A method for monitoring a 5G wireless network health state

Abstract

The application discloses a kind of 5G wireless network health state monitoring method, belong to network optimization technical field, method includes: 5G wireless network data is sampled;Sampling frequency offset model is constructed;Through sampling frequency offset model, the sampling frequency offset value is calculated to determine the influence degree of sampling frequency offset to wireless traffic;Through a variety of analysis methods, extract the abnormal characteristics in 5G wireless network traffic data;According to sampling frequency offset value and abnormal characteristics, the health state of 5G wireless network is comprehensively judged.In the application, the sampling frequency offset value and abnormal characteristics of 5G wireless network data are extracted, and the health state of 5G wireless network is automatically and comprehensively judged according to the sampling frequency offset value and abnormal characteristics, which avoids mechanical one by one investigation, improves the efficiency of troubleshooting, and has high investigation accuracy, which can meet the quality of a large number of signals in peak period.

Description

Technical Field

[0001] This invention belongs to the field of network optimization technology, specifically relating to a method for monitoring the health status of a 5G wireless network. Background Technology

[0002] As 5G wireless signals have an increasingly significant impact on signal quality in cell access, existing network optimization solutions are gradually failing to meet the quality requirements of a large number of signals in cells during peak hours. The main reason is that while wireless performance access optimization relies on automated troubleshooting to locate problems before they can be addressed, current automated troubleshooting techniques often involve mechanical, step-by-step checks, resulting in low efficiency and accuracy, thus failing to meet the required signal quality standards for a large number of cells during peak hours. Summary of the Invention

[0003] The purpose of this invention is to provide a method for monitoring the health status of a 5G wireless network, which can solve the technical problems of low efficiency and low accuracy in troubleshooting existing methods, making it difficult to meet the quality requirements of a large number of signals in the cell during peak hours.

[0004] To solve the above-mentioned technical problems, the present invention is implemented as follows:

[0005] This invention provides a method for monitoring the health status of a 5G wireless network, including:

[0006] S101: Sample 5G wireless network data;

[0007] S102: Construct a sampling frequency offset model;

[0008] S103: Calculate the sampling frequency offset value using the sampling frequency offset model to determine the degree of impact of the sampling frequency offset on wireless traffic;

[0009] S104: Extracting abnormal features from 5G wireless network traffic data using multiple analysis methods;

[0010] S105: Determine the health status of the 5G wireless network based on the sampling frequency offset value and the abnormal characteristics.

[0011] Furthermore, S103 specifically includes:

[0012] S1031: Set the sampling time of the OFDM transmitter to be... The corresponding sampling time of the OFDM receiver is Calculate the sampling frequency offset value α l :

[0013] Formula 1

[0014] Where sin c() represents the sampling function, ;

[0015] S1032: Adjust the sampling frequency offset value α l Compared with the normal offset threshold of the sampling frequency, at the sampling frequency offset value α l When the sampling frequency offset exceeds the normal offset threshold, it is determined that the sampling frequency offset has a significant impact on wireless traffic, and the probability of OFDM system performance degradation is high.

[0016] Furthermore, the wireless network traffic data includes explicit data and implicit data. The explicit data includes: data such as faulty equipment manufacturers and online reviews, ratings, and comments on the equipment model, as well as descriptions of related fault information. The implicit data includes: abnormal software data deployed on the equipment during use, abnormal network data between the equipment's network port and the peer device, and abnormal software data.

[0017] Furthermore, the various analytical methods mentioned include: incremental method, process method, and 2 / 8 principle method.

[0018] Furthermore, S105 specifically includes:

[0019] S1051: Extract the first abnormal feature of the wireless network traffic data by a progressive method, extract the second abnormal feature of the wireless network traffic data by a process method, and extract the third abnormal feature of the wireless network traffic data by the 2 / 8 principle method.

[0020] S1052: Determine the health status of the 5G wireless network based on the sampling frequency offset value, the first abnormal feature, the second abnormal feature, and the third abnormal feature.

[0021] Furthermore, prior to S105, it also includes:

[0022] S106: Determine the priority of the incremental method, the process method, and the 2 / 8 principle method through cluster analysis.

[0023] Furthermore, S106 specifically includes:

[0024] S1061: Perform cluster analysis on the abnormal features extracted by the incremental method, the process method, and the 2 / 8 principle method respectively;

[0025] S1062: Calculate the RMSSTD index value, R-Square index value, and total difference value for the abnormal features extracted by the incremental method, process method, and 2 / 8 principle method, respectively.

[0026] S1063: Based on the sampling frequency offset value, the RMSSTD index value, the R-Square index value, and the total difference value, comprehensively determine the priority of the gradual method, the process method, and the 2 / 8 principle method.

[0027] Furthermore, S1062 specifically includes:

[0028] Calculate the RMS / TD index value according to Formula 2:

[0029] Formula 2

[0030] Where Si represents the sum of the standard deviations of the i-th variable in each group, and p represents the number of variables.

[0031] Calculate the R-Square index value according to Formula 3:

[0032] Formula 3

[0033] Where R_Square represents the R-Square index value, W represents the degree of difference within each group after clustering, B represents the degree of difference between each group after clustering, and T represents the total degree of difference of all data objects after clustering, T=W+B;

[0034] The total degree of difference T is calculated according to Formula 4:

[0035] Formula 4

[0036] Where p represents the number of variables and n represents the number of group members. This represents the overall average.

[0037] Furthermore, S1063 specifically includes:

[0038] Each analysis method is scored based on its frequency offset value, its RMSSTD index value, its R-Square index value, and its total difference value.

[0039] The frequency offset score, the RMSSTD score, the R-Square score, and the total difference score are added together to obtain the total scores for the progressive method, the process method, and the 2 / 8 principle method, respectively.

[0040] The priority of the incremental method, the process method, and the 2 / 8 principle method is determined by ranking the total score from highest to lowest.

[0041] Furthermore, S1063 also includes:

[0042] When multiple analysis methods have the same total score, the priority of the incremental method, the process method, and the 2 / 8 principle method is determined according to a preset order. Among these preset orders, the incremental method has the highest priority, the process method has the second highest priority, and the 2 / 8 principle method has the lowest priority.

[0043] In this embodiment of the invention, the sampling frequency offset value and abnormal features of 5G wireless network data are extracted. The health status of the 5G wireless network is automatically and comprehensively determined based on the sampling frequency offset value and abnormal features. This avoids mechanically checking one by one, improves the efficiency of troubleshooting, and has high accuracy. It can meet the quality requirements of a large number of signals in the cell during peak hours. Attached Figure Description

[0044] Figure 1 This is a flowchart illustrating a method for monitoring the health status of a 5G wireless network provided in an embodiment of the present invention.

[0045] The realization of the objective, functional characteristics and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0046] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0047] The following description, in conjunction with the accompanying drawings, details the method for monitoring the health status of a 5G wireless network provided by the present invention through specific embodiments and application scenarios.

[0048] Reference Figure 1 The diagram illustrates a flowchart of a method for monitoring the health status of a 5G wireless network according to an embodiment of the present invention.

[0049] This invention provides a method for monitoring the health status of a 5G wireless network, comprising:

[0050] S101: Sample 5G wireless network data.

[0051] In one possible implementation, preliminary preparations are required before sampling 5G wireless network data. These preparations mainly involve understanding relevant information about network planning and collecting relevant data, including but not limited to parameter configuration and principles, indicator definition formulas and assessment methods, call statistics, DT testing, and event alarms.

[0052] S102: Construct the sampling frequency offset model.

[0053] Sampling Frequency Offset (SFO) refers to the degree of mismatch between the transmitter oscillator and the receiver oscillator.

[0054] S103: Calculate the sampling frequency offset value using the sampling frequency offset model to determine the degree of impact of the sampling frequency offset on wireless traffic.

[0055] It should be noted that determining the impact of sampling frequency offset on wireless traffic can be used to judge whether the trend of wireless signal coverage and interference in a cell is improving, providing corresponding indicators for performance evaluation. In this embodiment of the invention, the sampling frequency offset value is one of the important indicators for evaluating and predicting whether wireless network traffic is abnormal.

[0056] In one possible implementation, S103 specifically includes:

[0057] S1031: Set the sampling time of the OFDM transmitter to be... The corresponding sampling time of the OFDM receiver is Calculate the sampling frequency offset value α l :

[0058] Formula 1

[0059] Where sin c() represents the sampling function, ;

[0060] OFDM (Orthogonal Frequency Division Multiplexing) is actually a type of MCM (Multi-Carrier Modulation). It achieves high-speed parallel transmission of serial data through frequency division multiplexing. OFDM has good resistance to multipath fading and can support multi-user access.

[0061] S1032: Set the sampling frequency offset value α l Compared with the normal offset threshold of the sampling frequency, at the sampling frequency offset value α lWhen the sampling frequency offset exceeds the normal offset threshold, it is determined that the sampling frequency offset has a significant impact on wireless traffic, and the probability of OFDM system performance degradation is high.

[0062] S104: Extract abnormal features from 5G wireless network traffic data using multiple analysis methods.

[0063] Call traffic statistics are an important part of network optimization, used to analyze and assess the health of the network and locate network problems.

[0064] Furthermore, wireless network traffic data includes explicit and implicit data. Explicit data includes: data on faulty equipment manufacturers and online reviews, ratings, and comments about the equipment model, as well as descriptions of related fault information. Implicit data includes: software anomaly data deployed on the equipment during use, network anomaly data between the equipment's network ports and other devices, and software anomaly data. This implicit data reveals clues for tracing the faults of the equipment and its reputation. However, implicit data also has certain problems, such as how to identify whether the user deployed it for their own use or gave it to other customers, which affects the tracing and tracking process.

[0065] Furthermore, various analytical methods are employed, including the incremental approach, the process approach, and the 2 / 8 principle.

[0066] Among these methods, there are several approaches: The incremental approach: starting with an overall analysis of network performance and gradually narrowing down the scope to pinpoint the problem. The process approach: analyzing the message flow and the relationships between corresponding counters to determine the stage at which the problem occurred and further investigating potential causes. The 2 / 8 rule (TOPN): addressing the most severe problems in the network first, those with the greatest impact on overall network performance.

[0067] S105: Determine the health status of the 5G wireless network based on the sampling frequency offset value and abnormal characteristics.

[0068] In this embodiment of the invention, the sampling frequency offset value and abnormal features of 5G wireless network data are extracted. The health status of the 5G wireless network is automatically and comprehensively determined based on the sampling frequency offset value and abnormal features. This avoids mechanically checking one by one, improves the efficiency of troubleshooting, and has high accuracy. It can meet the quality requirements of a large number of signals in the cell during peak hours.

[0069] In one possible implementation, S105 specifically includes:

[0070] S1051: Extract the first abnormal feature of wireless network traffic data using a progressive method, extract the second abnormal feature of wireless network traffic data using a process method, and extract the third abnormal feature of wireless network traffic data using the 2 / 8 principle method.

[0071] S1052: The health status of the 5G wireless network is determined by comprehensively considering the sampling frequency offset value, the first abnormal feature, the second abnormal feature, and the third abnormal feature.

[0072] In one possible implementation, the process further includes the following step before S105:

[0073] S106: Determine the priority of the incremental method, the process method, and the 2 / 8 principle method through cluster analysis.

[0074] Furthermore, S106 specifically refers to:

[0075] S1061: Cluster analysis was performed on the abnormal features extracted by the incremental method, the process method, and the 2 / 8 principle method, respectively.

[0076] S1062: Calculate the RMSSTD index value, R-Square index value, and total difference value for the abnormal features extracted by the incremental method, process method, and 2 / 8 principle method, respectively.

[0077] Among them, RMSSTD (Root-Mean-Square Standard Deviation) is the combined standard deviation of all variables in the group. The smaller the RMSSTD, the higher the similarity of individual objects within the group (cluster), and the better the clustering effect.

[0078] R-Square: The magnitude of the difference between groups after clustering, that is, the proportion of variance of the original data that can be explained by the clustering results. The larger the R-Square, the higher the dissimilarity between groups (clusters) and the better the clustering effect.

[0079] In one possible implementation, S1062 specifically includes:

[0080] Calculate the RMS / TD index value according to Formula 2:

[0081] Formula 2

[0082] Where Si represents the sum of the standard deviations of the i-th variable in each group, and p represents the number of variables.

[0083] Calculate the R-Square index value according to Formula 3:

[0084] Formula 3

[0085] Where R_Square represents the R-Square index value, W represents the degree of difference within each group after clustering, B represents the degree of difference between each group after clustering, and T represents the total degree of difference of all data objects after clustering, T=W+B;

[0086] Calculate the total degree of difference T according to Formula 4:

[0087] Formula 4

[0088] Where p represents the number of variables and n represents the number of group members. This represents the overall average.

[0089] S1063: Based on the sampling frequency offset value, RMSSTD index value, R-Square index value, and total difference value, comprehensively determine the priority of the gradual method, the process method, and the 2 / 8 principle method.

[0090] In one possible implementation, S1063 specifically includes:

[0091] Each analysis method is scored based on its sampling frequency offset value, its RMSSTD index value, its R-Square index value, and its overall difference score.

[0092] The scores for frequency offset, RMSSTD, R-Square, and total difference are added together to obtain the total scores for the incremental method, the process method, and the 2 / 8 principle method.

[0093] The priority of the incremental method, the process method, and the 2 / 8 principle method is determined by ranking the total score from highest to lowest.

[0094] Furthermore, S1063 also includes:

[0095] When multiple analysis methods have the same total score, the priority of the incremental method, the process method, and the 2 / 8 principle method is determined according to a preset order. Among them, the incremental method has the highest priority, the process method has the second highest priority, and the 2 / 8 principle method has the lowest priority.

[0096] For example, refer to Table 1, which provides a possible clustering analysis table.

[0097] Table 1 Cluster Analysis Table

[0098]

[0099] In Table 1, the frequency offset score for the progressive method is 0, the RMSSTD score is 1, the R-Square score is 2, and the total difference score is 3; the frequency offset score for the process method is 1, the RMSSTD score is 2, the R-Square score is 9, and the total difference score is 1; the frequency offset score for the 2 / 8 principle method is 1, the RMSSTD score is 3, the R-Square score is 3, and the total difference score is 2. The RMSSTD score, R-Square score, and total difference score are added together to obtain the incremental approach (6 points), the process approach (7 points), and the 2 / 8 principle approach (7 points). At this point, the incremental approach is determined to have the lowest priority according to the total score from highest to lowest. For the process approach and the 2 / 8 principle approach, which both have a score of 7, their priorities need to be determined according to a preset order. In the preset order, the incremental approach has the highest priority, the process approach has the second highest priority, and the 2 / 8 principle approach has the lowest priority. Therefore, although both have a score of 7, the process approach has a higher priority than the 2 / 8 principle approach.

[0100] The above are merely embodiments of the present invention and are not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of the present invention should be included within the scope of the claims of the present invention.

Claims

1. A method for monitoring the health status of a 5G wireless network, characterized in that, include: S101: Sample 5G wireless network data; S102: Construct a sampling frequency offset model; S103: Calculate the sampling frequency offset value using the sampling frequency offset model to determine the degree of impact of the sampling frequency offset on wireless traffic; S104: Extracting abnormal features from 5G wireless network traffic data using multiple analysis methods; S105: Determine the health status of the 5G wireless network based on the combined sampling frequency offset value and the abnormal characteristics; Specifically, S103 includes: S1031: Set the sampling time of the OFDM transmitter to be... The corresponding sampling time of the OFDM receiver is Calculate the sampling frequency offset value α l : Official 1 Where sin c() represents the sampling function, ; S1032: Adjust the sampling frequency offset value α l Compared with the normal offset threshold of the sampling frequency, at the sampling frequency offset value α l When the sampling frequency offset exceeds the normal offset threshold, it is determined that the sampling frequency offset has a significant impact on wireless traffic, and the probability of OFDM system performance degradation is high. Specifically, S105 includes: S1051: Extract the first abnormal feature of the wireless network traffic data by a progressive method, extract the second abnormal feature of the wireless network traffic data by a process method, and extract the third abnormal feature of the wireless network traffic data by the 2 / 8 principle method. S1052: Determine the health status of the 5G wireless network based on the sampling frequency offset value, the first abnormal feature, the second abnormal feature, and the third abnormal feature.

2. The method for monitoring the health status of a 5G wireless network according to claim 1, characterized in that, The wireless network traffic data includes explicit data and implicit data. The explicit data includes: faulty equipment manufacturers and online reviews, ratings, comments on the equipment, and descriptions of related fault information. The implicit data includes: abnormal software data deployed on the equipment during use, abnormal network data between the equipment's network port and the peer device, and abnormal software data.

3. The method for monitoring the health status of a 5G wireless network according to claim 1, characterized in that, The various analytical methods mentioned include: incremental method, process method and 2 / 8 principle method.

4. The method for monitoring the health status of a 5G wireless network according to claim 3, characterized in that, Prior to S105, it also included: S106: Determine the priority of the incremental method, the process method, and the 2 / 8 principle method through cluster analysis.

5. The method for monitoring the health status of a 5G wireless network according to claim 4, characterized in that, Specifically, S106 is: S1061: Perform cluster analysis on the abnormal features extracted by the incremental method, the process method, and the 2 / 8 principle method respectively; S1062: Calculate the RMSSTD index value, R-Square index value, and total difference value for the abnormal features extracted by the incremental method, process method, and 2 / 8 principle method, respectively. S1063: Based on the sampling frequency offset value, the RMSSTD index value, the R-Square index value, and the total difference value, comprehensively determine the priority of the gradual method, the process method, and the 2 / 8 principle method.

6. The method for monitoring the health status of a 5G wireless network according to claim 5, characterized in that, S1062 specifically includes: Calculate the RMS / TD index value according to Formula 2: Official 2 Where Si represents the sum of the standard deviations of the i-th variable in each group, and p represents the number of variables; Calculate the R-Square index value according to Formula 3: Official 3 Where R_Square represents the R-Square index value, W represents the degree of difference within each group after clustering, B represents the degree of difference between each group after clustering, and T represents the total degree of difference of all data objects after clustering, T=W+B; The total degree of difference T is calculated according to Formula 4: Official 4 Where p represents the number of variables and n represents the number of group members. This represents the overall average.

7. The method for monitoring the health status of a 5G wireless network according to claim 6, characterized in that, S1063 specifically includes: Each analysis method is scored based on its frequency offset value, its RMSSTD index value, its R-Square index value, and its total difference value. The frequency offset score, the RMSSTD score, the R-Square score, and the total difference score are added together to obtain the total scores for the progressive method, the process method, and the 2 / 8 principle method, respectively. The priority of the incremental method, the process method, and the 2 / 8 principle method is determined by ranking the total score from highest to lowest.

8. The method for monitoring the health status of a 5G wireless network according to claim 7, characterized in that, S1063 further includes: When multiple analysis methods have the same total score, the priority of the incremental method, the process method, and the 2 / 8 principle method is determined according to a preset order. Among these preset orders, the incremental method has the highest priority, the process method has the second highest priority, and the 2 / 8 principle method has the lowest priority.

Images