Traffic suppression identification method and device, equipment, storage medium and computer program product
By acquiring and cleaning the engineering parameter data and performance data of the target cell, a linear regression model is constructed. By comprehensively considering multiple factors, traffic suppression identification is performed, which solves the problem of low identification accuracy in the existing technology and achieves higher identification accuracy.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHINA MOBILE GROUP DESIGN INST
- Filing Date
- 2025-12-10
- Publication Date
- 2026-04-24
AI Technical Summary
Existing technologies, when identifying traffic suppression, only use the number of users as a single variable, failing to fully consider other key factors that affect user perception, resulting in low identification accuracy.
By acquiring the target cell's engineering parameters, performance data, and traffic data, cleaning and correlation processing are performed, and a linear regression model is constructed. Taking into account factors such as the average number of users in the cell, physical resource block utilization, control channel occupancy rate, and allocation failure rate, traffic suppression identification is performed.
It improves the accuracy of traffic suppression identification, avoids interference from missing or abnormal data on the results, and ensures the comprehensiveness and accuracy of the identification.
Smart Images
Figure CN121924518A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of traffic suppression identification, and in particular to a traffic suppression identification method, apparatus, device, storage medium, and computer program product. Background Technology
[0002] With the popularization of 5G networks, the number of users and data traffic are growing explosively, and the types of services carried by mobile communication networks are becoming increasingly diverse. Against this backdrop, users' demand for wireless data services is no longer limited to basic signal coverage, but rather they are paying more attention to the stability, speed and overall experience of network usage.
[0003] Against this backdrop, users' demands for wireless data services are no longer limited to basic signal coverage, but rather they are paying more attention to the stability, speed, and overall experience of network usage.
[0004] However, in densely populated areas such as popular commercial districts and large venues, the data traffic carried by 5G base stations often experiences explosive growth, easily leading to network resource strain and subsequent traffic suppression. Specifically, when cell traffic exceeds a certain threshold, user experience speeds significantly decrease, resulting in issues such as video stuttering, slow downloads, and unstable connections, severely impacting user experience and service quality.
[0005] Currently, several technical solutions have attempted to identify and assess network traffic suppression. For example, one publicly disclosed method for detecting mobile network traffic suppression uses clustering and linear fitting techniques, with the number of connected users as the primary variable, to determine whether traffic suppression exists in a cell. While this method can identify traffic-limited situations to some extent, its model construction relies mainly on the number of users as a single variable, failing to fully consider other key factors affecting user perception. This results in limitations in the comprehensiveness of analysis and the accuracy of identification, making it difficult to accurately reflect the combined impact of service and control channels in actual networks.
[0006] Therefore, it is evident that how to achieve a traffic suppression identification method that can adapt to complex and ever-changing 5G network scenarios while taking into account user experience has become a technical problem that needs to be solved by existing technologies. Summary of the Invention
[0007] This invention provides a method, apparatus, device, storage medium, and computer program product for identifying traffic suppression, in order to solve the problem that existing technologies, when identifying traffic suppression, only use the number of users as a single variable and fail to fully consider other key factors affecting user perception, resulting in low identification accuracy.
[0008] This invention provides a method for identifying traffic suppression, comprising: Obtain the operating parameters, performance data, and traffic data of the target cell for each sub-time period; The engineering parameter data, performance data, and traffic data are cleaned and correlated to determine the correlated dataset; Traffic suppression is identified in the associated dataset using a preset recognition model.
[0009] A flow suppression detection method provided by the present invention, The engineering parameter data includes the average number of users in the community; The performance data includes: Average utilization of uplink physical resource blocks (PRBs); Average utilization of downlink physical resource blocks (PRBs); Physical downlink control channel (PDCCH) channel control channel element (CCE) occupancy rate; Uplink channel control channel element line CCE allocation failure rate; Downlink channel control channel element (CCE) allocation failure rate.
[0010] The traffic data includes: Uplink packet data aggregation protocol layer (PDCP) traffic; Downlink packet data aggregation protocol layer (PDCP) traffic.
[0011] According to a flow suppression detection method provided by the present invention, the step of cleaning and correlating the operating parameter data, the performance data, and the flow data to determine the correlating dataset includes: The working parameter data and performance data of each sub-period are sequentially associated to obtain an initial associated dataset, which includes the initial associated data corresponding to each sub-period. The first associated dataset is obtained by deleting any one or more of the following initial associated datasets: the average number of users in the cell, the average utilization rate of the uplink PRB, the average utilization rate of the downlink PRB, and the occupancy rate of the PDCCH channel CCE. Determine whether the uplink CCE allocation failure rate and / or the downlink CCE allocation failure rate of the initial associated data corresponding to each sub-time period in the first associated dataset are missing; When the judgment result is yes, the missing uplink CCE allocation failure rate and / or downlink CCE allocation failure rate in the initial associated data are supplemented by specified values to obtain the associated dataset.
[0012] According to the present invention, a method for detecting traffic suppression includes: the preset model is a linear regression model constructed by an intercept term, regression coefficients, and a random error term; The step of using a preset identification model to identify traffic suppression in the associated dataset includes: using the engineering parameter data and performance data of all data entries in the associated dataset as independent variables, and the traffic data of all data entries in the associated dataset as dependent variables, fitting the model through the linear regression model; if the proportion of entries whose fitted values deviate from the actual values by a first preset value to the number of data entries in the associated dataset is greater than a second preset value, then the cell is considered to have traffic suppression.
[0013] According to a traffic suppression detection method provided by the present invention, before performing traffic suppression identification on the associated dataset using a preset identification model, the method includes: The data in the associated dataset are arranged according to the average number of users in the community, and the first N data entries are selected to form the first dataset; The data in the associated dataset are arranged according to the average utilization rate of the uplink PRB or the average utilization rate of the downlink PRB, and the first N data entries are selected to form the second dataset. The data in the associated dataset are arranged according to the uplink PDCP layer traffic or the downlink PDCP layer traffic, and the first N data entries are selected to form a third dataset. A model training sample set is determined based on the first dataset, the second dataset, and the third dataset, and the preset recognition model is trained using the model training sample set; Where N is an integer greater than or equal to 2.
[0014] The present invention also provides a flow suppression identification device, comprising: The data acquisition unit is used to acquire the operating parameters, performance data, and traffic data of the target cell for each sub-time period; The association processing unit is used to clean and associate the engineering parameter data, the performance data, and the traffic data to determine the associated dataset. The suppression identification unit is used to identify traffic suppression in the associated dataset using a preset identification model.
[0015] According to a flow suppression identification device provided by the present invention, the association processing unit is specifically used for: The working parameter data, performance data, and traffic data of each sub-period are sequentially correlated to obtain an initial correlation dataset, which includes the initial correlation data corresponding to each sub-period. The first associated dataset is obtained by deleting any one or more of the following initial associated datasets: the average number of users in the cell, the uplink PDCP layer traffic, the downlink PDCP layer traffic, the average utilization rate of the uplink PRB, the average utilization rate of the downlink PRB, and the CCE occupancy rate of the PDCCH channel. Determine whether the uplink CCE allocation failure rate and / or the downlink CCE allocation failure rate of the initial associated data corresponding to each sub-time period in the first associated dataset are missing; When the judgment result is yes, the missing uplink CCE allocation failure rate and / or downlink CCE allocation failure rate in the initial associated data are supplemented by specified values to obtain the associated dataset.
[0016] The present invention also 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 any of the traffic suppression identification methods described above.
[0017] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements any of the above-described traffic suppression identification methods.
[0018] The present invention also provides a computer program product, including a computer program that can be stored on a non-transitory computer-readable storage medium, characterized in that the computer program, when executed by a processor, implements any of the above-described traffic suppression identification methods.
[0019] This invention provides a traffic suppression identification method, apparatus, device, storage medium, and computer program product, which has at least the following beneficial effects: When constructing the identification model, parameters related to user perception and the number of common users are considered together as dependent variables, and the comprehensive analysis of traffic factors by integrating information from the service channel and control channel is more comprehensive. Before traffic identification, the data is correlated, completed, and filtered, which effectively avoids the interference caused by missing or abnormal data on the results, and helps to improve the accuracy of traffic suppression identification. Attached Figure Description
[0020] Figure 1 This is one of the schematic diagrams illustrating the flow suppression identification method of the present invention; Figure 2 This is a second schematic diagram of the flowchart of an embodiment of the traffic suppression identification method of the present invention; Figure 3 This is a schematic diagram of the specific structure of the flow suppression identification device of the present invention; Figure 4 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation
[0021] The technical solutions in this application will now be described with reference to the accompanying drawings.
[0022] This invention provides a method for identifying traffic suppression, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of a traffic suppression identification method according to the present invention.
[0023] In this embodiment, the traffic suppression identification method includes the following steps: Step S10: Obtain the operating parameters, performance data, and traffic data for each sub-time period of the target cell.
[0024] It should be noted that the execution subject of the method in this embodiment can be a device with data acquisition, program execution, and data processing functions, such as a computer or control computer, or other devices that can achieve the same or similar functions. This embodiment does not impose specific limitations on this. The operating parameter data, performance data, and traffic data of the target cell in each sub-time period can come from different systems. For example, the operating parameter data may come from the network management system, and the operating parameter data and traffic data may come from the northbound automatic interface of the base station. This embodiment does not impose specific limitations on this.
[0025] It is worth noting that the target cell can be any of the 2G, 3G, 4G, and 5G cells that conform to the 3GPP mobile specifications; this embodiment does not impose specific restrictions on this. This embodiment will use a 5G cell as an example for explanation. The target cell is uniquely identified based on CGI (Cell Global Identity).
[0026] It is understood that a sub-time period can be a fixed duration, such as 1 minute, 1 hour, or 1 day, and this embodiment does not impose specific limitations on it. In this embodiment and the embodiments described below, a sub-time period of 1 hour will be used as an example for explanation. The time range for acquiring the operating parameter data, performance data, and traffic data of each sub-time period of the target cell can be 1 week, 1 month, 1 year, etc., and this embodiment does not impose specific limitations on it. In this embodiment and the embodiments described below, a time range of 1 week will be used as an example for explanation. That is, in the description of this embodiment, within the 1-week collection time range, operating parameter data, performance data, and traffic data of 7*24=168 sub-time periods of the target cell should be acquired respectively.
[0027] In this embodiment, the engineering parameter data includes: the average number of users in the cell.
[0028] In this embodiment, the performance data includes: average utilization of uplink physical resource blocks (PRBs), average utilization of downlink physical resource blocks (PRBs), occupancy rate of physical downlink control channel (PDCCH) channel control channel element (CCE), uplink channel control channel element (PDCCH) line CCE allocation failure rate, and downlink channel control channel element (PDCCH) CCE allocation failure rate.
[0029] In this embodiment, the traffic data includes: uplink packet data aggregation protocol layer (PDCP) traffic and downlink packet data aggregation protocol layer (PDCP) traffic.
[0030] Please refer to Table 1 for details.
[0031] Table 1
[0032] Step S20: Clean and correlate the engineering parameter data, the performance data, and the traffic data to determine the correlated dataset.
[0033] It should be noted that in actual production environments, due to various reasons such as software or hardware issues, the operating parameter data, performance data, and traffic data of a cell may have missing fields or entries. Therefore, data cleaning is necessary. Furthermore, since the operating parameter data, performance data, and traffic data of the target cell may originate from different systems, it is necessary to correlate these multiple data sources to create a unified format for analysis, forming a correlated dataset. In one implementation, this can be done as follows: Figure 2 The method shown performs association processing, specifically including the following sub-steps: Sub-step S21: Sequentially associate the working parameter data, the performance data, and the traffic data of each sub-period to obtain an initial associated dataset, which includes the initial associated data corresponding to each sub-period.
[0034] It should be noted that since the target cell's operating parameter data, performance data, and traffic data may come from different systems, it is necessary to correlate the various cleaned data according to sub-time periods to form data in a unified format. Specifically, the sub-time periods in each original data collection are used as the key values for correlation; that is, multiple data with the same sub-time period value are correlated into initial correlated data in a unified format, as shown in Table 2. Table 2
[0035] The initial associated dataset is composed of the initial associated data from all sub-time periods within the collection time range.
[0036] Sub-step S22: Delete the initial association data that is missing any one or more of the following from the initial association dataset: the average number of users in the cell, the average utilization rate of the uplink PRB, the average utilization rate of the downlink PRB, and the occupancy rate of the PDCCH channel CCE, to obtain the first association dataset.
[0037] Specifically: If any one or more of the following data are missing from a certain initial association data in the initial association dataset: average number of users in the cell, uplink PDCP layer traffic, downlink PDCP layer traffic, average uplink PRB utilization, average downlink PRB utilization, or PDCCH channel CCE occupancy rate, then that initial association data is deleted from the initial association dataset to obtain the first association dataset.
[0038] Sub-step S23: Determine whether the uplink CCE allocation failure rate and / or the downlink CCE allocation failure rate of the initial associated data corresponding to each sub-period in the first associated dataset are missing; when the determination result is yes, supplement the missing uplink CCE allocation failure rate and / or the downlink CCE allocation failure rate in the initial associated data with specified values to obtain the associated dataset.
[0039] Specifically: In this embodiment, if the uplink CCE allocation failure rate and the downlink CCE allocation failure rate are missing in a certain piece of associated data in the first associated dataset, then the missing uplink CCE allocation failure rate and the downlink CCE allocation failure rate are filled with 0.
[0040] After performing the above completion process on all sub-time period data in the first associated dataset, the associated dataset is obtained.
[0041] It is worth noting that, to ensure the overall validity of the data and the accuracy of traffic suppression identification, the number of data entries in the associated dataset of the target cell should be at least 60% greater than or equal to the number of data entries in the initial associated dataset of the target cell, i.e., 7 * 24 * 60% = 100 entries. The more data entries, the higher the accuracy of traffic suppression identification. If the number of data entries in the associated data of the target cell is less than 100, the target cell is considered to have data quality problems and the data is invalid. The data of the target cell should be analyzed and investigated, and the above steps S10, S21, and S22 should be repeated until the number of data entries in the associated data of the target cell is greater than or equal to 100, at which point the associated dataset is considered valid.
[0042] Step S30: Use a preset recognition model to identify traffic suppression in the associated dataset.
[0043] It should be noted that downlink PDCP traffic and uplink PDCP traffic are relatively independent scenarios. Therefore, the same method is used to build models for uplink and downlink data respectively. This embodiment uses downlink data as an example for illustration; the model and identification method for uplink data can be referenced from that for downlink data.
[0044] It is worth noting that the preset model in this embodiment can be any model that achieves data fitting, such as a linear regression model, a multinomial fitting model, a spline fitting model, a logistic regression / Poisson regression model, etc. This embodiment does not impose specific limitations on this. This embodiment uses a linear regression model as an example for illustration.
[0045] The preset identification model in this embodiment is a linear regression model constructed with an intercept term, multiple regression coefficients, multiple independent variables, and a random error term. The model uses the engineering parameter data and the performance data as independent variables and the traffic flow data as the dependent variable. Through linear fitting of the independent variables and the dependent variable, the identification of traffic suppression is achieved.
[0046] Specifically, the formula for this linear regression model is: ; in: Dependent variable: It represents downlink PDCP layer traffic, reflects the actual load of user data transmission, and is a core indicator for measuring the efficiency of network resource utilization.
[0047] Independent variable: This represents the average number of users in a community, which directly affects network traffic; an increase in the number of users is usually accompanied by an increase in traffic. This represents the average utilization rate of downlink PRBs, reflecting the density of wireless resource allocation. High utilization rates may lead to congestion or traffic bottlenecks. This indicates the CCE occupancy rate of the PDCCH channel, which measures the consumption of control signaling resources. A high occupancy rate may limit data transmission resources. This indicates the downlink CCE allocation failure rate, reflecting the degree of insufficient control channel resources, which may indirectly affect the data transmission success rate and throughput. Other parameters and error terms: This represents the intercept term, indicating the baseline flow rate when all independent variables are 0. , , , Represents the regression coefficients, reflecting the marginal effect of each independent variable on the dependent variable; Let represent the random error term, which is assumed to follow a normal distribution with a mean of 0.
[0048] Furthermore, based on the above recognition model, calculations are performed on all associated data entries in the associated dataset. The specific calculation formula is as follows: (fitted value - actual value) / fitted value.
[0049] If the percentage exceeds 5%, the associated data entry is added to set P; the proportion of entries in P to the total number of data entries in the associated dataset is calculated. If the percentage exceeds 10%, the cell is considered to have traffic suppression. This ensures that traffic suppression is not an isolated incident or interference from inaccurate data statistics. For cells identified as having traffic suppression, the maximum value and sum of (fitted value - actual value) among all data entries in P are calculated as a measure of potential traffic suppression in the cell this week.
[0050] Furthermore, based on the information from the above process, a corresponding assessment of the target cell is provided, including the following types: a) There is no traffic suppression; b) Traffic suppression exists, and corresponding indicators of traffic suppression (maximum suppression value, total suppression amount, etc.) are provided. c) Data quality issues exist; further investigation is recommended. d) For any unidentified special scenarios, further analysis and investigation are recommended.
[0051] It is worth noting that, such as Figure 3 As shown, to ensure the accuracy of the recognition model, a model training dataset can be constructed before step S30 in this embodiment. Specifically: Step S40: Before performing traffic suppression identification on the associated dataset using the preset identification model, the following steps are included: The data in the associated dataset are arranged according to the average number of users in the community, and the first N data entries are selected to form the first dataset; The data in the associated dataset are arranged according to the average utilization rate of the uplink PRB or the average utilization rate of the downlink PRB, and the first N data entries are selected to form the second dataset. The data in the associated dataset are arranged according to the uplink PDCP layer traffic or the downlink PDCP layer traffic, and the first N data entries are selected to form a third dataset. A model training sample set is determined based on the first dataset, the second dataset, and the third dataset, and the preset recognition model is trained using the model training sample set; Where N is an integer greater than or equal to 2.
[0052] In this embodiment, to ensure the validity of the training data, the number of data entries (i.e., N) in the first dataset, the second dataset, and the third dataset is required to reach 2 / 3 of the total number of entries in the associated dataset. That is, after sorting according to different parameters, the first 2 / 3 of the data entries are selected to construct the first dataset, the second dataset, and the third dataset, respectively.
[0053] For data entries that are simultaneously in the first dataset, the second dataset, and the third dataset, it can be assumed that the traffic is low and the business pressure is low, and all of them correspond to the free growth range of traffic, and are used as training sets to build models.
[0054] Step S50: If the preset identification model satisfies the validity requirement, perform traffic suppression identification on the associated dataset.
[0055] Specifically, from the perspective of statistical indicators, the model is evaluated based on its fit. If the fit is too poor, it is identified as a data problem or cannot be judged. Therefore, the model is considered effective only when the determination coefficient of the preset identification model is greater than or equal to M.
[0056] In this embodiment, the formula for the determination coefficient is: ; in: , representing the total sum of squares; , representing the sum of squared residuals; Indicates the sample size used for fitting; Indicates the first One observation value; , representing the sample mean; , representing the model's predicted value.
[0057] It should be noted that M can take different values depending on factors such as scenario, target cell, and data collection. For example, M can take values of 0.1, 0.3, 0.5, 1, etc. This embodiment does not impose any restrictions on this.
[0058] if If the value is less than M, the model is considered inapplicable to the scenario of this cell, and the cell is identified as belonging to an unidentified special scenario, with further analysis and investigation recommended; conversely, if... If the value is greater than or equal to M, the model is considered to be effective in analyzing the scenario of that community.
[0059] In this embodiment, when constructing the traffic suppression identification model, parameters related to user perception and the number of common users are considered as dependent variables. This comprehensive analysis of traffic factors, integrating information from both service and control channels, provides a more complete picture. Before traffic identification, data is correlated, completed, and filtered to effectively avoid interference from missing or abnormal data, thus improving the accuracy of traffic suppression identification. Furthermore, before using the preset model for traffic suppression identification, the data in the correlated dataset is sorted and a model training set is constructed. The model's determination coefficient is used to ensure its effectiveness, further enhancing the accuracy of traffic suppression identification.
[0060] In one embodiment, this application also provides a traffic suppression identification device to address the problem that existing technologies, when identifying traffic suppression, only use the number of users as a single variable, failing to fully consider other key factors affecting user perception, resulting in low identification accuracy. A schematic diagram of the specific structure of this traffic suppression identification device is shown below. Figure 3 As shown, it includes: a data acquisition unit 31, an association processing unit 32, and a suppression recognition unit 33.
[0061] Among them, the data acquisition unit 31 is used to acquire the working parameter data, performance data and traffic data of each sub-time period of the target cell; The association processing unit 32 is used to clean and associate the engineering parameter data, the performance data, and the flow data to determine the associated dataset; The suppression identification unit 33 is used to identify traffic suppression in the associated dataset using a preset identification model.
[0062] In one implementation, the engineering parameter data includes the average number of users in the cell; the performance data includes: average utilization rate of uplink physical resource blocks (PRBs); average utilization rate of downlink physical resource blocks (PRBs); occupancy rate of physical downlink control channel (PDCCH) channel control channel element (CCE); uplink channel control channel element (CCE) allocation failure rate; downlink channel control channel element (CCE) allocation failure rate; the traffic data includes: uplink packet data aggregation protocol layer (PDCP) layer traffic; downlink packet data aggregation protocol layer (PDCP) layer traffic.
[0063] In one embodiment, the association processing unit 32 is specifically configured to: sequentially associate the engineering parameter data, the performance data, and the traffic data for each sub-time period to obtain an initial association dataset, the initial association dataset including the initial association data corresponding to each sub-time period; delete initial association data from the initial association dataset that is missing any one or more of the following: average number of users in the cell, uplink PDCP layer traffic, downlink PDCP layer traffic, average utilization rate of uplink PRB, average utilization rate of downlink PRB, and CCE occupancy rate of PDCCH channel, to obtain a first association dataset; determine whether the uplink CCE allocation failure rate and / or the downlink CCE allocation failure rate of the initial association data corresponding to each sub-time period in the first association dataset are missing; when the determination result is yes, supplement the missing uplink CCE allocation failure rate and / or downlink CCE allocation failure rate in the initial association data by a specified value to obtain the association dataset.
[0064] In one embodiment, the suppression identification unit 33 is specifically used to: use the engineering parameter data and performance data of all data entries in the associated dataset as independent variables, and the traffic data of all data entries in the associated dataset as dependent variables, and fit the linear regression model. If the proportion of entries whose fitted values deviate from the actual values by a first preset value to the number of data entries in the associated dataset is greater than a second preset value, then it is considered that there is traffic suppression in the cell.
[0065] also, Figure 4 A schematic diagram of the physical structure of an electronic device is provided. This electronic device may include a processor 810, a communications interface 820, a memory 830, and a communication bus 840. The processor 810, communications interface 820, and memory 830 communicate with each other via the communication bus 840. The processor 810 can call logical instructions stored in the memory 830 to execute the aforementioned traffic suppression identification method.
[0066] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is able to execute a traffic suppression identification method provided by the above methods.
[0067] In another aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to perform a traffic suppression identification method provided by the methods described above.
[0068] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.
Claims
1. A method for identifying traffic suppression, characterized in that, include: Obtain the operating parameters, performance data, and traffic data of the target cell for each sub-time period; The engineering parameter data, performance data, and traffic data are cleaned and correlated to determine the correlated dataset; Traffic suppression is identified in the associated dataset using a preset recognition model.
2. The flow suppression detection method according to claim 1, characterized in that, The engineering parameter data includes the average number of users in the community; The performance data includes: Average utilization of uplink physical resource blocks (PRBs); Average utilization of downlink physical resource blocks (PRBs); Physical downlink control channel (PDCCH) channel control channel element (CCE) occupancy rate; Uplink channel control channel element line CCE allocation failure rate; Downlink channel control channel element (CCE) allocation failure rate; The traffic data includes: Uplink packet data aggregation protocol layer (PDCP) traffic; Downlink packet data aggregation protocol layer (PDCP) traffic.
3. The flow suppression detection method according to claim 2, characterized in that, The step of cleaning and associating the engineering parameter data, performance data, and traffic data to determine the associated dataset includes: The working parameter data, performance data, and traffic data of each sub-period are sequentially correlated to obtain an initial correlation dataset, which includes the initial correlation data corresponding to each sub-period. The first associated dataset is obtained by deleting any one or more of the following initial associated datasets: the average number of users in the cell, the uplink PDCP layer traffic, the downlink PDCP layer traffic, the average utilization rate of the uplink PRB, the average utilization rate of the downlink PRB, and the CCE occupancy rate of the PDCCH channel. Determine whether the uplink CCE allocation failure rate and / or the downlink CCE allocation failure rate of the initial associated data corresponding to each sub-time period in the first associated dataset are missing; When the judgment result is yes, the missing uplink CCE allocation failure rate and / or downlink CCE allocation failure rate in the initial associated data are supplemented by specified values to obtain the associated dataset.
4. The flow suppression detection method according to claim 3, characterized in that, The preset model is a linear regression model constructed using an intercept term, regression coefficients, and a random error term. The step of using a preset recognition model to identify traffic suppression in the associated dataset includes: Using the engineering parameter data and performance data of all data entries in the associated dataset as independent variables, and the traffic data of all data entries in the associated dataset as dependent variables, the linear regression model is used for fitting. If the proportion of entries whose fitted values deviate from the actual values by a first preset value to the number of data entries in the associated dataset is greater than a second preset value, then the cell is considered to have traffic suppression.
5. A flow suppression detection method according to any one of claims 1 to 4, characterized in that, Before performing traffic suppression identification on the associated dataset using a preset identification model, the following steps are included: The data in the associated dataset are arranged according to the average number of users in the community, and the first N data entries are selected to form the first dataset; The data in the associated dataset are arranged according to the average utilization rate of the uplink PRB or the average utilization rate of the downlink PRB, and the first N data entries are selected to form the second dataset. The data in the associated dataset are arranged according to the uplink PDCP layer traffic or the downlink PDCP layer traffic, and the first N data entries are selected to form a third dataset; A model training sample set is determined based on the first dataset, the second dataset, and the third dataset, and the preset recognition model is trained using the model training sample set; Where N is an integer greater than or equal to 2.
6. A flow suppression identification device, characterized in that, include: The data acquisition unit is used to acquire the operating parameters, performance data, and traffic data of the target cell for each sub-time period; The association processing unit is used to clean and associate the engineering parameter data, the performance data, and the traffic data to determine the associated dataset. The suppression identification unit is used to identify traffic suppression in the associated dataset using a preset identification model.
7. The apparatus according to claim 6, characterized in that, The association processing unit is specifically used for: The working parameter data, performance data, and traffic data of each sub-period are sequentially correlated to obtain an initial correlation dataset, which includes the initial correlation data corresponding to each sub-period. The first associated dataset is obtained by deleting any one or more of the following initial associated datasets: the average number of users in the cell, the uplink PDCP layer traffic, the downlink PDCP layer traffic, the average utilization rate of the uplink PRB, the average utilization rate of the downlink PRB, and the CCE occupancy rate of the PDCCH channel. Determine whether the uplink CCE allocation failure rate and / or the downlink CCE allocation failure rate of the initial associated data corresponding to each sub-time period in the first associated dataset are missing; When the judgment result is yes, the missing uplink CCE allocation failure rate and / or downlink CCE allocation failure rate in the initial associated data are supplemented by specified values to obtain the associated dataset.
8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements a traffic suppression identification method as described in any one of claims 1 to 5.
9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements a traffic suppression identification method as described in any one of claims 1 to 5.
10. A computer program product comprising a computer program, the computer program being storeable on a non-transitory computer-readable storage medium, characterized in that, When the computer program is executed by the processor, it implements a traffic suppression identification method as described in any one of claims 1 to 5.