WO2021093823A1

WO2021093823A1 - Pseudo base station identification method and system, and computer readable storage medium

Info

Publication number: WO2021093823A1
Application number: PCT/CN2020/128476
Authority: WO
Inventors: 陆纬
Original assignee: 中兴通讯股份有限公司
Priority date: 2019-11-14
Filing date: 2020-11-12
Publication date: 2021-05-20
Also published as: CN112804701A

Abstract

A pseudo base station identification method and system, and a computer readable storage medium. The pseudo base station identification method comprises: determining a key counter from performance data counters of a base station, wherein data in the key counter is affected by a pseudo base station (S101); establishing a corresponding regression analysis model according to the key counter (S102); extracting statistical data corresponding to the key counter from a network management database, substituting same into the regression analysis model for machine learning, to determine a threshold of a correlation coefficient of the pseudo base station (S103); and calculating performance data of the key counter and comparing same with the threshold, to determine whether the pseudo base station exists (S104).

Description

Pseudo base station identification method, system and computer readable storage medium

Cross-references to related applications

This application is filed based on a Chinese patent application with an application number of 201911114985.X and an application date of November 14, 2019, and claims the priority of the Chinese patent application. The entire content of the Chinese patent application is hereby incorporated into this application by reference.

Technical field

The present invention relates to but not limited to the field of communication technology, and more specifically, relates to but not limited to a pseudo base station identification method, system, and computer-readable storage medium.

Background technique

In the communication equipment system, in order to be able to comprehensively evaluate the operating status of the base station system and to monitor and optimize network services, it will define the parameters that evaluate and identify the normal, pros and cons of the system’s various important operating conditions, namely the counter, and then The network management system regularly collects counter data from network equipment, and these counter data are the source of the base station performance big data mentioned in the solution in this article.

Pseudo base stations controlled by illegal organizations or individuals, independent of the public mobile network, disguised as a base station of a regular mobile communication operator (for example, broadcasting a mobile operator’s public land mobile network ID (PLMN ID)) , So as to trick the user terminal to initiate a network registration or location update request to it, and then extract the terminal's information. In addition, pseudo base stations often send fraudulent short messages, malicious URL links or advertising short messages to the terminal, thereby endangering users.

The existing methods for locating pseudo base stations are generally divided into two types. The first is to use SMS data for analysis. This method is to detect the user's received short message, extract the characteristics of the short message and compare it with the big data of the cloud short message. If the characteristic information of the problem short message is met, the short message is placed in the intercepted state and the current terminal position is reported to locate the pseudo base station. . This method needs to extract the user's text messages and location information. Today, when network security and privacy are increasingly emphasized, users tend to resist. The other is to use a certain detection terminal to extract the characteristic information of the pseudo base station, such as frequency point, signal strength and other information. This method can detect the existence of pseudo base stations in real time, but because pseudo base stations are often mounted in vehicles and are highly mobile, public security agencies often do not have enough manpower and material resources to conduct large-scale net-based searches.

Summary of the invention

The embodiments of the present invention provide a pseudo base station identification method, system, and computer-readable storage medium.

The pseudo base station identification method according to the embodiment of the present invention includes: determining a key counter from the performance data counter of the base station, and the data in the key counter is affected by the pseudo base station; establishing a corresponding regression analysis model according to the key counter; The statistical data corresponding to the key counter is extracted from the database and substituted into the regression analysis model for machine learning to determine the threshold value of the correlation coefficient of the pseudo base station; the performance data of the key counter is calculated and compared with the threshold value to determine whether The pseudo base station exists.

According to another embodiment of the present invention, a pseudo base station identification system is provided. The pseudo base station identification system includes: a base station and an analysis device; the base station is used to generate performance data using a performance data counter; and the analysis device includes data extraction Module, rule learning module, and data storage module; the data extraction module is connected to the network management database, and extracts data corresponding to key counters from the network management database; the planning learning module is used to formulate analysis models and calculate based on historical performance data The threshold value of the correlation coefficient; the data storage module is used to store the data required by the pseudo base station identification method described in the embodiment of the present invention.

According to still another embodiment of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium stores one or more programs, and the one or more programs can be executed by one or more processors, In order to realize the steps of the pseudo base station identification method described in the embodiment of the present invention.

Other features and corresponding beneficial effects of the present invention are described in the latter part of the specification, and it should be understood that at least part of the beneficial effects will become apparent from the description in the specification of the present invention.

Description of the drawings

The present invention will be further described below in conjunction with the accompanying drawings and embodiments. In the accompanying drawings:

FIG. 1 is a schematic flowchart of a method for identifying a pseudo base station according to Embodiment 1 of the present invention;

2 is a schematic diagram of the influence of the pseudo base station on the data of the LTE average user count counter of the normal base station provided by the second embodiment of the present invention;

3 is a schematic diagram of the influence of a user's normal handover of a cell on the counter data of the number of successful handovers in and out of the base station according to the second embodiment of the present invention;

4 is a schematic diagram of the influence of the pseudo base station on the counter data of the number of successful handovers in and out of the normal base station according to the second embodiment of the present invention;

5 is a flowchart of machine learning based on performance data provided by Embodiments 3, 4, and 5 of the present invention;

6 is a flow chart of drawing pseudo base station trajectory in an unknown area according to the confidence interval obtained by learning according to the sixth embodiment of the present invention;

FIG. 7 is a schematic diagram of a pseudo base station identification system provided by Embodiment 7 of the present invention;

FIG. 8 is a schematic diagram of another pseudo base station identification system provided by Embodiment 3 of the present invention.

Detailed ways

In order to make the objectives, technical solutions, and advantages of the present invention clearer, the following further describes the embodiments of the present invention in detail through specific implementations in conjunction with the accompanying drawings. It should be understood that the specific embodiments described here are only used to explain the present invention, but not to limit the present invention. At the same time, it should be noted that the embodiments in the application and the features in the embodiments can be combined with each other if there is no conflict.

Example one:

The present invention provides a pseudo base station identification method and system based on base station performance big data and machine learning. At the same time, the location of the pseudo base station can be realized through the location information of the base station. The method uses a machine learning algorithm to detect fluctuations in the performance data of the base station. Study and analyze, determine whether there is a pseudo base station, and combine the location of the base station and the road network information to draw the movement trajectory of the pseudo base station.

In order to realize the identification of pseudo base stations, this embodiment provides a method for identifying pseudo base stations. The pseudo base station identification method includes: determining a key counter from the performance data counter of the base station, and the data in the key counter is affected by the pseudo base station; The counter establishes the corresponding regression analysis model; extracts the statistical data corresponding to the key counter from the network management database, and substitutes it into the regression analysis model for machine learning to determine the threshold value of the correlation coefficient of the pseudo base station; calculates the performance data of the key counter and compares the threshold value, Determine whether there is a pseudo base station.

Please refer to FIG. 1. FIG. 1 is a schematic flowchart of a pseudo base station identification method provided in this embodiment. The process of the pseudo base station identification method includes:

S101. Determine a key counter from the performance data counter of the base station.

The key counter determined from the base station requires that the data collected by the key counter will be affected by the pseudo base station, and the greater the impact, the more obvious the more conducive to the identification of the pseudo base station.

The performance data of the base station cannot directly feed back whether the user has accessed the pseudo base station. If a user terminal is connected to the pseudo base station, what the normal base station perceives is a normal user switch-out operation. However, when the pseudo base station passes by the normal base station, more users are often affected, which will cause the performance data of the normal base station to fluctuate or deteriorate. These counters that produce data fluctuations are the key counters. Therefore, these key counters can be used. The data fluctuation of the counter speculates whether there is a pseudo base station.

S102: Establish a corresponding regression analysis model according to the key counter.

After determining the key counter, it is necessary to determine the corresponding regression analysis model according to the key counter. The regression analysis model can clearly show the changes and influencing factors of the data in the key counter.

S103. Perform machine learning to determine the threshold value of the correlation coefficient of the pseudo base station.

The statistical data corresponding to the key counter is extracted from the network management database and substituted into the regression analysis model for machine learning to determine the threshold value of the correlation coefficient of the pseudo base station. The network management database contains the performance data counted by the performance data counters of all base stations .

Machine learning requires a large amount of data. Therefore, after the regression analysis model is determined in step S102, in addition to machine learning on local data, a large amount of data and statistical data related to key counters can also be obtained from the network management database. Machine learning makes the results of machine learning more accurate, and the obtained threshold value of the correlation coefficient of the pseudo base station is closer to the actual situation.

S104. Calculate the performance data of the key counter and compare it with the threshold to determine whether the pseudo base station exists.

Through the machine learning the threshold value of the correlation coefficient of the pseudo base station, calculate the current performance data from the key counter data collected by the local base station, and compare the performance data with the threshold value to determine whether there are pseudo base stations currently around the base station.

In this embodiment, the base station includes a single base station or a base station in an area, and the selected key counter may also be a performance counter or multiple performance counters, which are integrated to obtain data. In practical applications, when a base station is working, a single base station can be selected to work, or multiple base stations in an area can be selected to work together.

In this embodiment, after determining that there is a pseudo base station, it further includes: determining the location of the pseudo base station and drawing the movement track of the pseudo base station; determining the location of the pseudo base station according to the location information of the base station where the pseudo base station is found, within a preset time interval , Calculate multiple location information of the pseudo base station and connect them to obtain the movement trajectory of the pseudo base station. After determining the presence of pseudo base stations around the base station, the location of the pseudo base station can be roughly determined through the location information of the base station. When using base stations in an area to determine the pseudo base station, the location information of multiple base stations can also be analyzed for comparison. Accurately analyze the location of the pseudo base station.

In real life, pseudo base stations are usually mobile base stations and rarely stay in one place. Therefore, in order to effectively combat criminal activities of pseudo base stations, it is convenient to understand and analyze the movement trend of pseudo base stations. Monitoring within a time interval and determining the movement trajectory of the pseudo base station can easily determine the next trend of the pseudo base station, thereby providing assistance in combating the pseudo base station.

In this embodiment, calculating the analysis data of the key counter of the base station to compare with the threshold includes, but is not limited to: substituting the statistical data of the key counter into the regression analysis model to obtain performance data, and comparing the performance data with the threshold to determine whether there is false Base station. It is understandable that the method for judging a pseudo base station provided above is only one of many methods. In practical applications, other correction parameters or other optional parameters can be added in the calculation process to make the result more accurate. Even in some cases, in order to prevent the results from being leaked and used, encryption algorithms can be added to make the calculation results more secure.

In this embodiment, the performance data counter of the base station includes, but is not limited to: the average number of users, the number of successful executions of intra-cell intra-frequency handover of the X2 interface between base stations, the number of successful intra-cell intra-frequency handovers of the X2 interface between base stations, The number of successful executions of inter-cell inter-frequency handover of the X2 interface between base stations, the number of successful inter-cell inter-frequency handovers of the X2 interface between base stations, the number of successful inter-cell inter-frequency handovers of the S1 interface between base stations, and the number of successful inter-cell inter-frequency handovers of the S1 interface between base stations. The number of successful executions of inter-cell inter-frequency handover, the number of successful executions of inter-cell intra-frequency handover of the S1 interface between base stations, and the number of successful inter-cell intra-frequency handovers of the S1 interface between base stations.

In this embodiment, the regression analysis model is a calculation formula corresponding to the key counter, and the result of the calculation formula is the threshold value of the pseudo base station correlation coefficient of the key counter. The calculation formula also includes reasonable coefficients, which are made close to the actual values through machine learning.

In this embodiment, the performance data counter includes a performance data counter in a base station or a performance data counter between base stations. For example, when the pseudo base station does not exceed the coverage of a base station or a base station in an area when moving, the data of the performance data counter counted at this time is the data of the performance data counter in the base station; when the pseudo base station is moving beyond a base station Or when the coverage of the base stations in an area is the case, the data of the performance data counter counted at this time may be the data of the performance data counter between the base stations. Avoiding errors in data coordination between base stations, making data collection errors smaller and recognition results more accurate.

This embodiment provides a pseudo base station identification method, including: determining a key counter from the performance data counter of the base station, and the data in the key counter is affected by the pseudo base station; establishing a corresponding regression analysis model according to the key counter; and from the network management database The statistical data corresponding to the key counter is extracted and substituted into the regression analysis model for machine learning to determine the threshold of the correlation coefficient of the pseudo base station; the performance data of the key counter is calculated and compared with the threshold to determine whether there is a pseudo base station. Judge and identify the pseudo base station by obtaining the key counter data of the base station, without obtaining the user's private data, avoiding the user's disgust; at the same time, the regression analysis model and machine learning are used to obtain the threshold value of the pseudo base station correlation coefficient, and the pseudo base station correlation coefficient is used The threshold value is used to identify whether there is a pseudo base station, so that the identification result is more accurate.

Further, this embodiment also determines the location of the pseudo base station through the location information of the base station, and draws the movement trajectory of the pseudo base station by acquiring the location information of the pseudo base station for a period of time, which is used to predict the subsequent movement position of the pseudo base station, which is helpful To combat the crime of fake base stations.

Embodiment two:

This embodiment takes a 4G LTE protocol base station as an example, and proposes an implementation manner of a pseudo base station identification method provided in an embodiment of the present invention.

It includes the following steps:

1) Filter performance data counters and determine key counters.

Taking the LTE (4G protocol) base station as an example, we select the following key counters: average number of LTE users, X2 (inter-base station interface) between eNBs (base station), the number of successful executions of simultaneous handover between cells, and X2 inter-eNB cells The number of successful executions of inter-frequency handover in the same frequency, the number of successful executions of inter-cell inter-frequency handover of X2 ports between eNBs, the number of successful inter-cell inter-frequency handovers of X2 ports between eNBs, and the number of inter-eNB S1 (inter-base station interface) ports. The number of successful executions of handover, the number of successful inter-frequency handovers between eNBs and S1 ports, the number of successful intra-frequency handovers between eNBs and S1 ports, the number of successful inter-eNB S1 inter-cell handovers, and so on. For the average number of LTE users, if a pseudo base station passes through the management range of a normal base station, the number of users picked up by the pseudo base station will result in a decrease in the average number of users of the base station’s statistical granularity. When the pseudo base station leaves the management range of the base station, The average number of users will rise again. See Figure 2. Figure 2 is a schematic diagram of the influence of a pseudo base station on the data of a normal base station LTE average number of users counter.

2) Establish a regression analysis model using the influence of the above pseudo base station on the fluctuation of the normal counter value.

Take the above-mentioned key counters related to LTE as an example.

Set the average number of LTE users at the current statistical granularity time point as X _now , and the number of users corresponding to the same statistical granularity on _{the nth day in the past is divided into X n} ; W _n is the difference between the value of the nth day in the past and the same time today Weights. The following basic formula can be obtained:

_Ru can be used as a correlation coefficient to determine the pseudo base station. W _n value may be closer and closer to a reasonable value through machine learning, and R _u calculated confidence interval. By calculating the performance data at a certain point in time, it can be directly inferred whether there is a pseudo base station at that point in time.

Further, the counters of the neighboring cell and the current cell can be used for regression analysis modeling at the same time.

When the user terminal switches to an adjacent cell, the number of handovers of the base station will be +1, and the number of handovers of the neighboring cell will be +1, as shown in Figure 3, Figure 3 shows the user's normal handover cell to the base station handover to and from the base station. Schematic diagram of the influence of the success counter data. However, when a user terminal is connected to a pseudo base station, the number of handovers out of the base station will be +1, while the number of handovers in adjacent cells will not be +1. Switch out the schematic diagram of the influence of the success counter data. If a large number of users are connected to the pseudo base station, the number of handovers of this base station will increase more than normal data, but the number of handovers of adjacent cells will not increase significantly. Due to space limitations, the "number of handovers" is used here to represent the number of successful simultaneous and inter-frequency handovers of the S1 and X2 ports between eNBs, and the "number of handovers" means the same frequency and different frequency handovers of the S1 and X2 ports between eNBs. The number of successful entries. Sets the base station in the switching times of the current statistical particle size interval is O _now, the current cell in the past n days on the same switching frequency and the statistical particle size _n is O; the current frequency handover statistics size k-th adjacent cell is IK _now , the number of switch-ins with the same statistical granularity on the nth day in the past is IK _n , and the following two formulas can be obtained:

In the above formula R _o is a cut-out station number of the correlation coefficient, R _o larger value indicates a pseudo base stations may exist. R _i is the correlation coefficient of the number of cell cut-in times of adjacent base _{stations. The smaller the R i, the} more likely there is a pseudo base station.

Depending on the meaning of the counter, there are other counters that can be used to build a regression model. This method is not exhaustive. For GSM (2G protocol), UMTS (3G protocol), TD-SCDMA (3G protocol), and 5G base stations or controllers, similar counters can be selected to establish regression analysis models and the above methods can be used for machine learning. The data analysis algorithm can be specified by the user or obtained through machine deep learning

.

3) Use historical performance data and labeled base station information for machine learning to obtain the threshold value of the correlation coefficient of the pseudo base station.

Machine learning is performed through historical performance data of known pseudo base stations. To give R _u, R _o, R _i or the like correlation user-defined threshold model.

4) Using the existing analysis model and threshold, calculate the performance data of the area to be analyzed, and determine whether the pseudo base station exists. Obtain a list of base stations suspected of having pseudo base stations.

It can be connected to the network management system to obtain performance data of base stations in unknown areas, and use analysis models and threshold information to calculate and determine whether there are pseudo base stations around the target base station.

5) Finally, combine the road network data and base station location information to draw the movement trajectory of the pseudo base station.

In the pseudo base station identification method provided in this embodiment, compared with the existing pseudo base station positioning solution, there is no need to specifically design a terminal to locate the pseudo base station. A set of systems can be used in multiple places, just input the model and performance data. The performance data analysis model is universal and can be customized by users. It is applicable to 2G, 3G, 4G and 5G networks, regardless of the communication equipment of any manufacturer, as long as the performance data related to the model is statistically analyzed, the system can be applied. This method does not collect data on the user terminal, and the performance data it relies on is all control plane data and does not involve user plane data. There is no risk of user information leakage.

Example three:

This embodiment uses the average number of LTE users to establish an analysis model. FIG. 5 is a flowchart of machine learning based on performance data in an embodiment of the present invention. The process mainly includes the following steps (S501-S505):

Step S501, screening the performance data counters that the pseudo base station will affect. For this embodiment, that is, the average number of LTE users.

In step S502, a regression analysis model is established according to the selected counter. In this embodiment, the above-mentioned _Ru formula is used to build a model, and the model is stored in the database.

The average number of LTE users at the current statistical granularity time point is X _now , and the number of users corresponding to the same statistical granularity in the past n days is divided into X _n ; W _n is the weight of the difference between the value of the past few days and the same time today.

In step S503, the counter data required for this learning is extracted from the network management system, and the marked pseudo base station data is imported.

Table 1 shows the performance counter data of the average number of LTE users imported into the system

Table 1 shows the data of the average number of LTE users of the base station with ID equal to 301 at the same statistical granularity during four days. Among them, Collectime is the data collection time; Subnetwork is the subnet ID; EnbId is the base station ID; the column of LTE average users represents the number of users counted at the corresponding time. The counter data collected from the network management system will be stored using the structure in Table 1. The Mark field indicates whether there is a mark of a pseudo base station near the base station. This part of the data comes from the data of a known pseudo base station in the public security system.

In step S504, machine learning is performed based on historical performance data and information of known pseudo base stations with tags. The value of W _n weight should be closer to the current time, the higher the weight. To simplify the description in this embodiment, _{the values of W n} are all set to 1, and n is 3. R _u can be calculated to obtain the value -248, plus as an error of plus or minus 10, to obtain the confidence interval R _min R _u correlation coefficient ~ R _max = -258 ~ -238. Repeat this learning process, continuously adjust the weight and error values, and finally get a more accurate confidence interval. For the performance data is determined to be a base station, if the calculated value by the equation R _u R _min ~ R _max range, it may be determined that the base station in the presence of the corresponding data outside the time point pseudo base stations.

The more the learned performance data and the number of samples of the pseudo base station, the more accurate the confidence interval obtained. For holidays, edge base stations and other data that will have a special impact on the data, it will be eliminated before learning.

In step S505, the confidence interval obtained by learning is stored in the database.

For 4G base stations, the algorithm is the formula established in this article using the average number of LTE users between eNBs, the number of switching in and out of the S1 and X2 interfaces. The machine learning model involved in determining the pseudo base station can be pre-modeled or customized by the user. Analysis models can be stored in the cloud, and a cloud model library can be established.

Embodiment four:

In this embodiment, the number of handovers in and out between eNBs is used to establish an analysis model. The fourth embodiment is similar to the third embodiment, and the steps in FIG. 5 can also be referred to.

In step S501, the filtered counter is the number of successful handovers in and out of the S1 and X2 interfaces between eNBs.

In step S502, the above-mentioned R _o and R _{i are used to} build a model.

Wherein in the formula R _o is a cut-out station number of the correlation coefficient, R _o larger value indicates a pseudo base stations may exist. R _i is the correlation coefficient of the number of cell cut-in times of adjacent base _{stations. The smaller the R i, the} more likely there is a pseudo base station.

In step S503, the counter data required for this learning is extracted from the network management system, and the marked pseudo base station data is imported. For simplification, the data of S1 interface is used as an example:

Table 2 shows the performance counter data of the number of times the S1 port is switched out between eNBs imported into this system

Table 2 shows the performance counter data of the number of times that the base station with ID equal to 301 has switched out the S1 port between eNBs at the same statistical granularity during four days.

Table 3 shows the performance counter data of the number of S1 port handovers between two adjacent base stations eNB

Table 3 shows the performance counter data of the number of inter-eNB S1 port handover times of two base stations with base station IDs equal to 302 and 303. The two base stations 302 and 303 are adjacent to the base station 301.

In step S504, machine learning is performed based on historical performance data and information of known pseudo base stations with tags.

To simplify the description, the weight of each correlation coefficient is still set to 1. Putting the data into the formula for calculation _{, the value of R o} is 705 and the value of R _i is -11. Then the base station to be determined, if it is greater than or R _o equal to 705 and the value of the neighboring base station is R _i is less than or equal to -11. It can be determined that there is a pseudo base station around the base station. Similarly, the more sample data is learned, the more accurate the threshold and weight values corresponding to _{R o} and R _{i will be.}

Embodiment five:

This embodiment integrates the average number of LTE users and the number of handovers in and out of eNBs to establish a regression analysis model. The fifth embodiment is the same as the third embodiment and the fourth embodiment. In order to make the result more accurate, the data of the above two embodiments are comprehensively calculated.

Step S501, screening the performance data counters that the pseudo base station will affect. In this embodiment, the counters of the third embodiment and the fourth embodiment are selected.

In step S502, a regression analysis model is established according to the selected counter. In this embodiment, still using the model _{_{_{R u, R o, R i}}} . The thresholds of the coefficients also use the values in the above two embodiments. among them:

_Ru is the correlation coefficient of the LTE average number of users counter.

R _{o is} the correlation coefficient of the counter of the number of times that the base station has switched out the success.

R _i is the correlation coefficient of the counter of the number of times that the neighboring base station switches into the success.

Build the following model:

R=(-258＜R _u ＜-238)AND(R _o ≥705)AND(R _i ≤-11)

The meaning of the equation is as follows: If the data is calculated after a base station, R _u correlation coefficient falls confidence interval -258 ~ -238 and R _o equal to 705 and greater than the R _i -11 or less, it may be determined that the neighboring base stations A pseudo base station is suspected.

Step S503, step S504, and step S505 are the same as the third and fourth embodiments.

Embodiment 6:

In this embodiment, the pseudo base station is determined in the unknown area and the pseudo base station trajectory is drawn. Fig. 6 is a flowchart of drawing the pseudo base station trajectory in the unknown area according to the confidence interval obtained by learning. The process includes the following steps:

Step S601: Extract historical data of required performance counters from the network management system. And use the same format described in Table 1 to store in the database, leave the Mark field blank.

Step S602: Calculate the data of each statistical granularity of each base station according to the formula in Embodiment 3 or Embodiment 4 or Embodiment 5 to obtain the correlation coefficient R of the pseudo base station of each base station at different time points. If R falls _{within the range of the confidence interval R min} to R _max , modify the Mark field of this base station to True. After the data of all base stations are calculated, the records in which the Mark field is True are filtered out, and all base station IDs that are judged to have pseudo base stations nearby can be obtained. See Table 4 below.

Table 4 List of base stations with pseudo base stations nearby

In step S603, since the location of each base station is known, according to the longitude and latitude of the base station in Table 4, combined with the map data, a dot matrix can be marked on the map. Combined with the road network information, the trajectory diagram of the pseudo base station can be drawn.

Using the above pseudo base station positioning method based on base station performance big data and machine learning has universal applicability. Regardless of the communication equipment of any manufacturer, this system can be applied as long as the performance data related to the model is statistically analyzed. The analysis model can be modified according to the actual situation. After the machine learning is completed, the system can analyze data from different regions and different manufacturers. It has the characteristics of low cost and wide applicability. It is convenient for the public security system to carry out targeted arrests in the areas under its jurisdiction.

Embodiment Seven:

Please refer to FIG. 7. FIG. 7 is a schematic diagram of a pseudo base station identification system provided by this embodiment. The pseudo base station identification system includes: a base station, a network management server, and a network management database;

The network management database stores performance data counted by performance data counters of all base stations;

The base station is configured to use the performance data counter to generate performance data corresponding to the key counter;

The network management server includes a data extraction module, a rule learning module, and a data storage module;

The data extraction module is connected to the network management database, and extracts the performance data corresponding to the key counter from the network management database;

The planning learning module is used to formulate an analysis model, and calculate the threshold value of the correlation coefficient of the pseudo base station according to the performance data corresponding to the key counter extracted from the network management database;

The data storage module is used to store data required to implement the pseudo base station identification method described in Embodiment 1 to Embodiment 6 of the present application.

In this embodiment, the network management server further includes: a user interaction module and a graphics management module;

The user interaction module is used to make a pseudo base station reminder through the connection of the base station and the user terminal;

The graphic management module is used to generate the movement trajectory of the pseudo base station.

In view of the actual situation, this embodiment also provides a schematic diagram of another pseudo base station identification system for reference. Refer to FIG. 8, which is a schematic diagram of another pseudo base station identification system provided by this embodiment. Figure 8 includes: user interaction module, data extraction module, data storage module, database, rule learning module and graphics management module. The rule learning module includes analysis model management module, threshold management module, counter model management module and base station model management module.

In order to ensure that the data in the network management database can cover a larger area and provide more data information, the network management database can be updated by connecting to an external system. The external systems connected to this system include the network management system and the external database. The network management system is responsible for providing historical performance data of the base station. The external database is responsible for providing information such as maps, road networks, and known pseudo base stations.

In Figure 8, the data extraction module includes: docking with the existing network network management system, extracting the required counters, base station location information, base station and cell correspondence, cell neighboring relationship and other information from the network management system database and converting them into the format required by the system And stored in the data storage module.

The data storage module includes: storage of historical data of performance counters, known pseudo base station location information, road network data, regression analysis model, judgment threshold, pseudo base station trajectory graph and other data.

The rule learning module includes: formulating an analysis model and calculating the threshold of the correlation coefficient based on historical performance data.

User interaction module includes: user interaction and display of pseudo base station trajectory.

The graphics management module includes: matching of base station location information with road network information, and drawing of pseudo base station trajectory graphs.

This embodiment also provides a computer-readable storage medium. The computer-readable storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the embodiments of the present application. One to the steps of the pseudo base station identification method described in the sixth embodiment.

Beneficial effect

The embodiments of the present invention provide a pseudo base station identification method, system, and computer-readable storage medium, so as to realize pseudo base station identification without involving user privacy. The pseudo base station identification method includes: determining the key counter from the performance data counter of the base station, and the data in the key counter is affected by the pseudo base station; establishing the corresponding regression analysis model according to the key counter; extracting the statistical data corresponding to the key counter from the network management database , And substituted into the regression analysis model for machine learning to determine the threshold of the correlation coefficient of the pseudo base station; calculate the performance data of the key counter and compare with the threshold to determine whether there is a pseudo base station. The identification of pseudo base stations by obtaining key counter data from the base station avoids problems involving user privacy. At the same time, machine learning is also introduced to determine pseudo base stations, making the identification of pseudo base stations more accurate.

It can be seen that those skilled in the art should understand that all or some of the steps, functional modules/units in the system, and devices in the methods disclosed above can be implemented as software (which can be implemented by computer program code executable by a computing device). ), firmware, hardware and their appropriate combination. In the hardware implementation, the division between functional modules/units mentioned in the above description does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may consist of several physical components. The components are executed cooperatively. Some physical components or all physical components can be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or as hardware, or as an integrated circuit, such as an application specific integrated circuit .

In addition, as is well known to those of ordinary skill in the art, communication media usually contain computer-readable instructions, data structures, computer program modules, or other data in a modulated data signal such as carrier waves or other transmission mechanisms, and may include any information delivery medium. Therefore, the present invention is not limited to any specific combination of hardware and software.

The above content is a further detailed description of the embodiments of the present invention in combination with specific implementations, and it cannot be considered that the specific implementations of the present invention are limited to these descriptions. For those of ordinary skill in the technical field to which the present invention belongs, several simple deductions or substitutions can be made without departing from the concept of the present invention, which should be regarded as belonging to the protection scope of the present invention.

Claims

A method for identifying pseudo base stations, including:

Determining a key counter from the performance data counter of the base station, and the data in the key counter is affected by the pseudo base station;

Establishing a corresponding regression analysis model according to the key counter;

The statistical data corresponding to the key counter is extracted from the network management database and substituted into the regression analysis model for machine learning to determine the threshold value of the correlation coefficient of the pseudo base station. The network management database contains the statistics of the performance data counters of all base stations. Performance data;

Calculate the performance data of the key counter and compare it with the threshold to determine whether the pseudo base station exists.
The pseudo base station identification method according to claim 1, wherein the base station includes a single base station or a base station in an area, and the key counter includes at least one type of performance data counter.
3. The pseudo base station identification method according to claim 2, wherein after determining that the pseudo base station exists, the method further comprises: determining the location of the pseudo base station, and drawing the movement trajectory of the pseudo base station;

The location of the pseudo base station is determined according to the location information of the pseudo base station found, and within a preset time interval, multiple location information of the pseudo base station is counted and connected to obtain the movement track of the pseudo base station.
3. The pseudo base station identification method according to claim 2, wherein the calculating the analysis data of the key counter of the base station and comparing the threshold value comprises:

Substituting the statistical data of the key counter into the regression analysis model to obtain performance data, and comparing the performance data with the threshold value to determine whether there is a pseudo base station.
The pseudo base station identification method according to claim 2, wherein the performance data counter of the base station includes: the average number of users, the number of successful executions of intra-cell simultaneous handover of the X2 interface between the base stations, and the inter-cell synchronization of the X2 interface between the base stations. The number of successful executions of frequency handover, the number of successful executions of inter-cell inter-frequency handover of the X2 interface between base stations, the number of successful inter-cell inter-frequency handovers of the X2 interface between base stations, and the successful execution of inter-cell inter-frequency handover of the S1 interface between base stations. The number of times, the number of successful executions of inter-cell inter-frequency handover of the S1 interface between base stations, the number of successful executions of inter-cell intra-frequency handover of the S1 interface between base stations, and the number of successful inter-cell intra-frequency handovers of the S1 interface between base stations.
3. The pseudo base station identification method according to claim 2, wherein the regression analysis model is a calculation formula corresponding to the key counter, and the result of the calculation formula is the pseudo base station correlation coefficient of the key counter.
7. The pseudo base station identification method according to claim 6, wherein the calculation formula further includes a reasonable coefficient, and the reasonable coefficient makes the reasonable coefficient close to the actual value through the machine learning.
7. The pseudo base station identification method according to any one of claims 1-7, wherein the performance data counter comprises a performance data counter in a base station or a performance data counter between base stations.
A pseudo base station identification system, including:

Network management database, which stores performance data counted by performance data counters of all base stations;

The base station is configured to use the performance data counter to generate performance data corresponding to the key counter;

Network management server, which includes:

The data extraction module is configured to connect with the network management database, and extract the performance data corresponding to the key counter from the network management database;

The planning learning module is configured to formulate an analysis model, and calculate the threshold of the correlation coefficient of the pseudo base station according to the performance data corresponding to the key counter extracted from the network management database;

The data storage module is configured to store the data required by the pseudo base station identification method according to any one of claims 1-8.
The pseudo base station identification system according to claim 9, wherein the network management server further comprises:

The user interaction module is configured to perform pseudo base station reminders through the connection between the base station and the user terminal;

The graphics management module is configured to generate the movement trajectory of the pseudo base station.
A computer-readable storage medium storing one or more programs, and the one or more programs can be executed by one or more processors to realize the pseudo base station according to any one of claims 1 to 8 Identify the steps of the method.