Method for IP geolocation based on multi-server collaborative measurement

Abstract

The application discloses an IP geolocation method based on multi-server cooperative measurement. The application maps the terminal IP to a more stable proxy node in the topology through path detection and last hop identification, combines operator information analysis and conflict resolution, constructs a reliable network home determination basis, reduces the uncertainty caused by information distortion, path anonymity and dynamic change, cooperatively schedules multiple measurement servers under the constraints of operators and autonomous systems, constructs a multi-point delay measurement system with spatial coverage, and suppresses the interference of instantaneous jitter and abnormal samples through multiple rounds of detection and time delay stability evaluation. Through the introduction of abnormal delay cleaning and multi-factor adaptive weighting strategy, the measurement results are weighted according to stability and reliability, and the robustness to noise and complex network state is enhanced.

Description

Technical Field

[0001] This application relates to the field of computer networks and information processing, and specifically to an IP geolocation method based on multi-server collaborative measurement. Background Technology

[0002] IP geolocation is a technology that maps Internet Protocol (IP) addresses to physical geographical locations. With the widespread adoption of internet applications, this technology has become an indispensable foundation for fields such as network security management, targeted advertising, regionalized content distribution, and logistics tracking.

[0003] Existing IP geolocation technologies mainly include geolocation methods based on geographic information database queries and network measurement.

[0004] Location methods based on geographic information databases typically obtain the geographic location information of a target IP address by querying commercial or open IP geographic location databases. However, due to the dynamic nature of IP address allocation and network topology, there may be discrepancies between the geographic location information in the database and the actual location of the target IP address, especially in scenarios with dynamically allocated IP addresses.

[0005] Network-based location methods typically infer the geographical location of a target IP address by performing network probing operations. Common measurement methods include path probing and latency measurement. These methods rely on network measurement results for inference, but these results are susceptible to factors such as routing changes, link fluctuations, and cross-carrier transmissions, leading to discrepancies between measurement nodes and affecting the consistency and stability of the location results.

[0006] In addition, querying based on IP registration information can be used to obtain the registration or attribution information of IP addresses, but the relevant information usually reflects the address at the management or registration level, which may be inconsistent with the actual location of use, and is difficult to use as a stable and reliable basis for positioning.

[0007] Therefore, existing technologies struggle to obtain stable and accurate IP geolocation results when faced with multiple challenges such as dynamic changes in network topology, fluctuations in measurement results, and insufficient reliability of available data sources. This remains a technical problem that urgently needs to be solved. Summary of the Invention

[0008] The purpose of this application is to provide an IP geolocation method based on multi-server collaborative measurement, so as to at least solve some of the problems mentioned in the background art.

[0009] According to one aspect of this application, an IP geolocation method based on multi-server collaborative measurement is provided, comprising the following steps:

[0010] S1. Send a probe message for path probing to the target IP address to perform hop-by-hop route tracing, obtain path data containing multiple routing nodes, and determine the IP address of the previous hop router corresponding to the target IP address based on the path data. S2. Based on the registration information of the IP address of the previous hop router, determine the network operator to which it belongs; S3. Based on the determined network operator, schedule multiple measurement servers with known geographical locations to perform network latency measurement on the IP address of the previous hop router and obtain multiple latency measurement values. S4. Perform abnormal data removal processing on the multiple delay measurement values ​​to generate valid delay data. Based on a preset weight model, perform weighted processing on the valid delay data to generate a comprehensive delay value. S5. Based on the comprehensive delay value and the known geographical locations of the multiple target measurement servers, calculate the geographical location coordinates of the IP address of the previous hop router using a multi-point delay positioning algorithm, and output the geographical location information of the target IP address based on the calculated geographical location coordinates of the IP address of the previous hop router.

[0011] Preferably, in step S1, when the target IP address blocks ICMP protocol responses, a route tracing method based on TCP or UDP protocols is adopted.

[0012] Preferably, step 2 includes the following methods: S21. Extract multiple identifier fields containing the operator's name based on the operator registration information; S22. Based on the identifier field, perform character decomposition, keyword extraction, and field credibility calculation; S23. Sort the identifier fields according to their credibility. S24. If there are differences between the top-ranked identifier fields, conflict resolution will be performed based on the field keyword matching degree and field type.

[0013] Preferably, within the determined network operator, multiple geographically known measurement servers are scheduled, specifically including the following steps: S31. Based on the network topology attribution information of the IP address of the previous hop router, and using the measurement resources under the jurisdiction of the network operator, determine the candidate measurement server that is adjacent to its network topology. S32. Based on the preset geographical distribution requirements, select measurement servers distributed in different geographical locations from the candidate measurement servers; S33. Confirm that the number of selected measurement servers reaches the preset threshold for the number of measurement nodes.

[0014] Preferably, step S32 includes the following steps: S321. Control each of the measurement servers to send multiple rounds of probe messages to the IP address of the previous hop router; S322. Collect and record the round-trip delay corresponding to each round of detection by each measurement server; S323. For each measurement server, evaluate latency stability based on its multi-round round-trip latency data; S324. Determine the corresponding delay measurement value based on the round-trip delay data and delay stability of each measurement server.

[0015] Preferably, step S4 includes the following steps: S41. Based on the preset absolute delay threshold, remove delay values ​​that exceed the absolute delay threshold; S42. Calculate the central trend value based on the remaining effective delay data; S43. Determine the deviation magnitude based on the deviation of each time delay value relative to the central trend value; S44. Calculate the dynamic deviation coefficient by combining the number of measurement servers and the current network latency fluctuations; S45. Based on the dynamic deviation coefficient, remove time delay values ​​that exceed the deviation limit to generate an effective delay data set.

[0016] Preferably, if the number of valid delayed data obtained after delayed cleaning is less than a preset lower limit, then the following steps are performed: Based on the ISP to which the IP address of the previous hop router belongs, retrieve the corresponding latency record from historical measurement data; Based on the delay records, determine the historical delay reference value; Generate compensatory delay data based on historical delay reference values; If the amount of compensatory delay data is still insufficient, then obtain the geographical location information corresponding to the target IP address; The compensatory delay data and geographic location information are then provided together in step S5.

[0017] Preferably, in step S4, based on a preset weighting model, weighted processing is performed on the effective delay data to generate a comprehensive delay value, specifically including: in, This indicates the number of target measurement servers participating in valid measurements. Indicates the first The latency measurement results corresponding to each target measurement server This represents the weighting parameter corresponding to the delay measurement result. This represents the calculated overall delay value.

[0018] Preferably, in step S5, the multi-point delay positioning algorithm is as follows: construct an objective function by the difference between the estimated geographical distance corresponding to each delay measurement value and the theoretical distance calculated based on the coordinates of the measurement server and the coordinates to be determined, establish a weighted nonlinear least squares optimization model, and obtain the optimal geographical coordinates by solving the algorithm through iterative optimization.

[0019] Preferably, the method further includes the step: S6, comparing the deviation between multiple consecutive location results of the same IP address range with a preset distance threshold; if the deviation is less than the threshold, the location result is marked as a high-confidence result and persistently stored.

[0020] This application addresses the technical bottlenecks in existing IP geolocation technologies, such as frequent dynamic changes in network topology, susceptibility of measurement results to path fluctuations and cross-carrier transmission, and the difficulty in balancing positioning accuracy and stability with single data source or single-point measurement methods. It aims to overcome the limitations of traditional static database queries or single-node network measurements, while ensuring real-time large-scale positioning, and improving the collaborative utilization capability of multi-source network measurement data to achieve stable and accurate inference of IP geographical location in complex network environments.

[0021] Specifically, this application maps terminal IPs to more stable proxy nodes in the topology through path probing and previous-hop identification. Combined with operator information parsing and conflict resolution, it constructs a reliable basis for network affiliation determination, reducing uncertainties caused by information distortion, path anonymity, and dynamic changes. Under the constraints of operators and autonomous systems, it coordinates the scheduling of multiple measurement servers to construct a multi-point delay measurement system with spatial coverage. Through multiple rounds of probing and delay stability assessment, it suppresses interference from instantaneous jitter and abnormal samples. By introducing abnormal delay cleaning and multi-factor adaptive weighting strategies, it assigns weights to measurement results according to stability and reliability, enhancing robustness to noise and complex network conditions. In addition, based on comprehensive delay constraints, it establishes a multi-point positioning model, jointly models and iteratively solves the results from multiple servers, and combines historical measurement compensation and multi-level caching to solidify and reuse stable positioning results, thereby improving the stability and consistency of positioning under different times and network conditions. Attached Figure Description

[0022] Figure 1 This is a block diagram of an IP geolocation method based on multi-server collaborative measurement according to an embodiment of this application; Figure 2 This is a flowchart of an IP geolocation method based on multi-server collaborative measurement according to an embodiment of this application; Figure 3 This is a schematic diagram illustrating the acquisition of IP publishing information using an IP geolocation method based on multi-server collaborative measurement according to an embodiment of this application. Figure 4 This is a schematic diagram illustrating a method for IP geolocation based on multi-server collaborative measurement according to an embodiment of this application, which involves querying a target IP operator using WHOIS information. Figure 5 This is a schematic diagram illustrating the acquisition of routes through a multi-line server in an IP geolocation method based on multi-server collaborative measurement according to an embodiment of this application. Figure 6 This is a schematic diagram illustrating a method for IP geolocation based on multi-server collaborative measurement according to an embodiment of this application, in which the previous hop operator of the route is queried through WHOIS information. Figure 7 This is a schematic diagram of an IP BGP multiline server according to an embodiment of the IP geolocation method based on multi-server collaborative measurement according to an embodiment of this application. Figure 8 This is a schematic diagram of an IP BGP multiline server according to an embodiment of the IP geolocation method based on multi-server collaborative measurement according to an embodiment of this application. Figure 9 This is a schematic diagram of another IP's BGP multiline server in an IP geolocation method based on multi-server collaborative measurement according to an embodiment of this application. Detailed Implementation

[0023] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0024] Please see Figure 1 and Figure 2 According to one embodiment of this application, an IP geolocation method based on multi-server collaborative measurement is provided. Specifically, the method includes the following steps S1-S5.

[0025] like Figure 3 As shown, in step S1, for the target IP address to be located, the scheduling node sends a probe message for path probing to the target IP address and performs a hop-by-hop route tracing operation to obtain the complete route path from the scheduling node to the target IP address.

[0026] In one embodiment, the probe packets are preferably sent using the ICMPTraceroute method. When the target IP address does not respond to the ICMP packets, the probe is automatically switched to a TCP or UDP-based path probing method, and commonly used service ports are selected for probing to ensure path reachability.

[0027] The system records the response status of different probe protocols and preferentially selects the probe protocol type that can return a complete hop-by-hop path as the effective path probing method for the current target IP. By parsing the returned path data, the penultimate hop routing node in the path is identified, and the IP address of the router corresponding to the penultimate hop is determined as the IP address of the previous hop router of the target IP address (Last-Hop-IP). This previous hop router IP address is used as the proxy target for subsequent network measurement and geolocation.

[0028] In one embodiment, the system further records the Autonomous System (AS) number corresponding to the IP address of the previous hop router for subsequent measurement server screening and location range constraints.

[0029] like Figure 4 As shown, in step S2, after obtaining the IP address of the previous hop router, the system performs a registration information query operation based on the IP address of the previous hop router to obtain the corresponding operator registration information.

[0030] In one embodiment, the operator registration information includes at least fields such as NetName, org-name, desc, and country. The system parses multiple fields containing operator names or operator identifiers based on the operator registration information and performs character decomposition and keyword matching processing.

[0031] When there are multiple identifier fields in the registration information that may point to different operators, the system further calculates the field credibility of each identifier field and sorts them according to the field credibility. If there are differences between the identifier fields that are ranked first, conflict resolution is performed based on the field keyword matching degree and field type to determine the operator category to which the IP address of the previous hop router belongs.

[0032] After determining the carrier category, the system uses that carrier category as a priority constraint for subsequent measurement server scheduling.

[0033] Furthermore, a preset weighting model is used to weight the effective delay data obtained in step S4, and its weighting parameters are determined based on a variety of factors defined in the claims.

[0034] Specifically, for each target measurement server, the system obtains its corresponding latency stability factor, error compensation factor, latency sensitivity factor, and topology distance factor based on the latency measurement results.

[0035] Among them, the delay stability factor is determined based on the dispersion of the delay data and is used to reflect the stability of the delay measurement result in multiple rounds of detection; the error compensation factor is determined based on the historical positioning error statistics of the target measurement server and is used to reflect the reliability of the measurement server in the historical positioning process; the delay sensitivity factor is generated based on the reciprocal of the effective delay data and is used to reflect the constraint strength of the delay measurement result on the positioning calculation; and the topology distance factor is determined based on the topology complexity of the network path between the target measurement server and the target IP address and is used to reflect the degree of influence of the network path on the delay measurement result.

[0036] The system performs normalization on the above factors to eliminate dimensional differences and generates weight parameters for weighted calculations. These weight parameters are used to perform weighted processing on the corresponding effective delay data, thereby allowing delay measurement results with higher stability, smaller errors, and better topological paths to have a greater impact on positioning calculations during the subsequent positioning model construction process.

[0037] like Figure 5 and Figure 6 As shown, in step S3, after determining the carrier category to which the IP address of the previous hop router belongs, the system obtains multiple candidate measurement servers and reads the geographical coordinates corresponding to each candidate measurement server.

[0038] Subsequently, the system filters out measurement servers located within the same operator's network from the candidate measurement servers based on the operator category, and further combines autonomous system information to preferentially select measurement servers located within the autonomous system range that has the same IP address as or is adjacent to the previous hop router's IP address.

[0039] The system uses the AS number of the IP address of the previous hop router as one of the filtering criteria to eliminate measurement servers that are too far away from its network topology.

[0040] Based on this, the system statistically analyzes the distribution of measurement servers in multiple different geographical areas to determine whether they meet the preset regional coverage conditions. Among the measurement servers that meet the regional coverage conditions, a quantity check is performed to determine the set of measurement servers that meet the preset quantity threshold, and this set is identified as multiple target measurement servers.

[0041] In this embodiment, the number of target measurement servers is preferably no less than 30.

[0042] Subsequently, the system controls each target measurement server to send multiple rounds of probe messages to the IP address of the next-hop router.

[0043] Furthermore, each target measurement server sends 3 to 5 rounds of probe messages.

[0044] The system records the delay measurement values ​​corresponding to each round of detection, and calculates the minimum, maximum, average and dispersion values ​​based on the delay measurement values. Furthermore, based on the dispersion, it determines the corresponding delay stability index, and determines the delay measurement result corresponding to each target measurement server based on the minimum round-trip delay and the delay stability index.

[0045] like Figure 7 and Figure 8 As shown, in step S4, after obtaining the latency measurement results of multiple target measurement servers, the system performs abnormal data removal processing on multiple latency measurement values.

[0046] The system first performs preliminary screening of the delay data based on a preset absolute delay threshold. In this embodiment, the absolute delay threshold is set to 500 milliseconds.

[0047] The system calculates the central trend value based on the remaining delay data, which is the median in this embodiment, and calculates the deviation of each delay value from the median.

[0048] The system calculates a dynamic deviation coefficient by combining the number of measurement servers currently participating in the measurement and the network latency fluctuation, and then removes latency data with deviations exceeding the limit.

[0049] If the number of valid delay data is less than a preset lower limit (e.g., less than 20), the system extracts delay reference values ​​from historical measurement data based on the operator to which the IP address of the previous hop router belongs, and generates compensatory delay data.

[0050] After obtaining valid delay data, the system performs weighted processing on the valid delay data based on a preset weight model, where the weight is inversely proportional to the corresponding delay value, and further combines the delay stability factor, historical error compensation factor and network topology factor for comprehensive determination.

[0051] In one embodiment, the system uses a weighted average algorithm to calculate the overall delay, with the specific calculation formula as follows: in, This indicates the number of target measurement servers participating in valid measurements. Indicates the first The latency measurement results corresponding to each target measurement server This represents the weighting parameter corresponding to the delay measurement result. This represents the calculated overall delay value. The weighting parameter is inversely proportional to the corresponding delay value (delay sensitivity factor), and is further combined with the delay stability factor, historical error compensation factor and network topology distance factor to determine the overall delay. This gives higher weight to delay measurement results with higher stability, smaller error and better topology path in the overall delay calculation.

[0052] The weights are determined based on the following factors: The latency sensitivity factor is generated based on the reciprocal of the effective latency data. The delay stability factor is determined based on the degree of dispersion of the delay data; Error compensation factor determined based on historical positioning error statistics of the target measurement server; The topological distance factor is determined based on the topological complexity of the network path between the target measurement server and the target IP address. The factors are then normalized to generate weight parameters for the weighting calculation.

[0053] like Figure 9 As shown, in step S5, after obtaining the comprehensive delay value, the system constructs a multi-point delay positioning model based on the comprehensive delay value and the geographical coordinates of multiple target measurement servers, and solves the model to obtain the geographical coordinates corresponding to the IP address of the previous hop router.

[0054] Since the IP address of the upstream router is usually located in the same autonomous system or the same data center as the target IP address, the system determines the geographical coordinates corresponding to the IP address of the upstream router as the geographical location information of the target IP address and outputs it.

[0055] When the target IP address corresponds to a CDN node or has DNS city identification information, the system further combines the DNS resolution records to perform secondary correction on the geographical coordinates.

[0056] Meanwhile, the system writes the geographic coordinates into the in-process cache, node-level cache, and distributed cache respectively. When the distance deviation of multiple consecutive positioning results is less than a preset threshold (for example, the deviation is less than 30 kilometers for 3 consecutive times), the geographic coordinates are solidified as high-confidence results and cached for a long time.

[0057] In this embodiment, the positioning model is a multi-point delay-constrained positioning model, which uses the geographic coordinates of the target location as the variable to be determined, and the known geographic coordinates of each target measurement server and the corresponding comprehensive delay value as input conditions.

[0058] Specifically, the system converts the comprehensive delay value into a corresponding distance constraint based on the network propagation delay relationship between each target measurement server and the geographic location to be determined, and constructs a weighted positioning constraint model with the geographic coordinates of the geographic location to be determined as the variable and the geographic coordinates of each measurement server as the known parameter.

[0059] During the model building process, different weights are assigned to measurement results from different time delays. The weights are inversely proportional to the corresponding comprehensive delay values ​​to reduce the impact of high-latency or unstable measurements on the positioning results.

[0060] The system obtains the geographic location coordinates that minimize the overall error by iteratively solving the weighted location constraint model, and uses these coordinates as the location result of the IP address of the previous hop router.

[0061] One embodiment is a geolocation scenario targeting the IP address of a typical home broadband user.

[0062] In step S1, a traceroute probe is performed on the target IP address 114.114.114.114, and its Last-Hop-IP is resolved to be 58.241.XX.XX.

[0063] In step S2, a Whois query is performed on 58.241.XX.XX to extract fields such as NetName and org-name, and to determine that it belongs to the CHINANET-JS network.

[0064] In step S3, 120 telecom probe nodes deployed nationwide within a certain operator's network are scheduled to perform multiple rounds of Ping measurements on 58.241.XX.XX, with each probe node sending 3 to 5 probe data packets.

[0065] In step S4, the collected delay data is cleaned to remove outliers with RTT greater than 500 milliseconds and MAD values ​​that deviate from the median by 3 times, and finally 98 valid probe nodes are retained.

[0066] In step S5, the comprehensive delay value is calculated based on the delay data of 98 valid probe nodes, and weighted by combining the geographical location information of each probe node. The geographical location of the target IP address is inferred to be 118.78°E, 32.04°N. After multiple location results verification, the final location error is less than 8 kilometers, and the city hit rate is 100%.

[0067] One embodiment addresses a server IP location scenario in a cloud computing environment.

[0068] In step S1, a traceroute probe is performed on the target IP address 47.98.6.235 to determine that its Last-Hop-IP is the entry node of a company's BGP network.

[0069] In step S2, a Whois query is used to confirm that the Last-Hop-IP belongs to a certain company's BGP network.

[0070] In step S3, multiple measurement nodes distributed in different cities are scheduled within the operator's network to perform multiple rounds of delay measurement on Last-Hop-IP.

[0071] In step S4, outliers are removed from the measurement results, and stable, valid delayed samples are retained.

[0072] In step S5, the comprehensive latency value is calculated based on the effective latency samples, and the geographical location of the target IP address is inferred to be 118.89°E, 30.24°N. After caching and multiple verifications, the location result deviates from the actual location of the data center in a certain region by less than 3 kilometers.

[0073] One embodiment is a geolocation scenario targeting CDN edge node IPs.

[0074] In step S1, a traceroute probe is performed on the target IP address to obtain its Last-Hop-IP.

[0075] In step S2, a Whois query is used to confirm that the operator to which Last-Hop-IP belongs is the backbone network of the CDN service provider.

[0076] In step S3, probe nodes distributed across multiple geographical areas are scheduled within the operator's network to perform collaborative latency measurements on Last-Hop-IP.

[0077] In step S4, the measurement data is cleaned to obtain a stable set of effective delay data.

[0078] In step S5, the geographical location of Last-Hop-IP is corrected based on the comprehensive latency value and combined with the DNS resolution city information, and the target IP address is finally determined to be located in a certain region with a positioning error of 11 kilometers.

[0079] All parts not covered in this application are the same as or can be implemented using existing technology. Although embodiments of this application have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of this application, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An IP geolocation method based on multi-server collaborative measurement, characterized in that, Includes the following steps: S1. Send a probe message for path probing to the target IP address to perform hop-by-hop route tracing, obtain path data containing multiple routing nodes, and determine the IP address of the previous hop router corresponding to the target IP address based on the path data. S2. Based on the registration information of the IP address of the previous hop router, determine the network operator to which it belongs; S3. Based on the determined network operator, schedule multiple measurement servers with known geographical locations to perform network latency measurement on the IP address of the previous hop router and obtain multiple latency measurement values. S4. Perform abnormal data removal processing on the multiple delay measurement values ​​to generate valid delay data. Based on a preset weight model, perform weighted processing on the valid delay data to generate a comprehensive delay value. S5. Based on the comprehensive delay value and the known geographical locations of the multiple target measurement servers, calculate the geographical location coordinates of the IP address of the previous hop router using a multi-point delay positioning algorithm, and output the geographical location information of the target IP address based on the calculated geographical location coordinates of the IP address of the previous hop router.

2. The method according to claim 1, characterized in that, In step S1, when the target IP address prohibits ICMP protocol responses, a route tracing method based on TCP or UDP protocols is adopted.

3. The method according to claim 1, characterized in that, Step 2 includes the following methods: S21. Based on the operator registration information, extract multiple identifier fields containing the operator name; S22. Based on the identifier field, perform character decomposition, keyword extraction, and field credibility calculation; S23. Sort the identifier fields according to their credibility. S24. If there are differences between the top-ranked identifier fields, conflict resolution will be performed based on the field keyword matching degree and field type.

4. The method according to claim 1, characterized in that, Within the identified network operator, multiple geographically known measurement servers are scheduled, specifically including the following steps: S31. Based on the network topology attribution information of the IP address of the previous hop router, and using the measurement resources under the jurisdiction of the network operator, determine the candidate measurement server that is adjacent to its network topology. S32. Based on the preset geographical distribution requirements, select measurement servers distributed in different geographical locations from the candidate measurement servers; S33. Confirm that the number of selected measurement servers reaches the preset threshold for the number of measurement nodes.

5. The method according to claim 4, characterized in that, Step S32 includes the following steps: S321. Control each of the measurement servers to send multiple rounds of probe messages to the IP address of the previous hop router; S322. Collect and record the round-trip delay corresponding to each round of detection by each measurement server; S323. For each measurement server, evaluate latency stability based on its multi-round round-trip latency data; S324. Determine the corresponding delay measurement value based on the round-trip delay data and delay stability of each measurement server.

6. The method according to claim 1, characterized in that, Step S4 includes the following steps: S41. Based on a preset absolute delay threshold, remove delay values ​​that exceed the absolute delay threshold; S42. Calculate the central trend value based on the remaining effective delay data; S43. Determine the deviation magnitude based on the deviation of each delay value relative to the central trend value; S44. Calculate the dynamic deviation coefficient by combining the number of measurement servers and the current network latency fluctuations; S45. Based on the dynamic deviation coefficient, remove delay values ​​that exceed the deviation limit to generate the effective delay data set.

7. The method according to claim 1, characterized in that, In step S4, if the number of valid delayed data obtained after delayed cleaning is less than a preset lower limit, then the following steps are performed: Based on the operator to which the IP address of the previous hop router belongs, retrieve the corresponding latency record from historical measurement data; Based on the aforementioned delay records, determine the historical delay reference value; Based on the historical delay reference value, compensated delay data is generated; If the amount of compensatory delay data is still insufficient, then obtain the geographical location information corresponding to the target IP address; The compensatory delay data and the geographic location information are then provided together in step S5.

8. The method according to claim 1, characterized in that, In step S4, based on the preset weight model, the effective delay data is weighted to generate a comprehensive delay value, specifically including: in, This indicates the number of target measurement servers participating in valid measurements. Indicates the first The latency measurement results corresponding to each target measurement server This represents the weighting parameter corresponding to the delay measurement result. This represents the calculated overall delay value.

9. The method according to claim 1, characterized in that, In step S5, the multi-point delay positioning algorithm is as follows: construct an objective function by the difference between the estimated geographical distance corresponding to each delay measurement value and the theoretical distance calculated based on the coordinates of the measurement server and the coordinates to be determined, establish a weighted nonlinear least squares optimization model, and obtain the optimal geographical coordinates by solving the algorithm through iterative optimization.

10. The method according to claim 1, characterized in that, It also includes the following steps: S6. Compare the deviation between multiple consecutive location results of the same IP address range with a preset distance threshold. If the deviation is less than the threshold, mark the location result as a high-confidence result and store it persistently.

Images