Cross-country submarine cable and autonomous system mapping large-scale inference method and device

Abstract

This application relates to the field of network measurement technology, and in particular to a method and apparatus for large-scale inference of the mapping between transnational submarine optical cables and autonomous systems. The method includes: acquiring the global router-level topology and the global transnational submarine optical cable topology; inputting the global router-level topology and the global transnational submarine optical cable topology into a trained classifier, outputting the mapping relationship between routers and transnational submarine optical cables; converting the router's IP address into an autonomous system number, and generating the mapping relationship between the transnational submarine optical cable and the autonomous system based on the autonomous system number and the mapping relationship between routers and transnational submarine optical cables. This solves the problems of excessively small measurement range and high detection overhead in related inference methods for mapping transnational submarine optical cables to autonomous systems.

Description

Technical Field

[0001] This application relates to the field of network measurement technology, and in particular to a method and apparatus for large-scale inference of mapping between transnational submarine optical cables and autonomous systems. Background Technology

[0002] As one of the world's most fundamental physical communication infrastructures, transnational submarine optical cables carry over 95% of global cross-border traffic. Due to the unique nature of their deployment locations, these cables are frequently susceptible to failures caused by natural disasters or human factors, leading to sudden increases in inter-regional access delays and even complete disconnection of certain regions from the outside world. Therefore, measuring and analyzing the role of transnational submarine optical cables in global network communication is of great significance in preventing serious consequences from cable failures. Further analysis of this role requires first obtaining the mapping between transnational submarine optical cables and autonomous systems (AS / RS).

[0003] Existing methods for mapping transnational submarine optical cables to autonomous systems primarily involve deploying probe points at both ends of a transnational submarine optical cable, allowing these probe points to perform traceroute probes against each other, and inferring the routers at both ends of the transnational submarine optical cable based on changes in the geographical location of IP addresses (Internet Protocol addresses) along the traceroute path. This method suffers from limitations in measurement range and excessive measurement overhead. Summary of the Invention

[0004] This application provides a method and apparatus for large-scale inference of mapping between transnational submarine optical cables and autonomous systems, in order to solve the problems of small measurement range and high detection overhead in inference methods for mapping between transnational submarine optical cables and autonomous systems in related technologies.

[0005] The first aspect of this application provides a method for large-scale inference of the mapping between transnational submarine optical cables and autonomous systems, comprising the following steps: obtaining a global router-level topology and a global transnational submarine optical cable topology; inputting the global router-level topology and the global transnational submarine optical cable topology into a trained classifier, and outputting the mapping relationship between routers and transnational submarine optical cables; converting the IP address of the router into an autonomous system number, and generating the mapping relationship between the transnational submarine optical cable and the autonomous system based on the autonomous system number and the mapping relationship between the router and the transnational submarine optical cable.

[0006] Optionally, before converting the router's IP address to an Autonomous System number, the method further includes: identifying the distance between landing stations of transnational submarine optical cables within the same geographical location; if the distance between the landing stations is less than or equal to a first preset distance, then determining that the transnational submarine optical cable corresponding to the landing station is an adjacent transnational submarine optical cable, and determining the mapping relationship between the router and the transnational submarine optical cable based on the link changes in the global router-level topology before and after the construction of the adjacent transnational submarine optical cables.

[0007] Optionally, the training process of the classifier includes: obtaining positive samples that meet preset screening conditions, wherein the preset screening conditions include that there is one and only one transnational submarine optical cable between the two countries and there is no land between the two countries; using the positive samples to train the positive sample learning classifier until the training stopping condition is met, and obtaining the trained classifier.

[0008] Optionally, before training the positive sample learning classifier using the positive samples, the method further includes: removing samples from the positive samples where the distance between the router and the login station is greater than a second preset distance.

[0009] A second aspect of this application provides a large-scale inference apparatus for mapping transnational submarine optical cables to autonomous systems, comprising: an acquisition module for acquiring a global router-level topology and a global transnational submarine optical cable topology; an input module for inputting the global router-level topology and the global transnational submarine optical cable topology into a trained classifier and outputting a mapping relationship between routers and transnational submarine optical cables; and a generation module for converting router IP addresses into autonomous system numbers and generating a mapping relationship between the transnational submarine optical cables and the autonomous systems based on the autonomous system numbers and the mapping relationship between routers and transnational submarine optical cables.

[0010] Optionally, it further includes: an identification module, used to identify the distance between landing stations of transnational submarine optical cables within the same geographical location before converting the router's IP address to an autonomous system number; if the distance between the landing stations is less than or equal to a first preset distance, then the transnational submarine optical cable corresponding to the landing station is determined to be an adjacent transnational submarine optical cable, and the mapping relationship between the router and the transnational submarine optical cable is determined based on the link changes in the global router-level topology before and after the construction of the adjacent transnational submarine optical cables.

[0011] Optionally, the training process of the classifier includes: obtaining positive samples that meet preset screening conditions, wherein the preset screening conditions include that there is one and only one transnational submarine optical cable between the two countries and there is no land between the two countries; using the positive samples to train the positive sample learning classifier until the training stopping condition is met, and obtaining the trained classifier.

[0012] Optionally, it further includes: a rejection module, used to reject samples in the positive samples whose distance between the router and the login station is greater than a second preset distance before training the positive sample learning classifier using the positive samples.

[0013] A third aspect of this application provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the large-scale inference method for mapping transnational submarine optical cables to autonomous systems as described in the above embodiments.

[0014] A fourth aspect of this application provides a computer-readable storage medium having a computer program stored thereon, which is executed by a processor to implement the large-scale inference method for mapping transnational submarine optical cables to autonomous systems as described in the above embodiments.

[0015] Therefore, this application has at least the following beneficial effects:

[0016] This application's embodiments can infer from publicly available global router-level topology, expanding the measurement range; they can utilize positive sample learning to classify samples that are difficult to judge using strict rules; they reduce the interference of the MPLS (Multi-Protocol Label Switching) protocol on the inference results based on distance features; by comparing the changes in global router-level topology before and after the construction of submarine optical cables, they distinguish submarine optical cables with very similar landing stations and routes, further improving measurement accuracy; they can expand the measurement range while avoiding detection overhead, and through this mapping, they combine the actual transnational submarine optical cables transmitting data (physical communication links) with the autonomous systems that route the data forwarding direction (logical communication links), providing data support for a better understanding of the role of transnational submarine optical cables in the global network. Therefore, this solves the technical problems of excessively small measurement range and high detection overhead in related technologies' inference methods that map transnational submarine optical cables to autonomous systems.

[0017] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0018] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0019] Figure 1 This is a flowchart of a large-scale inference method for mapping transnational submarine optical cables to autonomous systems provided in an embodiment of this application;

[0020] Figure 2This is a schematic diagram illustrating the excessive distance between the router and the login station caused by an MPLS tunnel, according to an embodiment of this application.

[0021] Figure 3 This is an overall architecture diagram of the large-scale inference method for mapping transnational submarine optical cables and autonomous systems provided in the embodiments of this application;

[0022] Figure 4 This is an example diagram of a large-scale inference device for mapping transnational submarine optical cables to autonomous systems provided according to embodiments of this application;

[0023] Figure 5 This is a schematic diagram of the structure of an electronic device provided according to an embodiment of this application. Detailed Implementation

[0024] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0025] The following describes a method and apparatus for large-scale inference of mapping between transnational submarine optical cables and autonomous systems, based on embodiments of this application, with reference to the accompanying drawings. Addressing the issues mentioned in the background art regarding current inference methods for mapping transnational submarine optical cables and autonomous systems, which primarily involve deploying probe points at both ends of a transnational submarine optical cable and having these probe points perform traceroute probes against each other, inferring the routers at both ends of the transnational submarine optical cable based on changes in the geographical location of IP addresses along the traceroute path, this application provides a method for large-scale inference of mapping between transnational submarine optical cables and autonomous systems. In this method, the global router-level topology (ITDK) disclosed by CAIDA is matched geographically with the global transnational submarine optical cable topology. For transnational submarine optical cables with very similar landing station locations and routes, differentiation is achieved by comparing changes in the ITDK before and after construction. This solves the problems of insufficient measurement range and high probe overhead in related inference methods for mapping between transnational submarine optical cables and autonomous systems.

[0026] Specifically, Figure 1 This is a flowchart illustrating a large-scale inference method for mapping transnational submarine optical cables to autonomous systems, provided in an embodiment of this application.

[0027] like Figure 1 As shown, the large-scale inference method for mapping transnational submarine optical cables to autonomous systems includes the following steps:

[0028] In step S101, the global router-level topology and the global transnational submarine optical cable topology are obtained.

[0029] Among them, the global router-level topology (ITDK) was made public by CAIDA after traceroute. Specifically, CAIDA periodically uses the global detection platform Ark to traceroute all / 24 prefixes and then uses alias resolution tools such as MIDAR, iffinder, and kapar to generate the global router-level topology, which is also the largest publicly available global router-level topology. The global transnational submarine optical cable topology was made public by Telegeography.

[0030] It should be noted that the embodiments of this application use ITDK and its historical data for inference, which can both avoid introducing a large amount of probe traffic into the global network and affect data transmission efficiency, and expand the measurement range to the greatest extent.

[0031] In step S102, the global router-level topology and the global transnational submarine optical cable topology are input into the trained classifier, and the mapping relationship between the router and the transnational submarine optical cable is output.

[0032] It is understood that, in the embodiments of this application, a classifier trained with global router-level topology probes and global transnational submarine optical cable topologies can be used to output the mapping relationship between routers and transnational submarine optical cables.

[0033] The classifier training process includes: obtaining positive samples that meet preset screening conditions, including that there is one and only one transnational submarine optical cable between the two countries and there is no land between the two countries; training the positive sample learning classifier using the positive samples until the training stopping condition is met, and obtaining the trained classifier.

[0034] It should be noted that this application needs to distinguish between terrestrial optical cables and submarine optical cables: when there is land between the two landing stations of a transnational submarine optical cable, it is difficult to determine whether a link in the router-level topology is a terrestrial optical cable or a submarine optical cable. Therefore, it is necessary to screen positive samples that meet the conditions.

[0035] Specifically, the classifier training process includes: First, selecting positive samples based on strict screening criteria: there must be one and only one transnational submarine optical cable between two countries, and the two countries must not share land. If there is also a direct link between these two countries in the ITDK (Internet Data Center), we consider this link to pass through the corresponding unique transnational submarine optical cable. These links and the mapping relationship between the transnational submarine optical cable are considered positive samples. Second, the positive sample learning classifier is trained using these positive samples until the training stopping condition is met, resulting in a fully trained classifier. In addition, unlabeled samples that cannot be classified according to the above two conditions are classified.

[0036] The process of training a positive sample learning classifier using positive samples is as follows: Input positive samples and unlabeled samples; perform N iterations, in each iteration taking all positive samples and a portion of unlabeled samples to train a binary classifier, and calculating the probability of the remaining unlabeled samples being judged as positive based on the trained binary classifier; for all unlabeled samples, if the average probability of being judged as positive after N iterations is greater than 0.5, then it is judged as positive by the positive sample classifier, otherwise it is judged as negative.

[0037] In this embodiment of the application, before training the positive sample learning classifier using positive samples, the method further includes: removing samples from the positive samples in which the distance between the router and the login station is greater than a second preset distance.

[0038] The second preset distance can be set according to specific circumstances, and is not limited here.

[0039] It is understandable that, before training the positive sample learning classifier using positive samples in this embodiment, it is necessary to identify samples in the positive samples where the distance between the router and the login station is greater than a second preset distance. This means removing samples where the distance between the router and the login station is too large. The reason is that in the pipeline mode, the entire MPLS tunnel appears as two directly connected routers in ITDK, which can affect the measurement results. Therefore, the selected positive samples are further cleaned, removing samples where the distance between the router and the login station is too large. The cleaned positive samples are then used to train the classifier to reduce the impact of the MPLS protocol on the inference results. Specifically, the MPLS tunnel causes the distance between the router and the login station to be too large... Figure 2 As shown.

[0040] In step S103, the router's IP address is converted to an Autonomous System number, and a mapping relationship between the transnational submarine optical cable and the Autonomous System is generated based on the Autonomous System number, the mapping relationship between the router and the transnational submarine optical cable.

[0041] It is understood that, in the embodiments of this application, the mapping relationship between the router and the transnational submarine optical cable can be obtained based on step S102, the IP address of the router can be converted into an autonomous system number, and the mapping relationship between the transnational submarine optical cable and the autonomous system can be generated based on the autonomous system number and the mapping relationship between the router and the transnational submarine optical cable.

[0042] In this embodiment of the application, before converting the router's IP address to an autonomous system number, the method further includes: identifying the distance between landing stations of transnational submarine optical cables within the same geographical location; if the distance between landing stations is less than or equal to a first preset distance, then determining that the transnational submarine optical cable corresponding to the landing station is an adjacent transnational submarine optical cable, and determining the mapping relationship between the router and the transnational submarine optical cable based on the link changes in the global router-level topology before and after the construction of the adjacent transnational submarine optical cables.

[0043] The first preset distance can be set according to specific circumstances, and there is no limitation on it.

[0044] It should be noted that since the feature used for inference is only the distance between the router and the landing station, when the landing stations of two transnational submarine optical cables are geographically close, a link in the ITDK may be assigned to both transnational submarine optical cables simultaneously. This application's embodiments distinguish this situation by comparing the changes in the ITDK topology before and after the construction of a certain submarine optical cable. If a router link appears only after the construction of a certain submarine optical cable, this router link is assigned to the newly constructed submarine optical cable, thereby obtaining the mapping relationship between the router and the transnational submarine optical cable.

[0045] Understandably, only when the distance between the landing stations is less than or equal to the first preset distance is it necessary to compare historical data to determine which transnational submarine optical cable the router link is assigned to. If the distance between the landing stations is greater than the first preset distance, a portion of positive samples are first selected based on the above strict conditions, and then the positive samples are used to train a classifier to classify other samples that cannot be selected by strict conditions, so as to obtain the mapping relationship between the router and the transnational submarine optical cable.

[0046] In summary, the large-scale inference method for mapping transnational submarine optical cables to autonomous systems proposed in this application mainly includes four aspects:

[0047] (1) Expanding the measurement range while avoiding probe overhead: CAIDA periodically uses the global probe platform Ark to perform traceroute probes on all / 24 prefixes, and then uses alias resolution tools such as MIDAR, iffinder, and kapar to generate a global router-level topology (ITDK), which is also the largest publicly available global router-level topology. Using the ITDK and its historical data for inference can avoid introducing a large amount of probe traffic into the global network, which would affect data transmission efficiency, while maximizing the expansion of the measurement range.

[0048] (2) Distinguishing between terrestrial and submarine optical cables: When there is land between two landing stations of a transnational submarine optical cable, it is difficult to determine whether a link in the router-level topology follows a terrestrial or submarine optical cable. To address this issue, positive samples are first selected based on strict conditions: ① There is one and only one transnational submarine optical cable between the two countries, and ② There is no land between the two countries. If there is a link directly connecting these two countries in the ITDK, we consider this link to have passed through the corresponding unique transnational submarine optical cable. Then, a positive sample learning classifier is trained based on the selected positive samples to classify other unlabeled samples that cannot be classified according to the above two conditions.

[0049] (3) Reduce the impact of MPLS protocol on inference results: Since the entire MPLS tunnel in pipeline mode is represented by two directly connected routers in ITDK, it has a certain impact on the measurement results. Therefore, the positive samples selected in (2) were further cleaned (samples with too large a distance between the router and the login station were removed), and the cleaned positive samples were used to train the classifier to reduce the impact of MPLS protocol on inference results.

[0050] (4) Differentiating Adjacent Transnational Submarine Cables: Since the only feature used for inference is the distance between the router and the landing station, when the landing stations of two transnational submarine cables are geographically close, a link in the ITDK may be assigned to both transnational submarine cables simultaneously. This situation is differentiated by comparing the changes in the ITDK topology before and after the construction of a certain submarine cable. If a router link appears after the construction of a certain submarine cable, this router link is assigned to the newly constructed submarine cable, thereby obtaining the mapping relationship between the router and the transnational submarine cable.

[0051] Specifically, the large-scale inference method for mapping transnational submarine optical cables to autonomous systems proposed in this application has the following overall architecture diagram: Figure 3 As shown, it mainly includes: an input module for inputting the global router-level topology and the global transnational submarine optical cable topology; a distance calculation module for calculating the distance from the router to the landing station; a machine learning module for training the classifier; an adjacent submarine cable differentiation module for differentiating adjacent transnational submarine optical cables; and an output module that, based on BGP (Border Gateway Protocol) announcement data, converts the router's IP address into an Autonomous System number and outputs the mapping relationship between transnational submarine optical cables and Autonomous Systems.

[0052] It should be noted that this application infers the mapping relationship between transnational submarine optical cables and autonomous systems based on the latest ITDK topology published by CAIDA, the global transnational submarine optical cable topology published by Telegeography, and the BGP declared data published by RIS and Routeviews. Experimental results show that the measurement range of this application can cover 77.6% of the global transnational submarine optical cables.

[0053] The large-scale inference method for mapping transnational submarine optical cables to autonomous systems proposed in this application can expand the measurement range by inferring based on publicly available global router-level topology; it can classify samples that are difficult to judge by strict rules by using positive sample learning; it reduces the interference of MPLS protocol on the inference results based on distance features; by comparing the changes in global router-level topology before and after the construction of submarine optical cables, it distinguishes submarine optical cables with very similar landing stations and routes, further improving measurement accuracy; it can expand the measurement range while avoiding detection overhead; and it combines transnational submarine optical cables that actually transmit data with autonomous systems that route data forwarding directions through this mapping, providing data support for a better understanding of the role of transnational submarine optical cables in the global network.

[0054] Next, referring to the accompanying drawings, a large-scale inference apparatus for mapping transnational submarine optical cables and autonomous systems according to embodiments of this application is described.

[0055] Figure 4 This is a block diagram of a large-scale inference device for mapping transnational submarine optical cables and autonomous systems according to an embodiment of this application.

[0056] like Figure 4 As shown, the transnational submarine optical cable and autonomous system mapping large-scale inference device 10 includes: an acquisition module 100, an input module 200 and a generation module 300.

[0057] The acquisition module 100 is used to acquire the global router-level topology and the global transnational submarine optical cable topology; the input module 200 is used to input the global router-level topology and the global transnational submarine optical cable topology into the trained classifier and output the mapping relationship between routers and transnational submarine optical cables; the generation module 300 is used to convert the IP address of the router into an autonomous system number and generate the mapping relationship between transnational submarine optical cables and autonomous systems based on the autonomous system number and the mapping relationship between routers and transnational submarine optical cables.

[0058] In this embodiment of the application, the device 10 further includes an identification module.

[0059] The identification module is used to identify the distance between landing stations of transnational submarine optical cables within the same geographical area before converting the router's IP address to an autonomous system number. If the distance between landing stations is less than or equal to a first preset distance, the transnational submarine optical cable corresponding to the landing station is determined to be an adjacent transnational submarine optical cable. The mapping relationship between the router and the transnational submarine optical cable is determined based on the changes in links in the global router-level topology before and after the construction of adjacent transnational submarine optical cables.

[0060] In this embodiment of the application, the training process of the classifier includes: obtaining positive samples that meet preset screening conditions, wherein the preset screening conditions include that there is one and only one transnational submarine optical cable between the two countries and there is no land between the two countries; using the positive samples to train the positive sample learning classifier until the training stop condition is met, and obtaining the trained classifier.

[0061] In this embodiment of the application, the apparatus 10 further includes a rejection module.

[0062] The elimination module is used to eliminate samples in the positive samples whose distance between the router and the login station is greater than a second preset distance before training the positive sample learning classifier using positive samples.

[0063] It should be noted that the foregoing explanation of the embodiment of the large-scale inference method for mapping transnational submarine optical cables and autonomous systems also applies to the large-scale inference device for mapping transnational submarine optical cables and autonomous systems in this embodiment, and will not be repeated here.

[0064] The large-scale inference apparatus for mapping transnational submarine optical cables to autonomous systems proposed in this application can perform inferences based on publicly available global router-level topology, thus expanding the measurement range; it can utilize positive sample learning to classify samples that are difficult to judge by strict rules; it reduces the interference of the MPLS protocol on the inference results based on distance features; by comparing the changes in global router-level topology before and after the construction of submarine optical cables, it distinguishes submarine optical cables with very similar landing stations and routes, further improving measurement accuracy; it can expand the measurement range while avoiding detection overhead; and through this mapping, it combines transnational submarine optical cables that actually transmit data with autonomous systems that route data forwarding directions, providing data support for a better understanding of the role of transnational submarine optical cables in the global network.

[0065] Figure 5 A schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device may include:

[0066] The memory 501, the processor 502, and the computer program stored on the memory 501 and capable of running on the processor 502.

[0067] When the processor 502 executes the program, it implements the large-scale inference method for mapping transnational submarine optical cables and autonomous systems provided in the above embodiments.

[0068] Furthermore, electronic devices also include:

[0069] Communication interface 503 is used for communication between memory 501 and processor 502.

[0070] The memory 501 is used to store computer programs that can run on the processor 502.

[0071] The memory 501 may include high-speed RAM (Random Access Memory) memory, and may also include non-volatile memory, such as at least one disk storage.

[0072] If the memory 501, processor 502, and communication interface 503 are implemented independently, then the communication interface 503, memory 501, and processor 502 can be interconnected via a bus to complete communication between them. The bus can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 5 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0073] Optionally, in a specific implementation, if the memory 501, processor 502, and communication interface 503 are integrated on a single chip, then the memory 501, processor 502, and communication interface 503 can communicate with each other through an internal interface.

[0074] Processor 502 may be a CPU (Central Processing Unit), an ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement embodiments of this application.

[0075] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for large-scale inference of mapping between transnational submarine optical cables and autonomous systems.

[0076] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0077] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0078] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more N executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0079] It should be understood that the various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (FPGAs), field-programmable gate arrays (FPGAs), etc.

[0080] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0081] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A method for large-scale inference of mapping between transnational submarine optical cables and autonomous systems, characterized in that, Includes the following steps: Obtain global router-level topology and global transnational submarine fiber optic cable topology; The global router-level topology and the global transnational submarine optical cable topology are input into the trained classifier, and the mapping relationship between the router and the transnational submarine optical cable is output. The training process of the classifier includes: First, positive samples are selected based on strict screening conditions: there is one and only one transnational submarine optical cable between two countries and there is no land between the two countries. If there is also a link directly connecting the two countries in the ITDK between the two countries, that is, this link passes through the corresponding unique transnational submarine optical cable, the mapping relationship between these links and the transnational submarine optical cable is used as positive samples; Second, the positive sample learning classifier is trained using positive samples until the training stopping condition is met, and the trained classifier is obtained. The process of training a positive sample learning classifier using positive samples is as follows: Input positive samples and unlabeled samples; perform N iterations, in each iteration taking all positive samples and a portion of unlabeled samples to train a binary classifier, and calculating the probability of the remaining unlabeled samples being classified as positive based on the trained binary classifier; for all unlabeled samples, if the average probability of being classified as positive after N iterations is greater than 0.5, then it is classified as positive by the positive sample classifier, otherwise it is classified as negative. The router's IP address is converted to an Autonomous System number (AS number), and a mapping relationship between the transnational submarine optical cable and the AS number is generated based on the AS number and the mapping relationship between the router and the transnational submarine optical cable.

2. The large-scale inference method for mapping transnational submarine optical cables to autonomous systems according to claim 1, characterized in that, Before translating the router's IP address to an Autonomous System number, the following steps are also included: Identify the distances between landing stations of transnational submarine optical cables within the same geographical area; If the distance between the landing stations is less than or equal to the first preset distance, the transnational submarine optical cable corresponding to the landing station is determined to be an adjacent transnational submarine optical cable. The mapping relationship between the router and the transnational submarine optical cable is determined based on the link changes in the global router-level topology before and after the construction of the adjacent transnational submarine optical cable.

3. The large-scale inference method for mapping transnational submarine optical cables to autonomous systems according to claim 1, characterized in that, The training process for the classifier includes: Obtain positive samples that meet preset screening conditions, wherein the preset screening conditions include that there is one and only one transnational submarine optical cable between the two countries and there is no land between the two countries; The positive samples are used to train a positive sample learning classifier until the training stopping condition is met, resulting in a fully trained classifier.

4. The large-scale inference method for mapping transnational submarine optical cables to autonomous systems according to claim 3, characterized in that, Before training the positive sample learning classifier using the positive samples, the method further includes: Remove samples from the positive samples where the distance between the router and the login station is greater than a second preset distance.

5. A large-scale inference device for mapping transnational submarine optical cables to autonomous systems, characterized in that, include: The acquisition module is used to acquire global router-level topology and global transnational submarine optical cable topology; The input module is used to input the global router-level topology and the global transnational submarine optical cable topology into the trained classifier, and output the mapping relationship between the router and the transnational submarine optical cable. The training process of the classifier includes: first, selecting positive samples based on strict screening conditions: there is one and only one transnational submarine optical cable between two countries and the two countries have no land. If there is also a link directly connecting the two countries in the ITDK between the two countries, that is, this link passes through the corresponding unique transnational submarine optical cable, the mapping relationship between these links and the transnational submarine optical cable is used as positive samples; second, the positive sample learning classifier is trained using positive samples until the training stopping condition is met, and the trained classifier is obtained. The process of training a positive sample learning classifier using positive samples is as follows: Input positive samples and unlabeled samples; perform N iterations, in each iteration taking all positive samples and a portion of unlabeled samples to train a binary classifier, and calculating the probability of the remaining unlabeled samples being classified as positive based on the trained binary classifier; for all unlabeled samples, if the average probability of being classified as positive after N iterations is greater than 0.5, then it is classified as positive by the positive sample classifier, otherwise it is classified as negative. The generation module is used to convert the router's IP address into an Autonomous System number (AS number), and generate a mapping relationship between the transnational submarine optical cable and the AS number based on the AS number and the mapping relationship between the router and the transnational submarine optical cable.

6. The large-scale inference apparatus for mapping transnational submarine optical cables to autonomous systems according to claim 5, characterized in that, Also includes: The identification module is used to identify the distance between landing stations of transnational submarine optical cables within the same geographical area before converting the router's IP address to an Autonomous System number; If the distance between the landing stations is less than or equal to the first preset distance, the transnational submarine optical cable corresponding to the landing station is determined to be an adjacent transnational submarine optical cable. The mapping relationship between the router and the transnational submarine optical cable is determined based on the link changes in the global router-level topology before and after the construction of the adjacent transnational submarine optical cable.

7. The large-scale inference device for mapping transnational submarine optical cables and autonomous systems according to claim 5, characterized in that, The training process for the classifier includes: Obtain positive samples that meet preset screening conditions, wherein the preset screening conditions include that there is one and only one transnational submarine optical cable between the two countries and there is no land between the two countries; The positive samples are used to train a positive sample learning classifier until the training stopping condition is met, resulting in a fully trained classifier.

8. The large-scale inference device for mapping transnational submarine optical cables and autonomous systems according to claim 7, characterized in that, Also includes: The elimination module is used to eliminate samples in the positive samples whose distance between the router and the login station is greater than a second preset distance before training the positive sample learning classifier using the positive samples.

9. An electronic device, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, the processor executing the program to implement the large-scale inference method for mapping transnational submarine optical cables and autonomous systems as described in any one of claims 1-4.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, The program is executed by the processor to implement the large-scale inference method for mapping transnational submarine optical cables to autonomous systems as described in any one of claims 1-4.

Images