; wherein the classification result satisfies the following condition: the classification number and the estimated splitting ratio of the PON port meet the preset production requirement, and the classification number is at least two types and at most not more than half of the number of the Optical Network Units (ONU) under the PON port; the light splitting ratio is determined based on loss data corresponding to experience data and loss data of actual data, the preset production requirement refers to an actual production condition, the actual production condition is a condition that the obtained classification number meets actual application, and the classification number comprises at least one of the following: 1:2,1:4,1:8,1:16,1:32,1:64;

10. A processor-readable storage medium, characterized in that the processor-readable storage medium stores a computer program for causing the processor to execute the method of determining the topology of the passive optical network according to any one of claims 1 to 7.